dynamic solid state lighting - mit media lab

Dynamic Solid State Lightingby

Matthew Aldrich

B.S., Electrical EngineeringYale University, 2004

Submitted to the Program in Media Arts and Sciences,School of Architecture and Planning,

in partial fulfillment of the requirements for the degree of,

Master of Science in Media Arts and Sciences

at the

Massachusetts Institute of Technology

June 2010

c© Massachusetts Institute of Technology 2010. All rights reserved.

Author

Program in Media Arts and SciencesMay 7, 2010

Certified by

Joseph A. ParadisoAssociate Professor

Program in Media Arts and SciencesThesis Supervisor

Accepted by

Pattie MaesAssociate Academic Head

Program in Media Arts and Sciences

Dynamic Solid State Lighting

by

Matthew Aldrich

Submitted to the Program in Media Arts and Sciences,School of Architecture and Planning,

on May 7, 2010, in partial fulfillment of therequirements for the degree of,

Master of Science in Media Arts and Sciences

Abstract

Energy conservation concerns will mandate near-future environments to regulate themselves toaccommodate occupants’ objectives and best tend to their comfort while minimizing energy con-sumption. Accordingly, smart energy management will be a needed and motivating application areaof evolving Cyber-Physical Systems, as user state, behavior and context are measured, inferred, andleveraged across a variety of domains, environments, sensors, and actuators to dynamically miti-gate energy usage while attaining implicit and explicit user goals. In this work, the focus in on theefficient control of a LED-based lighting network.

This thesis presents a first-of-its-kind pentachromatic LED-based lighting network that is ca-pable of adjusting its spectral output in response to ambient conditions and the user’s preferences.The control of the intensity is formulated as a nonlinear optimization problem and the mathemat-ics governing sensed illuminance, color, and corresponding control (feedback and adjustment) areformally defined. The prototype adjustable light source is capable of maintaining an average colorrendering index greater than 92 (nearly the quality of daylight) across a broad adjustable range(2800 K - 10,000 K) and offers two modes of control, one of which is an energy efficient mode thatreduces the total power consumption by 20%. The lighting network is capable of measuring theilluminance and color temperature at a surface and adjusting its output with an overall update rateof 11 Hz (limited by the MATLAB kernel). The sensor node features an optical suite of sensors witha dynamic range of 10000 : 1 lx (rms error: 2 lx). The sensor node measures the color temperatureof daylight within ±500K (kelvin). Device testing and validation were performed in a series of ex-periments in which the radiant power was collected using a radiometrically calibrated spectrometerwith an expanded uncertainty (k = 2) of 14% and validated against a model derived by measuringthe individual spectra of the system using custom MATLAB tools. A digital multimeter measuredthe current in the experiments. The work concludes by estimating the energy savings based on themeasured optical and electrical data. In environments with moderate ambient lighting, the net-worked control reduces power consumption by 44% with an additional 5-10% possible with spectraloptimization.

Thesis Supervisor: Joseph A. ParadisoTitle: Associate Professor, Program in Media Arts and Sciences

3

Dynamic Solid State Lighting

by

Matthew Aldrich

The following people served as readers for this thesis:

Thesis Reader

William J. MitchellProfessor


Thesis Reader

Ramesh RaskarAssociate Professor


5

Acknowledgments

There are many people and organizations to which I owe thanks and credit. To Joe Paradiso,for the opportunity to join the greatest research group at the Media Lab and second, for histime and guidance on helping me carve out this research area; my readers, Bill Mitchell andRamesh Raskar, for their support, patience, and flexibility; to Media Lab sponsors Philips-Color Kinetics and Schneider Electric for their pledges of support; to the Color Kineticsstaff in Burlington, whose generous donation of lighting equipment facilitated the creationof the first lighting testbed; in particular, Jeff Cassis at CK for his written support of theresearch and Jim Anderson, Paul Kennedy, John Warwick and Tracey Estabrook at CK;Phil London at Schneider for his written support; thanks to Peter Rombult for working onour grants; the Responsive Environments Group; Alex Reben, Mat Laibowitz, and GershonDublon for their help in the machine shop; Nan Wei Gong for Cargonet; to Mike Lapinski,Bo Morgan, and Clemens Satzger for the trips to Peoples’; to Nan Zhao for her work on theCK system; Laurel Pardue for her virtuosic musicianship; Mark Feldmeier for constructingthe first half of the building of the future; Spinner; and the ResEnv Country Tyme Band.

I owe a great deal of thanks to Dr. Peter Kindlmann, whose advice and guidance Ihave always counted on; Jack Rains, Jr. at Renaissance Lighting; for the opportunity towork directly in the design, manufacture, and research of solid state lighting; also, MikeGarbus and Steve Lyons who always provide sound engineering advice; credit goes to MaroSciacchitano and Josh Tor for their mechanical engineering prowess and the design of theintegrating sphere and heatsink; to Lisa Lieberson and Amna Carreiro for supporting theResponsive Environments Group, saying yes to overnight shipping, and for keeping the grouprunning smoothly.

This work is supported by funding from the MIT Media Laboratory.

I want to thank my family and Becky Davis for their love and support over the years.

7

Contents

Abstract 3

Acknowledgements 7

Contents 9

List of Figures 12

List of Tables 13

1 Introduction 141.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141.2 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2 Dynamic Lighting 172.1 Feedback-Controlled Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . 172.2 The Promise of Solid State Lighting . . . . . . . . . . . . . . . . . . . . . . . 192.3 White Light using LEDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202.4 Dynamic Solid State Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3 System Modeling 253.1 Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Estimating the Illuminance From a Point Source . . . . . . . . . . . . . . . . 25Room Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Sensing Changes to Operating Environment . . . . . . . . . . . . . . . . . . . 27Color Temperature and Color Rendering Index . . . . . . . . . . . . . . . . . 28Dynamic Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Formal Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Size and Scaling of the Search Space . . . . . . . . . . . . . . . . . . . . . . . 32

3.2 Linear Methods of Controlling the Spectral Power Distribution . . . . . . . . 32Exact Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33Overdetermined Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

9

3.3 Proposed Method of Control: Direct Search . . . . . . . . . . . . . . . . . . . 35Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Pattern Based Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Algorithm Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.4 Lighting Network Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Control by Optimization Workflow . . . . . . . . . . . . . . . . . . . . . . . . 37

3.5 Sensor Feedback in a Lighting Network . . . . . . . . . . . . . . . . . . . . . . 38

4 Physical Implementation 404.1 Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404.2 Sensor Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424.3 LED Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424.4 LED Array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5 Observations 495.1 Experiment Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495.2 LED Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Observing the Inverse Square Law . . . . . . . . . . . . . . . . . . . . . . . . 49Linearity and Superposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Experiment #1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51Experiment #2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

5.3 Theoretical Limits of Operation . . . . . . . . . . . . . . . . . . . . . . . . . . 51Predicted Results using Monte Carlo Simulations . . . . . . . . . . . . . . . . 58Predicted Results using Linear Methods . . . . . . . . . . . . . . . . . . . . . 58Predicted Results using Direct Search . . . . . . . . . . . . . . . . . . . . . . 58

5.4 Measured Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61Linear Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61Direct Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5.5 Sensor Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625.6 Estimating the Ambient Light . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

6 Analysis 686.1 Sources of Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Measurement Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Spectrometer Induced Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Error in Linearity Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . . 69

6.2 Performance of the Sensor Node . . . . . . . . . . . . . . . . . . . . . . . . . 706.3 Performance of Control Methods . . . . . . . . . . . . . . . . . . . . . . . . . 746.4 Comparison to Commercial Systems . . . . . . . . . . . . . . . . . . . . . . . 776.5 Energy Saving Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

10

7 Conclusions 827.1 Overview of Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

A Expanded Results 85

B Hardware Design 90B.1 LED Controller Hardware Schematics . . . . . . . . . . . . . . . . . . . . . . 90B.2 LED Controller PCB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95B.3 LED Controller Bill of Materials . . . . . . . . . . . . . . . . . . . . . . . . . 100B.4 Sensor Node Hardware Schematics . . . . . . . . . . . . . . . . . . . . . . . . 103B.5 Sensor Node PCB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108B.6 Sensor Node Bill of Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . 113B.7 LED Ring Schematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116B.8 LED Ring PCB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

C Firmware 123

D Radiometric Traceability 124

Bibliography 142

11

List of Figures

2.1 Levels of Abstraction in Lighting Control . . . . . . . . . . . . . . . . . . . . . . 24

3.1 Lighting Control Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4.1 Network Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414.2 Sensor Node and LED Controller Hardware . . . . . . . . . . . . . . . . . . . . . 444.3 LED Circuit Board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474.4 LED Heatsink and Dome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.1 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505.2 Distance Measurement Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525.3 LED Linearity Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535.4 Superposition of White Point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545.5 Measured LED Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555.6 Superposition of Irradiance and Effects of Temperature . . . . . . . . . . . . . . 565.7 Contour Plot of Five Wavelength Simulation . . . . . . . . . . . . . . . . . . . . 595.8 Contour Plot of Three Wavelength Simulation . . . . . . . . . . . . . . . . . . . . 605.9 Plot of Collected Lighting Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635.10 Dynamic Range of Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645.11 Dynamic Range of Digital Color Sensor . . . . . . . . . . . . . . . . . . . . . . . 65

6.1 Results of Removing Temperature Dependence . . . . . . . . . . . . . . . . . . . 716.2 Results of Digital Color Sensor Calibration . . . . . . . . . . . . . . . . . . . . . 736.3 Results of Measured Optical Parameters and Efficacy . . . . . . . . . . . . . . . 756.4 Results of Lux and Whitepoint Error . . . . . . . . . . . . . . . . . . . . . . . . 766.5 Office Illuminance Profile and Adaptive Lighting . . . . . . . . . . . . . . . . . . 80

A.1 Predicted Spectral Response Using Linear Methods . . . . . . . . . . . . . . . . . 86A.2 Measured Spectral Response For Linear Methods . . . . . . . . . . . . . . . . . . 87A.3 Predicted Spectral Response Using Nonlinear Methods . . . . . . . . . . . . . . . 88A.4 Measured Spectral Response for Nonlinear Methods . . . . . . . . . . . . . . . . 89

12

List of Tables

4.1 Bandwidth and Resolution of Irradiance Sensors . . . . . . . . . . . . . . . . . . 434.2 Dominant Wavelength and Flux of LEDs . . . . . . . . . . . . . . . . . . . . . . 46

5.1 Comparison of LED Array Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . 575.2 Predicted Results Using Linear Method . . . . . . . . . . . . . . . . . . . . . . . 585.3 Predicted Results Using Nonlinear Methods . . . . . . . . . . . . . . . . . . . . . 615.4 Nonlinear Solver Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615.5 Measured Linear Method Results . . . . . . . . . . . . . . . . . . . . . . . . . . 625.6 Measured Nonlinear Method Results . . . . . . . . . . . . . . . . . . . . . . . . . 625.7 Room Calibration Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

6.1 Estimated Flux and Efficacy of Prototype . . . . . . . . . . . . . . . . . . . . . . 78

13

Chapter One

Introduction

1.1 Motivation

We have the unique ability to modify and adapt to our surroundings, and for better orworse, shape our environment as we see fit. We live at a time where personal comfort isonly a switch, lever, or button press away. When there is not enough light, we flip a switch.When it is too hot or too cold, we immediately adjust the thermostat. As long as the lightis on or the room is cold, we are happy and for many of us, the problem ends the momentwe are gratified. What reason do we have to consider energy efficiency if our immediategoal is met? We are at the point where the word “green” hardly refers to a color anymore.What we read, watch, eat, and enjoy now comes with some reassurance that we are makingthe world better for ourselves and for our children. Yet it is not just the choices we make asconsumers that reflects our acute sensitivity to the color green. Being “green” means thatwe modify our behavior; we remember to turn off the lights, or shut down the A/C beforeleaving the house or office.

But, sometimes we forget. We create technology that attempts to minimize these mis-takes. We have programmable HVAC controllers and motion controlled lighting. We intro-duce new lighting technologies, such as compact florescent lighting (CFL) that are energyefficient, yet contain mercury and produce a low quality of light. Since the adoption ofincandescent technology, the benefits of adopting a more energy efficient light source hasbeen undermined by some implicit sacrifice in quality or control.

We have all had some experience in which the outcome of choosing the more energyefficient option fell short of our expectations. As consumers, we have become wary of greentechnologies. Instead, we prefer the solution that is immediate and attainable within ourown financial limits. We are conscious of solutions that contain the word “automatic” andaptly so, the weak automation found in modern lighting and HVAC technologies has leftthe building manager, the employees, and consumers confused. Why did this turn off?How do I program this? So we ignore these options. We leave the lights on and we turnthermostats to 50 F when we want to cool down quickly. We control the built environment

14

1.2. Contributions

using mechanisms that are inherently “deaf and mute” [22]. The disparity between “this iswhat I want” and “this is what I get” has led to inefficient control over utilities and errorladen user control. Adding feedback is one way we attempt to balance efficient operationand the user’s comfort.

In the academic community, the organization, feedback, control, and deployment ofthese types of distributed utility systems is a growing area of research, especially in theareas of sensor network utility control. In a broader sense, these intelligent systems areknown as cyber-physical systems (CPS) [34, 63]. In this thesis, the author argues thatthe system presented in this work embodies the very foundations of cyber-physical systemsthat actively monitor and regulate building utilities. Through the use of sensor networksand advanced control methodologies, the underpinnings of cyber-physical infrastructure arerealized: we move beyond reactive control, reduce human-in-loop error, and dynamicallyadjust and respond to changing physical and environmental conditions.

The goals of this thesis are two-fold. The primary goal is to thoroughly model, build,and test an intelligent lighting network capable of on-the-fly adjustments from both the userand the environment. No assumptions, other than a simple room calibration and a modelof the system, are required for run-time controlled optimization of the lighting parameters.The end result is a system capable of tuning its spectral output to not only achieve thecorrect intensity and color, but also selectively use less energy or maximize the quality oflight without user intervention.

In doing so, this work satisfies a second goal. While the efficacy of both active andphosphor based emitters will certainly improve, until we can modulate the band gap orreconfigure the LED phosphors on the fly, we will most likely turn to the control of multiplewavelengths to create precise and high quality light. The methods presented in this thesiscertainly apply to the control of a single fixture, but are general enough to apply to thecontrol of an entire network.

The next steps in utility management begin with the integration of these sensor richcyber-physical systems with hybrid lighting systems, but the approach must remain tech-nology agnostic. The less assumptions made about the operating environment, the better.For if the lighting network is to be truly transparent, then it must handle both the least andgreatest perturbation about its equilibrium with as little human intervention as possible.

1.2 Contributions

This thesis presents a first-of-its-kind pentachromatic LED-based lighting network that ad-justs its spectral output in response to environmental stimuli and the user’s preferences. Inthis work, the control of the wavelengths is formulated as a nonlinear optimization problemand to the best of the author’s knowledge, represents the first known published account ofusing this approach. This work extends previous research regarding sensor enabled lightcontrol, in that solid state lighting is considered to be the primary means of illumination.

15

1.2. Contributions

Consequently, the lighting network can dynamically optimize the light output for greaterefficiency or higher color quality depending on the measured ambient conditions. Whereasprevious research considered one linear parameter, intensity, we extend the control methodsto highly nonlinear photometric and color criteria that directly govern energy consumptionand visual quality. Additionally, we incorporate the ambient color temperature as a feed-back parameter and describe the performance of this method. In thoroughly presenting newmethods, insights, and approaches regarding the design and control of an intelligent lightingnetwork, a foothold in the area of sensor enabled utility management is established. In sum-mary, this thesis provides the foundation for future research in the areas of efficient lightingcontrol, dynamic utility regulation, persuasive energy, and human computer interaction.

16

Chapter Two

Dynamic Lighting

In this work, dynamic lighting is considered to be the control and feedback of a light sourceor group of light sources whose output directly reflects the measured ambient and user sup-plied operating conditions. In a larger context, the research of dynamic lighting draws fromseveral academic fields of study: applied optics and photonics, feedback and control systems,numerical optimization, electronics design, and human computer interaction. Dynamic light-ing is best labeled as a subset of research in the broad field of sensor network enabled utilitymanagement and control. In turn, these systems, intended to minimize wasted energy, re-duce human-in-loop error, and remain transparent to the user, represent a specific researcharea of cyber-physical systems (CPS) for intelligent utility management. This emerging fieldencapsulates any system that incorporates feedback and knowledge models of the environ-ment and user. In general, the goal of these intelligent systems is maximizing the utilityafforded by multiple networked subsystems whose augmented performance is derived frommeasuring or inferring the user’s needs and the present state of the operating environment.

2.1 Feedback-Controlled Lighting

Intelligent systems that compensate for varying optical and electrical properties, temper-ature stability, and the temporal variation of intensity in LEDs are crucial to achievingefficient and long lasting lighting. These systems have been identified as critical researchareas by the U.S. Department of Energy [45]. A typical solid-state lamp consists of fourmain components, each affecting the overall system efficiency. The driver is responsiblefor controlling the brightness of the LEDs. The LED package affects the efficiency of theLED device. The thermal efficiency refers to the loss of intensity due to thermal heatingof the LEDs, board, and electronics. Finally, the fixture optics sets the light loss due toinefficiencies or nonidealality in the reflector or secondary optics.

Increased interest from both academia and industry has led to numerous publicationsand research concerning efficient and optimal systems to drive LEDs. Van der Broek et al.[76] provide an overview of these topologies. LED intensity is controlled using either linear

17

2.1. Feedback-Controlled Lighting

current drivers or pulse width modulation (PWM). Green and blue LEDs shift towardslonger wavelengths, while red and yellow LEDs shift towards shorter wavelengths at lowerjunction temperatures. Therefore, PWM is the generally accepted method to control theintensity because current is held constant during the pulse and die temperature stability isimproved. Recently, Garcia et al. [23] published results on the overall effect of current andPWM fundamental frequency on the total luminous flux of systems.

As early is 2002, Muthu et al. incorporated optical and temperature feedback into thedriver architecture, allowing for compensation of optical, electrical and thermal variations ofthe system [42, 41]. Subsequent research evaluated the use of neural networks to maintainaccurate chromaticity [6] and explored optical feedback using three or more unique LEDwavelengths [41, 3]. However, in this work, a very simple (and ineffective) method of con-trolling the light output is presented for systems comprised of more than three wavelengths.This is achieved by fixing the ratio of intensity at one wavelength to another at the costof saturated colors and a decreased color gamut. This does not guarantee the best colorrendering index for a given white point.

Dynamic and adaptive lighting enabled by environmental sensors offers additional en-ergy savings. Early work by Crisp and Hunt in the 1970s focused on estimating internalilluminance from artificial and external light sources in order to reduce unnecessary lightingand excess energy expenditure [16, 28, 17, 29, 18]. This early work discussed the use ofphotosensors to monitor the natural daylight in the office place. As early as 1978, Crisp [17]was already considering repurposing the light switch, inspired by the affordances offered bytheir rudimentary incorporation of daylight harvesting. With the availability of low costlux sensors (perhaps more importantly, silicon improvements to reign in device variance)and an influx of low cost embedded devices, this simple form of intensity feedback was ex-tended to networks of fixed color incandescent and fluorescent lights. Most relevant is thework by Singhvi et al. [64], in which the group designed and tested closed loop algorithmsto maximize energy efficiency while meeting user lighting requirements in an incandescentlighting network. Park et al. [49] developed a lighting system to create high quality stagelighting to satisfy user profiles. Wen et al. [78] researched fuzzy decision making in lightingcontrol networks and Bayesian inference. Machado and Mendes [38] tested automatic lightcontrol using neural networks, as did Mozer in a pioneering study from 1992 [40], wherelight switches and thermostats in his house only provided reinforcement for a neural net-work that, also utilizing motion sensors, built a user activity model to automatically controlutilities.

In industry, Philips Lighting Research has investigated interactive lighting as well, yetthe sophistication of these systems is limited by cost and consumer interest. Recently,Philips Lighting (Sluis et al. [66]) produced a lighting demo in which a single device canadjust its output to match the color of a particular object. This device is not designedfor white lighting applications or networked control. Instead, its purpose is to enhance thechromaticity of merchandise on display at a retail location. The closest example to the

18

2.2. The Promise of Solid State Lighting

proposed work is Philip’s “Dynamic Lighting” suite of products [51]. Their basic conceptrevolves around a preprogrammed schedule of color temperature and lighting intensity tomimic a typical daylight cycle on Earth. However, the restrictions are rigid: the systemfeatures no feedback, requires programming, and offers only two color temperatures. Withthese shortcomings, it leaves plenty of open space for advances into this burgeoning field ofinteractive solid-state lighting control.

2.2 The Promise of Solid State Lighting

Each year, 915 TWh of energy powers incandescent and fluorescent technology in the US.This accounts for 79% of the total lighting energy consumption [45]. By comparing theluminous efficacy (the ratio of luminous flux to the input electric power) of incandescent,fluorescent, and LED technologies, it is clear that white LEDs have great potential to reduceenergy consumption and enhance the built environment. The potential savings in energyand reduction of environmental pollution by adopting solid state lighting (SSL) technologyis also well studied. Bergh et al. [12] provided early technical discussions of the economicbenefits of using LEDs for illumination. In a highly regarded paper concerning the futureuse of LEDs for illumination, Schubert et al. [61] provided a detailed study of how varyingadoption rates of solid state lighting reduced greenhouse gases, crude oil consumption, andincreased financial savings. The results of this study, known as the "5.5% and 11% scenarios"state that a 40% and 80% market penetration of solid state lighting technology in the UnitedStates would save $22.89 billion to $45.78 billion, depending on the adoption rate. Theelectrical energy used for lighting is decreased by 25% and 50% respectively. In the US,this translates into yearly savings of 12 million to 24 million barrels of oil and reductionof carbon emissions by 133 Mt to 267 Mt per year. Carbon dioxide emissions have alsobeen studied in [11] which highlights the adoption of solid state lighting an an attractivealternative to carbon capture and sequestration. Subsequent research focused on analyzingthe high initial cost of adopting solid state lighting. The US Department of Energy estimatesthe initial cost of an LED incandescent equivalent to be 560 times that of a traditionalincandescent [45]. While the initial cost is quite high, other methods of determining thetrue cost of the technology exist. The life-cycle cost or "cost of light" is another metric thattakes into account the operational lifetimes, labor cost, and energy usage. This is definedas dollars per kilolumen-hours ($/klm-hr). Azevedo, Morgan, and Morgan [11] provide adetailed analysis of both the initial cost and levelized annual cost of solid state lighting. Inaddition, their estimated economic, energy, and carbon emission savings are consistent withSchubert et al. The analysis concludes that rational consumers will find it cost effective toreplace incandescent and fluorescent lighting by 2010, and that solid state lighting will becompetitive with conventional lighting before 2015 [11, 21].

The principal focus of current LED research is on improving the maximum efficacyof phosphor based LEDs (white LEDs). At present, near-ultraviolet pumped phosphors

19

2.3. White Light using LEDs

remain the most efficient way to create white light. Yet, these devices, like their activeemitter counterparts, have their tradeoffs. In this research, a hybrid active and phosphorbased system is employed to create a color-adjustable, high quality white light source.

2.3 White Light using LEDs

Although initial cost impedes market penetration, solid state lighting presents new possi-bilities in terms of controlling the brightness, the spectral composition, and color renderingproperties of white light (the ability of an artificial light source to "render" the true colorsof an object) [61, 60, 59]. Although cost is important, the color rendering quality of thelight is known to be an important factor [72]. Indeed, the slow adoption of compact flu-orescent has, in part, been directly attributed to the fluorescent technology’s lower colorrendering ability and quality of light [26, 4, 58, 24]. LED-based lighting systems allow fordynamic and automatic modulation of these parameters to enhance the built environmentand provide additional energy savings [72, 77, 65]. An artificial white light source is judgedby two quantitative metrics, its Correlated Color Temperature or CCT (the temperaturein kelvin of an equivalent black body radiator) and its Color Rendering Index (CRI) [79].Warmer (red) colors have a lower color temperature and cooler (green and blue) colors havea higher color temperature. The CRI is a unitless measurement of how well an artificialilluminant renders the color of an object versus a perfect reference illuminant, where thereference CRI typically has a value of 100. Guo and Houser [25] review the application ofCRI for measurement of light sources. Davis and Ohno [20, 47] at the National Institute ofStandards and Technology (NIST) in the US are researching an improved color renderingmetric.

Boynton reviews the history and current status of research on color spaces derived fromhuman perception [14]. The Commission Internationale de L’Eclairage (CIE) approvedseveral color chromaticity spaces for mathematically representing colorimetric data. TheCIE 1931 XYZ color space is a common example (see Wyszecki and Stiles [79], pg. 137).This two dimensional space is based on a set of "color matching functions" that collectivelysum to unity. The outermost locus represents saturated primary colors and pure white lightfalls in the center of the space. The Plankian locus is added for a blackbody reference andthe vertical bars represent the correlated color temperature.

The optimal method to create white light using LEDs is still an open question, andwhile several strategies exist, there are tradeoffs between efficacy and color rendering ability[61]. White light can be created using either multiple current-injected emitters or by noncurrent-injected phosphor wavelength converters. The overall efficacy and color renderingability are determined by the LEDs selected. The spectral power of the system, defined bythe narrowband emission of the LEDs, determines the color rendering ability of the lightsource. Fewer unique wavelengths result in lower color rendering ability yet the efficacy ofthe source is often greater.

20

2.3. White Light using LEDs

Common approaches using active emitters include dichromatic (blue and yellow), tri-chomatic (red, green, blue), and tetrachromatic (red, green, blue and cyan or yellow) sources.These systems have been studied by Narendran et al. [43], Zukauskas et al. [81], Schubertand Kim [60], Stanikunas et al. [67], Lei et al. [35], Zukauskas et al. [82]. In this work theauthors report namely on the effects of dominant wavelength, correlated color temperature,and resulting color rendering index. Lei et al., uses a simplistic technique in the evaluationof four wavelengths and Zukauskas et al. [81] applies an exhaustive boundary search tomaximize certain optical properties of polychromatic light sources, yet does not provide anyinformation on completion time or provide any empirical results of the study. Ultimately, ineach of the papers, the authors provide the theoretical “best” combination of wavelengthsand intensities to maximize either the ratio of flux to radiant energy, or to maximize thecolor rendering index for a given color temperature. It is quite easy to stochastically deter-mine the right mixture of wavelengths using a Gaussian-based model of LED irradiance oncethe functions to measure the color rendering index and color temperature are programmed.

The problem is that these techniques do not correspond with the ranges of dominantwavelengths provided by the manufactures, nor do they take into account the variancein intensity and dominant wavelength across a particular bin of LEDs; in essence, theconclusions of these papers restrict the applicability of the results to either highly specializedconfigurations or particular laboratory and operating environments. While performing asensitivity analysis is one way to at least understand these effects, a better approach is toremain flexible from the the start, specifically, to concentrate on adaptive techniques thatrequire no assumptions of the system. For example, in [35], assumptions of the ratio of amberto red were made in order to resolve the system to three equations and three unknowns. Afixed ratio of these wavelengths to imposes harsh constraints on the range of achieveableCRI. The best ratio is nonlinear in the sense that the dominant wavelengths of all the LEDs,the desired color temperature, and required intensity must be taken into account. One finalconcern is the length of time required to find the appropriate solution. For example, in thisresearch, Monte Carlo simulations required up to eight hours to terminate. These techniquesdo not provide the run-time affordances required to control a solid state lighting network. Bytaking into account the variance of the LEDs and other uncontrollable parameters upfront(i.e., a method of control that requires no assumptions), we free ourselves from heuristicbased assumptions about the wavelengths and required intensity of the LEDs.

A different approach uses white LEDs of a fixed color temperature to achieve the illumi-nation goal. These devices do not require careful thought and control like active emitters,however there are other tradeoffs, which in turn, make active emitters a better option forcolor-adjustable lighting. White LEDs consist of an active emitter, primarily blue or nearultra violet and a down-converting phosphor. The first phosphor used in white LEDs ab-sorbed blue light and emitted yellow light typically between 5500 K and 6500 K [52, 33, 80].These devices provided adequate efficiency but low CRI. More recently, devices have utilizeda blue LED with a yellow and red phosphor. These resulted in warmer white LEDs with

21

2.4. Dynamic Solid State Lighting

improved CRI. However, the phosphor conversion always leads to an energy loss due tothe Stokes shift between absorption and emission. Scattering loss, or reflectance back ontothe chip or walls of the package also reduces efficiency. Stevenson [68] provides a recentoverview (2009) of Auger recombination (droop) and why LEDs become less efficient withincreased current. Present day efficacy and CRI for cool and warm white LEDs (not in afixture) are 101 lm/W and 70 (5500 K) and 72 lm/W and 90 (3300 K), respectively [45].Inversely, active emitter designs offer a lower efficacy, closer to 40 lm/W to 50 lm/W (dueto quantum efficiency and droop see Krames et al. [33]) but in some cases a higher colorrendering index. In this research, we present a state of the art light source capable of coloradjustable white points with an average color rendering index greater than 92.

There are practical limits to system performance. Burmen et al. [15] describe four ma-jor factors affecting the quality of LEDs: The initial variability of optical and electricalproperties, their temperature and electrical dependence, and temporal degradation withcorresponding variability. Initial optical and electrical properties such as luminous flux,chromaticity, and forward voltage can vary due to imperfections in the manufacturing pro-cess. Manufacturers typically "bin" LEDs to sub-classify the yield based on these threeproperties. However, higher flux bins add additional system cost. Within a particular bin,wavelength tolerance can adversely effect system performance. Lei et al. [35] reported thatred primaries of shorter wavelengths negatively affect CRI and that blue wavelength shiftsaffect the chromaticity. At the device level, the emission power, peak wavelength, and spec-tral width of LEDs vary with temperature. The luminous flux of low gap devices (red LEDs)decreases as the junction temperature rises, resulting in chromaticity and CRI shifts as thejunction temperature rises. To compensate in trichromatic and tetrachromatic systems, redLEDs are added, reducing efficacy and increasing cost. Finally, the temporal degradationof LEDs, even those within the same bin, can be different. Over time, this causes spatialcolor and intensity imhomogeneities in LED array light sources [15].

Blue and ultra violet pumped phosphor devices exhibit their own unique set of problems.Increases in junction temperature cause intensity degradation, however this effect is lessprominent in the high current region of the diode [15]. Narendran et al. studied the long termheating effects of white LEDs to properly evaluate their useful life and found that heatingof junction in white LEDs caused the yellowing of epoxy surrounding the die, resulting inchromaticity shift [43, 44]. Trevisanello et al. [75] also studied degration of white LEDs anddetermined that excessive heat at the junction led to a decrease in phosphor conversion.

2.4 Dynamic Solid State Lighting

While the optimal control of lighting networks has been researched in the past, it hasbeen done using traditional lighting technologies (e.g., incandescent and fluorescent). Thechallenges of efficient and optimal control of LED based lighting requires moving awayfrom the strictly linear techniques presented in [64, 78]. Although the goals of the user

22


may be satisfactorily represented as linear inequality constraints, the greatest challengeimposed by adopting LEDs is accounting for LED variance, temperature dependence, andthe difficulties of controlling a network comprised of a large number of variables (i.e., thenumber of wavelengths to be controlled). The control problem is further complicated byissues of optical sensor variance, occlusion of the sensor, and network update rates (e.g.,estimating dark lighting levels, polling sensor nodes, and preprocessing the data). To theuser, minimal control is required, as the lighting network must be capable of energy efficientoperation without the direct intervention and constant adjustment by the user. In previousresearch, the use of commercially available sensor nodes imposed limited flexibility andcontrol of these systems. Fixed resolution and no guarantees of timing or update rates wereamong the greatest problems faced by researchers who favored a top down approach.

Furthermore, as reported in [78], future systems should examine lighting conditions basedon tasks or inferred context – in cases where this fails or is inadequate, it is suggested thatmanual override or a user interface be incorporated into the control. In this research, weleverage multiple domains of expertise to design a custom lighting platform and novel sensornetwork to bridge the gap between ubiquitous sensing, user interfaces, and run-time control.Clearly, this is an ambitious undertaking, but doing so opens the opportunity to exploreideas and concepts unobtainable by previous research. For example, the system does notrequire fixed sensor positions, but in fact can dynamically recalibrate. Furthermore, weincorporate color measurements, color control, and closed loop LED control into our designusing polychromatic lighting.

In doing so, we embody the underpinnings of an intelligent system conscious of its outputand capable of reacting and interacting with dynamic environmental and human phenom-ena. By developing our tools and hardware from the ground up, we can begin to researchthe difficult challenges of cyber-physical systems: namely, the transparent integration ofubiquitous sensing and run-time control affordances needed to make these systems trulytransparent. This research is the first of many steps. In this embodiment, there is a prin-cipal focus on discovering and applying new methods of control and feedback in a lightingnetwork. Accordingly, the proposed techniques are mathematical and the intended purposeis to demonstrate their effectiveness by drawing conclusions from the physical measurements(e.g., energy, response speed, optical parameters). The natural progression is to greater lev-els of abstraction , where more emphasis is placed on interaction and true run-time control(see Figure 2.1). But first, at the lowest level, are the concepts, methods, and hardware thatcreate the foothold for these more complex systems, and it is these topics that are exploredin this thesis.

23


(a)

(b)

Figure 2.1: First-phase interactive testbeds to control illumination with diverse light sourcesexploiting mobile incident sensor/interfaces (a) and wearable first-person sensor/interfaces(b).

24

Chapter Three

System Modeling

In this chapter, we describe the control problem of estimating the illuminance from multiplelight fixtures in the presence of ambient light. We formally define the control problem as anoptimization and define the objective function subject to linear inequality constraints. Wereview existing linear methods of spectral control and propose a new method, direct search,to efficiently control the spectral output of an arbitrary number of wavelengths in a systemcomprised of an arbitrary number of light fixtures.

3.1 Problem Definition

In this section, a new method to control the spectral output of system comprised of anarbitrary number of light fixtures and wavelengths is presented. Setting the intensity andcolor of the lighting network is stated as an optimization problem, whereby the feasible setof solutions is determined by an objective function that either maximizes CRI or minimizesthe power consumption. The illuminance (Ev) is treated as an linear inequality constraint.

Estimating the Illuminance From a Point Source

First, consider a dark room in which there are n artificial light sources where each light sourceconsists of m wavelengths and j and i are indexing variables of the set j ∈ Z : 1 ≤ j ≤ nand i ∈ Z : 1 ≤ i ≤ m. If we consider the light source as a spherical point source, then forevery ith wavelength in each jth fixture there is a maximum luminous intensity designated,Iij . This is measured as candelas (lm · sr−1). In the presumed linear system, the intensity ofany ith wavelength in fixture j with corresponding maximum intensity Iij is set by columnvector x of dimesion 1 ×m where x ∈ Rm : 0 ≤ x ≤ 1. Therefore, the total candelas offixture j is the sum

∑xijIij where 1 ≤ i ≤ m. Formally defined, the intensity Iv for the

entire system of n lights is:

25

3.1. Problem Definition

Iv =n∑j=1

m∑i=1

xijIv,ij . (3.1)

If this luminous intensity is measured by some receiver at a distance d from the light cone,the receiver is said to measure to incident light on a plane perpendicular to the light cone.This intensity is formally known as illuminance (Ev) and is measured in lux (lm ·m−2) (see[79], pp. 265,266). Assuming this light is a point source, we can take advantage of the factthat illuminance decreases at a rate inversely proportional to the square of distance, suchthat:

Ev ∝Ivd2 , (3.2)

where d is the distance between the point source and apex of the solid angle. Using Eq. 3.2,we can calculate the distance between the light source and the receiver. In theory, we arealways assured to have sufficient knowledge of the light sources such that Iv for any lightsource is known and that Iv remains unchanged throughout the operational lifetime of thesource. We can then estimate the total Ev at the receiver from all the light sources:

Ev =n∑j=1

kj Iv,j , (3.3)

where k is a scaling factor found in a calibration step such that k = Ev/Iv. A complete wayof representing this relationship is:

Ev =n∑j=1

m∑i=1

kjxij Iv,ij . (3.4)

In this case, kj is measured for every individual light source in the room. Additionally,for a receiver rotated at some angle θ (the oblique angle of incidence) to the plane, theilluminance at the receiver is given as:

Ev,θ = Ivd2 cos(θ). (3.5)

It should be pointed out that prior knowledge of the intensity of the point source is notdirectly needed to measure the illuminance. The equations above describe the physical lawsof how a generic receiver measures illuminance. Such equations are useful when modelingthe linear inequality constraints. In the actual operation of the system, the illuminanceis measured directly using the sensor node. In turn, these values are used to solve theinequality Ax ≤ b. Therefore, only a calibration step (in which the percentage of lux fromeach light source and ambient light is measured) is required satisfy the inequality constraint.

In this linear system, for any given setpoint xij with maximum intensity Iv,ij , there isa corresponding maximum current, Ip,ij (C · s−1) calculated by multiplying the maximummeasured current by the setpoint xi for all 1 ≤ i ≤ m, where m is the total number of

26


wavelengths in the system. The subscript p is used to differentiate the equations from theluminous intensity and current. Using superposition, the total power consumed by a singlelight fixture is given by:

P = Vin

m∑i=1

xiIp,i. (3.6)

The development of Eqs. 3.1, 3.3 and 3.6 made use of several important assumptions.First, the system was presumed to be linear. This implies that the relationship betweensetpoint x and measured lux current draw is linear. Empirical results are used to establishthis criterion can be found in Chapter 5. At this point we have established the physicalrelationship between the operational setpoint of a light fixture, the corresponding powerconsumption, and the perceived illuminance of an optical sensor placed in the room. For theremainder of Chapter 3 we refine Eqs. 3.1 and 3.3, introduce additional lighting parametersTc and Ra (i.e., Correlated Color Temperature and the Color Rendering Index), and imposeconstraints on the values of x.

Room Calibration

In Eq. 3.5, it was assumed that Ev was measured without additional light. In order to mea-sure Ev (the lux measured by the optical sensor) reliably, several calibration measurementsare required for the n fixtures present in the room. First, a dark measurement must betaken. This measurement, denoted Edark, captures the present lighting conditions in theroom (e.g., sunlight, incandescent, or fluorescent sources). Second, the scaling factor kj for1 ≤ j ≤ n must be measured for all the n light sources in the room. In order to derive the Evcontributed by the n fixtures, we simply choose a set point x for each of the n fixtures andsubtract Edark for each of the measurements. Once the dark measurements are obtained,the true contribution of each of the n sources can be estimated as long as the optical sensordoes not move or the Edark does not change.

Sensing Changes to Operating Environment

In reality, the dark measurements performed by node are not static, but change over timein response to daylight, other light sources being turned on and off, shadows, and occlusion.Additionally, the distance between the receiver and the source may change as well. Thesensor node can be moved, lifted, or carried as it measures the lux in the room. Theonly requirement is that the sensor node remain stationary during calibration. Given thisdynamic behavior, it is necessary that the sensor board employ some type of mechanism orintelligence to sense changes to the operating environment.

27


Color Temperature and Color Rendering Index

If the lighting network in Eq. 3.1 is comprised of LEDs of different wavelengths, then thecolor rendering index (Ra) and correlated color temperature (Tc) of the system can becontrolled as well.

The color rendering index (CRI) is a unitless measurement of how an artificial lightsource represents that of a perfect illuminant (i.e., daylight) where the reference CRI istypically 100. Any incandescent system, or more generally, any black body radiator, emitstemperature-dependent visible energy and has spectral density (ueλ) given by Planck’s for-mula:

ueλ = 8πhcλ−5(ehc/kTλ − 1)−1 (J · m−3) (3.7)

where c is the velocity of light, h is the Planck constant, k is the Boltzmann constant, andT is the temperature in kelvin. Additionally, we can reconstruct the relative spectral energyof daylight using a set of equations derived from the first two eigenvectors of empiricallymeasured daylight (achieved using Principal Component Analysis, see Wyszecki and Stiles[79] pp. 143-147). These models represent ideal sources of radiant energy and serve as thereference illuminant in the calculation of Ra (we quantitatively compare an ideal source andthe source we wish to characterize).

The temperature, measured in kelvin is used to classify the color of an artificial lightsource. Warmer (red) colors have a lower color temperature and cooler (green and blue)colors have higher color temperature. By manipulating the intensity of the system’s indi-vidual wavelengths Ra and Tc can be tuned perfectly for the given application. The directcalculation of the quantities is a complicated procedure and is not treated formally here.However the formulas, uniform color spaces, and other colorimetric calculations are foundin [79]. For this discussion, we treat the calculation of these parameters as a black box,whose result depends on the intensity of the wavelengths.

Revisiting the problem detailed in Sec. 3.1, we introduce two new parameters to control,Ra and Tc. A single fixture consisting of m wavelengths, whose individual wavelengths aredenoted λi where i ∈ Z : 1 ≤ i ≤ m also has a CRI that is function of the m wavelengthsweighted by setpoint, xi. That is,

Ra = f(x1Iv,1, x2Iv,2 . . . , xmIv,m) , (3.8)

and has a corresponding color temperature denoted as

Tc = f(x1Iv,1, x2Iv,2 . . . , xmIv,m). (3.9)

It should also be noted that white light (of a fixed Tc) can be created using just twowavelengths and that the greater number of wavelengths in a system, the higher the CRI.This is intuitive; as more energy is spread throughout a larger visible range, a closer ap-proximation of sunlight can be achieved than concentrating a larger amount of energy in

28


a smaller visible range. Consequently, systems creating white light using only two wave-lengths have a lower CRI than systems employing many wavelengths. For an arbitrary colortemperature Tc, increasing the CRI requires that we spread the radiant energy out over alarger visible range.

In a system comprised of more than three wavelengths, arriving at the optimal setpointfor a given Tc is a difficult problem and has no unique solution. This limitation arises becauseof how we mathematically model color vision. The basis for all colorimetric computationstarts with the assumption that most humans are trichromats (unfortunately some aredichromats, and some even monochromats) and that we perceive color as a weighted sum ofthree variables over a visible range of 380 nm− 780nm. As an analogy, this can be thoughtof as an additive red, green, and blue system.

The problem arises when we try to decipher the color of a light source that is no longera mixture of three wavelengths but of four, five, six, etc. Biologically, this presents noproblems to us, yet mathematically, there is no exact solution. How is this dealt with?In the past, researchers fixed the setpoint of the fourth wavelength (typically an ambercolor) to ratio of another wavelength (typically red) [3, 42, 41]. In this case, a set of linearequations are solved for and the set point of the red, green, blue, and amber is determined(see Sec. 3.2). However, as will be shown in Chapter 5 this method actually leads to loweroverall CRI values. The situation arises because the color temperature of a system can beexpressed linearly, yet the optimal CRI for that particular color temperature is nonlinear.

Dynamic Efficacy

The optimization of the number of wavelengths used in the system is referred to in thiswork as dynamic efficacy. It is the method of selecting the optimal number of wavelengthsrequired to satisfy the lighting condition. For a light fixture n, given Tc and Ev, there existsa group of setpoints x that minimizes the total radiant energy, hence, the electrical energyof the lighting system. Conversely, for the same Tc and Ev, there exists another group ofsetpoints x′ that maximizes the color rendering index, requiring more radiant energy, hencemore electrical energy is needed.

Let us consider a simple example for two light sources A and B to illustrate thesepoints. Light source A is dichromatic and light source B is pentachromatic. This meansA is comprised of two dominant wavelengths and B is also comprised of these same twodominant wavelengths and three other wavelengths. Both light sources are set such theyhave the same color temperature, Tc(A) = Tc(B), and at some distance d have the sameilluminance Ev(A) = Ev(B). Which light source has the greater CRI? According to thelogic above, light source B will have a greater CRI. But which light source is more efficient?Or to rephrase the question: which light requires the least amount of radiant energy to havethe equivalent illuminance?

First, radiant energy measured on a plane some distance d from a point source is known

29


as irradiance. Irradiance has the symbol Ee and the units (W ·m−2) and it too, for a pointsource, obeys the inverse square law. It is analogous to the illuminance calculation in Eq. 3.2,except it describes the electromagnetic power incident on a surface.

Illuminance is calculated using irradiance by integrating Ee and V (λ), the photopicefficiency function. This efficiency function is a relative measurement and is dimensionless.The results of integrating Ee and V (λ) are scaled by Km, an empirically derived luminousefficacy function (lm ·m−1), which has a peak sensitivity at 555nm (Wyszecki and Stiles).Formally, this is written as:

Ev = Km

ˆλ

EeV (λ)dλ. (3.10)

The important conclusion here is that different amounts of radiant energy at different wave-lengths result in different amounts of visible energy. We can now answer the question posedat the beginning of this section. If the two dominant wavelengths in light source A requireless radiant energy to produce the same illuminance as light source B for a given Tc, thenlight source A is more efficient, that is, it has a higher efficacy.

In the previous example, we considered a system of fixed illuminance and fixed colortemperature. Using LEDs as a light source not only enables the illuminance at a pointof interest to be adjusted, but also the ability to dynamically optimize the number ofwavelengths required to satisfy the lighting goals. These lighting goals might be driven by auser’s preferences of illuminance, color, energy efficiency, or color rendering index. We nowproceed to formally describe these goals.

Formal Problem Definition

Consider n artificial light sources, where each light source consists of m wavelengths, wherethe ith wavelength is designated λi, the jth light source designated Lj and i and j areindexing variables of the set i ∈ Z : 1 ≤ i ≤ m and j ∈ Z : 1 ≤ j ≤ n. If we considerthe light source as a spherical point source, then for every λi in Lj there is a maximumilluminance designated Eij incident on a plane at some distance d from the point source.In the presumed linear system, the intensity of any single wavelength λi in light source Ljwith corresponding maximum intensity Iv,ij , a corresponding Tc, and a corresponding Ra,is set by x where x ∈ Rmn : 0 ≤ x ≤ 1.

The optimal setpoint x has the form:

minx∈Ω

f(x)

where f : Ω ⊂ Rmn → ∪∞ and Ω is set of feasible points subject to linear or nonlinearconstraints. The function f does not have a closed form and is non-differentiable. Becausethere is no direct form of f , an optimization algorithm is required that is driven accordingto the results of f (see Sec. 3.3).

30


We define f to be an scaled objective function to minimize square error with z repre-senting our minimization goal:

minx∈Ω

f(x) =

(Tc−Tc(x)

Tc

)2+ ∆uv(Tc)2 +

(∑n

j=1Vin,j

∑m

i=1xijIp,ij∑n

j=1Vin,j

∑m

i=1Ip,ij

)2

for z = 0(Tc−Tc(x)

Tc

)2+ ∆uv(Tc)2 −

(Ra(x)100

)2for z 6= 0

(3.11)

subject to linear inequality constraints

n∑j=1

m∑i=1

kjxij Iv,ij ≥ Edesired − Edark, (3.12)

n∑j=1

m∑i=1

kjxij Iv,ij ≤ Edesired − Edark + β(Edesired − Edark), (3.13)

where

kj = Ev,j − EdarkIv,j

. (3.14)

As stated earlier, the left hand side of Eq. 3.12 and Eq. 3.13 can be replaced with theilluminance data directly measured by the sensor node and appropriately weighted by itscontribution relative to the total measured lux. In completeness, the full formulation, con-sistent with Sec. 3.1 is given. The term Edesired describes the total lux required at the pointof measurement. The dual inequality constraints (rather than a single linear equality con-straint) are required to reduce the sensitivity to noise, and to allow the objection functionf some flexibility as Tc, Ra, and Edesired can not be completely decoupled from each other(Principal Component Analysis could be used to show the extent of which these terms arelinearly separable).

The optimization problem can be thought of as a piecewise objective function whose goalis either minimizing energy (i.e., maximize efficacy) or maximizing the color rendering indexfor a specified color temperature, Tc. The parameter ∆uv(Tc) is an error measurement ofTc designed to ensure that the white point defined by the feasible points is a high qualityone (it is possible to create a set of Tc, as the color temperature can be described in termsof distance to the black body curve in chromaticity space). This piecewise function satisfiesthe goals of dynamic efficacy. The desired illuminance Edesired is controlled by the user(e.g., a slider).

To conclude, the feasible points found using direct search coincide with our assumptionsabout the spectral distribution of the system; the efficacy of the system is dependent on theintensity and number of wavelengths used to satisfy the lighting conditions.

31

3.2. Linear Methods of Controlling the Spectral Power Distribution

Size and Scaling of the Search Space

The feasible solution, Ω contains setpoints x where x ∈ Rmn : 0 ≤ x ≤ 1, where m isthe number of wavelengths and n is number of light fixtures in the system. For example,consider a finite system with six bits of precision, then a single setpoint x could take on over10, 000, 000 different values. Hence, for one light comprised of five wavelengths, the searchspace becomes 1035. And this is only for a single fixture.

If a typical office has four light fixtures installed overhead, where each lamp containfive wavelengths, then the search grows to 10140. With each additional light fixture toconsider, the search space grows exponentially. Controlling a large workspace consistingof n = 25 fixtures is practically out of the question, unless we employ an algorithm thatefficiently narrows the search in a short amount of time. Still, future systems will benefitfrom some sort of localization to infer which light fixtures are closest to the optical sensornode measuring the illuminance. To keep the dimensionality of the search space low, multiplenodes must be installed in large deployments.

Yet there are simpler solutions that may also yield reasonably accurate results. Themicrocontroller (8, 16, or 32 bit) controlling the intensity of each wavelength in a lightsource has a finite resolution many times smaller than that of the machines performing theglobal search. Consider a system in which the setpoint x, where x ∈ Zmn : 0 ≤ x ≤ 2p−1,and p = 16. Again with m = 5 and n = 1, then there are (216 − 1)5 ≈ 1024 solutions. Byconsidering just unsigned 16-bit integers, our search space shrinks by 11 orders of magnitude.Similarly, if p = 8, m = 5, and n = 1, there are now just 1012 solutions. This approximationof the search space makes run-time substrates possible at the cost of system resolution.

3.2 Linear Methods of Controlling the Spectral PowerDistribution

Traditional methods of color control assume a linear relationship between the setpoint andthe output. These techniques often formulate the problem as a linear mixture of red, green,and blue components, whose tristimulus values were measured in a previous calibration step.

In these problems, the exact solution is found by solving the linear equality Ax = bfor x. However, the system could also be formulated with more equations than unknowns,hence x is overdetermined and no exact solution exists. In this case, a common techniqueuses the pseudoinverse to minimize the sum-squared error between Ax and b, resulting inan exact minimum square error solution if A is non-singular. As will be shown in Chapter 5,these methods allow the correct color temperature to be found, but often have a low colorrendering index.

While these approaches work well to calibrate most optical sensors (e.g., those in scan-ners, cameras, etc.), they are not necessarily optimal for lighting. They fail because thislinear approach does not take into account the nonlinearity of calculating the color rendering

32


index. In this case, the only real benefit of these methods is the relatively low computationalcomplexity to obtain an answer.

Exact Solutions

Most often, we are interested in the relationship between a setpoint and the resulting color.If the lighting system is comprised of three wavelengths, say red, green, and blue, thenthe relationship is of the form Ax = b. In this case A is a 3-by-3 matrix comprised oftristimulus values from the measured system

Xr Xg Xb

Yr Yg Yb

Zr Zg Zb

Rset 0 0

0 Gset 00 0 Bset

−1

R

G

B

=

X

Y

Z

, (3.15)

where x is the setpoint for controlling the intensity of the RGB system. Intensity valuesRset, Gset, and Bset represent the intensity setting used at the time when the tristimulusvalues were measured, most often scaled to one. Solving Eq. 3.15 yields XrRset XgGset XbBset

YrRset YgGset YbBset

ZrRset ZgGset ZbBset

−1 X

Y

Z

=

R

G

B

. (3.16)

Using Eq. 3.16, we can specify the desired tristimulus value. Quite often, this tristimulusvalue is derived from the two-dimensional chromaticity coordinate of the desired correlatedcolor temperature.

It is common to normalize the white point of matrix A such that the sum of the Ytristimulus sum to 100. In this case, every element in matrix A is multiplied by a scalingfactor c, where

c = 100Yr + Yg + Yb

. (3.17)

This centers the fixture’s whitepoint such that Y = 100. The solution x is of the setx ∈ Rn : 0 ≤ x ≤ 1. If x is outside of this set, then it represents a point in spaceunachievable with the current configuration.

If we add a fourth wavelength to the system, a simple approach is fix the ratio of thisfourth wavelength to its closest color. For example, if we add an amber color, we makeamber a fixed ratio, α, of the red setpoint. In this example, the solution is Xr(λ1) +Xr(λ2) Xg Xb

Yr(λ1) + Yr(λ2) Yg Yb

Zr(λ1) + Zr(λ2) Zg Zb

−1 X

Y

Z

=

R

G

B

, (3.18)

33


where R

A

G

B

=

R

αR

G

B

. (3.19)

Similarly, we can do the same for five wavelength system. However, these are naivetechniques, which constrain the radiant energy to a linear system in which the output coloris the only concern. This problematic approach is only slightly mitigated by removing thisfixed ratio, and minimizing the sum-squared error.

Overdetermined Systems

If we consider a linear system of the form Ax = b that contains four wavelengths than wecan rewrite matrix A as a 3-by-4 system. However, A is now rectangular and has moreequations than unknowns, x is overdetermined, and, no exact solution exists. But, we couldseek a solution of x that minimizes the sum-squared error between Ax and b , such asJs(x) = ‖Ax− b‖2. In this classical problem, the norm has a closed form solution, and wesolve for the Jacobian by taking the derivative of Js(x) and setting the solution to equal tozero. Solving this leads to the pseudoinverse:

x = (AA)−1ATb

= A†b. (3.20)

Using the four wavelength example to illustrate a red, amber, green, and blue system:

R

A

G

B

=

Xr Xa Xg Xb

Yr Ya Yg Yb

Zr Za Zg Zb

S1 0 . . . 00 S2 . . . 0...

.... . .

...0 0 . . . S4

−1

† X

Y

Z

, (3.21)

where the 3-by-4 tristimulus matrix is measured with some intensity setting denoted hereas Si, where i corresponds to the specific wavelength, with a typical intensity such thatsolution x is of the set x ∈ Rn : 0 ≤ x ≤ 1 where the dimension of the space is n, thenumber of wavelengths.

Using this method, we are guaranteed that for any Tc, a minimum square error solutionexists. Therefore, in contrast to the method in Eq. 3.18, we are no longer tasked with findingthe ratio α such that we minimize error for any arbitrary Tc.

The use of this method, which is a improvement over the ratio-based technique is nearlynon-existent in literature regarding spectral control of LEDs.

34

3.3. Proposed Method of Control: Direct Search

3.3 Proposed Method of Control: Direct Search

Here we consider an entirely new approach for solving the optimal setpoint in a system com-posed of an arbitrary number of wavelengths. The key insight here is that linear models areefficient (for the most part) at accurately controlling the intensity for a specified correlatedtemperature, yet this solution, due to the nonlinear nature of calculating the color renderingindex, leads to a chromatically accurate white light with a low color rendering index. Thisis not desireable.

The reason is that parameter Ra is the average color difference of eight different colorsamples illuminated by both the test light source and a perfect illuminant (perfect meaninga CRI of 100). Often times, a setting obtained using a linear method may have a goodscore (Ri > 80) for less than quarter of the color samples. Finding the optimal Ra meansfinding a feasible set of solutions in the search space, while at the same time meeting theilluminance constraints and desired color temperature.

In this section, we briefly introduce direct search algorithms, describe a subset of tworelatively new methods, pattern search and mesh adaptive search, and their application tosolve the problem detailed in Sec. 3.1.

Background

Direct search methods are a class of optimization algorithms that require only a numericalanswer, supplied by the iterative evaluation of the objective, that guides the search. Ingeneral, these methods are applicable to problems that lack a closed form solution, or non-convex, or discontinuous. In the words of Audet et al.,

“There is a wide spectrum of optimization methods. At one of the spectrum,there are those designed to exploit a specific known structure of the optimiza-tion problem. For example, the simplex method [19] is designed to exploit thelinearity of both the objective and constraint functions to solve large problemsefficiently. At the other end of the spectrum are methods that do not or cannotexploit any structure. This is often the case when the objective and/or the con-straints are evaluated through computer code (such as process simulator) or byan experiment...Direct search methods belong to the end of the spectrum [7].”

The first published papers on direct search appeared in the early 1960s, yet they lackedrigorous convergence analysis and were dismissed by the community for almost 30 years untilthe award winning work of Torczon [74] on the convergence of pattern search algorithms,which are a generalization of algorithms used by Hooke and Jeeves [27]. Powell [53] describesthe many forms of direct search: line search, discrete grid, simplex methods, conjugatedirection, linear approximation, quadratic approximation, and simulated annealing. Somewell known direct search methods documented by Audet et al. are: Jones et al. [31], Hooke

35

3.4. Lighting Network Control

and Jeeves [27], Rosenbrock [56], and Nelder and Mead [46]. Kolda, Lewis, and Torczon [32]also provide a historical and technical review of these methods.

Pattern Based Algorithms

The generalized pattern search algorithm described in [74] proves the convergence of aniterative algorithm most analogous to the classic “compass search.” In this work, Torczonestablished global convergence, that is, subsequent iterates (i.e., progressively decreasingthe objective function) that guaranteed first order convergence and hence proved the al-gorithm’s optimality. Recently (2006), the mesh adaptive direct search algorithm (MADS)published in [10, 9] extends the work of Torczon, but calculates the search points at random.Both methods allow for both linear and nonlinear constraints and bounds, however the al-gorithms perform exceedingly well in unconstrained cases. The algorithms are implementedin MATLAB as part of the Global Optimization Toolbox [71].

Algorithm Description

Both pattern search and the mesh adaptive algorithm are iterative algorithms that evaluatea mesh centered around the current point. A mesh is formed by the multiplication ofthe current point with a set of vectors called a pattern. Typically these patterns are of themaximum basis 2n or minimum basis n+1, where n is the total number of dimensions in thesearch space. The size of mesh, mesh size, is used along with the pattern and incumbentsolution in a poll step, where the new points in the mesh are tested to see if they aresmaller than the present solution. Both algorithms have the ability to grow or shrink thesize of mesh such that the search can expand and contract based upon the results of theuser-supplied evaluation function. With each incumbent solution, the algorithm shrinks themesh, and when the mesh size decreases beyond a threshold, the algorithm terminates. Ofcourse, there are many more details that are omitted in this discussion. These are found in[74, 37, 36, 1, 10, 8, 2].

3.4 Lighting Network Control

In order to evaluate the methods of spectral control and test the closed loop control of thelighting network (detailed in Secs. 3.1 and 3.2) a custom toolbox was written in MATLAB.The spectrometer processing and optimization toolbox (SPOT) is a collection of functionsthat calculate and evaluate illuminance, color temperature, color rendering, and the newcolor quality scale (CQS) [20]. These functions were used in the evaluation and control oflighting parameters in this work. It is planned to release this package to the communityunder open source license.

36

3.4. Lighting Network Control

1. Measurement (once per lamp)

a) Perform a calibrated radiometric measurement for each of the wavelengths in thesystem to obtain the spectral power distribution, Ee or Pe. Measure system qui-escent current, (Iq) and the maximum forward current (Iλ) for each wavelength.

2. Calibration

a) Perform room calibration. Measure Edark and Ev for an arbitrary set point forall the controllable fixtures in the area and calculate the contribution of eachlight source at the sensor node. (Sec. 3.1).

3. Calculate Optimal Setpoint

a) Specify an objective function (Eq. 3.11) and select what parameter will be max-imized (CRI or Efficacy), then specify a target color temperature Tc and desiredilluminance Edesired. The objective function operates using a model of the systemusing the spectral data acquired in step 1.

b) Obtain results and set light fixtures.

4. Update

a) Continuously monitor for changes in intensity or color temperature at the sensornode or requests for recalibration.i. If Edesired or Tc change but the sensor node is not displaced go back to

step ??.ii. If sensor node is displaced or recalibration is requested, go back to step 2.

Figure 3.1: The lighting control algorithm.

Control by Optimization Workflow

The control problem is summarized in Fig. 3.1. The process begins with step 1. Controllingthe spectral output of a single or group of light fixtures requires a one time step of measuringeither the radiant flux with symbol Pe (W), which requires the use of an integrating sphere,or the irradiance Ee (W ·m−2) and distance of the measurement, the latter of which is usedin this work. Additionally, in this step, the voltage, quiescent current Ip,q, and maximumcurrent Ip, for each wavelength are also recorded. Step 2 is the calibration step. Here,the Edark of the room is measured, as well as the individual illuminance of each of thefixtures measured by the sensor node. In step 3, the direct search algorithm is called usingthe information in collected in steps 1 and 2 with additional information about what tomaximize. Finally, in step 4, the system continuously monitors for changes and returns toeither step 2 or step 3, depending on what happened.

37

3.5. Sensor Feedback in a Lighting Network

3.5 Sensor Feedback in a Lighting Network

If a sensor was introduced that measures color, for example, red, green, and blue, thenEq. 3.22 can be written to include the color sensor information. Assuming a calibrationstep is taken where the intensity of the system is set to maximum, we omit writing thesetpoint matrix and define the problem as:

Xr Xg Xb

Yr Yg Yb

Zr Zg Zb

Rr Rg Rb

Gr Gg Gb

Br Bg Bb

−1

Rs

Gs

Bs

=

X

Y

Z

, (3.22)

where the vector of interest is found by the red, green, and blue measurements from thecolor sensor. In this case, the solution is of the form A−1b = x, where b represents thetristimulus values of the desired color and x the corresponding sensor values. Thus, theexpected sensor values can be obtained by first performing a calibration step where thefixture is set to a maximum intensity and the corresponding trimstimulus and sensor valuesare recorded for later computation. Once these values are obtained, closed loop operation ismaintained by comparing the expected color sensors values to the measured sensor values.Typically a proportional, integral and derivative (PID) type controller is used to minimizeerror between the setpoint and measured values [3, 5].

If the system is comprised of more than three wavelengths, then the color levels can beapproximated using a similar channel such as red for amber, or blue for cyan, at the costof decreased responsivity (this sacrifice in dynamic range reduces the ability to converge onparticular solutions). This could be augmented by using additional sensors and placing atinted filter over the array to reject wavelengths not in the desired bandpass. For example,using a five wavelength system, we pose to color-feedback and control as:

Rs

As

Gs

Cs

Bs

=

Xr Xa Xg Xc Xb

Yr Ya Yg Yc Yb

Zr Za Zg Zc Zb

S−1

† X

Y

Z

, (3.23)

where S is a square matrix containing the color sensor readings for the five channels eachmeasured for a particular wavelength,

S =

Rr Ra Rg Rc Rb

Ar Aa Ag Ac Ab

Gr Ga Gg Gc Gb

Cr Ca Cg Cc Cb

Br Ba Bg Bc Bb

. (3.24)

38

3.5. Sensor Feedback in a Lighting Network

The results of Equations 3.22 and 3.23 produce a result that is the sum of the sensorreadings for every channel. This means that the solution cannot be used to explicitly setthe intensity of each channel. If we take into account the sensitivity of a single wavelengthto the other wavelengths, then assuming there is no wavelength shift, we can decouple thesystem. Using Equation 3.23 as an example, we extend the results of [5] to five wavelengthsand compute the sensitivity matrix as

Γ =

1 Ra/Aa Rg/Gg Rc/Cc Rb/Bb

Ra/Rr 1 Ag/Gg Ac/Cc Ab/Bb

Rg/Rr Ga/Aa 1 Gc/Cc Gb/Bb

Rc/Rr Ca/Aa Cg/Gg 1 Cb/Bb

Rb/Rr Ba/Aa Bg/Gg Bc/Cc 1

, (3.25)

and solve Ax = b for x, that is,

Γ−1

Rs

As

Gs

Cs

Bs

=

Rs

As

Gs

Cs

Bs

. (3.26)

The decoupled sensor readings can now be used to directly set the intensity of the individualchannels.

39

Chapter Four

Physical Implementation

The hardware, firmware, and software developed in this research phase aims to break newground in the control of energy efficient lighting networks by researching and implementinghighly configurable light sources and determining the proper feedback sent from the sensornetwork. In this chapter, we present a novel pentachromatic lighting and control system totest the theory defined in Chapter 3.

4.1 Requirements

The ideal system is comprised of a sensor node, daylight, incandescent, fluorescent, and LEDtechnologies. The sensor node measures the intensity and color temperature of all sourcesand sends the data to a controller (in this research, a computer). The controller processesthese data and, taking into account the user’s preferences, sets the lighting to the optimalconditions (Figure 4.1 on page 41). This method can be extended for multiple controlnodes. However, with some adaptation, the enhanced processors utilized in both the sensorand lighting controllers can perform most of the lookup routines and some floating pointcalculations. Additionally, an embedded processor (i.e., ARM 9, Atom, or OMAP) runningan operating system could perform the same optimization algorithms as the computer.

To solve the problem outlined in Section 3.1, we designed a prototype LED lightingsystem and sensor board. The intensity of the LED fixture is controlled using pulse widthmodulation (PWM). This requirement is crucial, since much of the theory requires that thepower and intensity relationship of the light fixture be linear. A digital signal processor(DSP) with microcontroller-type peripherals controls the intensity, monitors onboard LEDoperating temperature, intensity and color, and has a bidirectional data link to the computer.The sensor node measures intensity and color. These data allow the controller to estimatethe illuminance and correlated color temperature of each light source. The sensor node alsofeatures basic controls for the user (e.g., intensity and color control), and detects occupancy.The LED array is comprised of five dominant wavelengths; four of the LEDs are activeemitters and and one is a phosphor based device. This enables a wide range of possible

40

4.1. Requirements

Intensity Data Link

Optional Sources

LEDLighting

SensorNode Controller

Figure 4.1: System diagram. The lighting network consists of LED light sources, optionalincandescent and fluorescent sources, and ambient room conditions (daylight), that aremeasured by a single or group of sensor nodes. The sensor nodes return intensity and colorinformation back to the control node (e.g., a computer) for processing. Here, the intensityis controlled via a link to the artificial light sources. The data link is bidirectional fromeither the sensor node or the LEDs.

41

4.2. Sensor Node

color temperatures and serves a second purpose – the broad spectral range ensures anexceptionally high color rendering index.

4.2 Sensor Node

The sensor node is controlled by a TMS320F28027 controller from Texas Instruments andis comprised of multiple sensors. The device detects occupancy via a passive infrared sensor(AMN11111, Panasonic), as well as measures the illuminance and color of incident light(Table 4.1 on page 43). The sensor also features a basic user interface, realized using linearpotentiometers and buttons. These serve to specify the intensity and color temperature ofthe adjustable light sources.

Measuring the illuminance is achieved using three analog linear illuminance sensors(ISL29006, Intersil) with different sensitivities and one digital sensor (TCS3414CS, Taos).The analog sensors are sampled by the system at 30Khz using a 12 bit successive approx-imation converter. Oversampling of the sensors is user configurable (up to 256×). Two ofthe sensors are filtered using a second order unity gain low pass with a cut off frequency of50Hz which ensures illuminance data is accurately measured when fixture intensity is lessthan 100% (the circuit integrates the duty cycle of the LEDs). The integration time (e.g.,update rate) of the digital color sensor is a function of an internal timer which controls theacquisition time of the digital color sensor. This is nominally set to 120Hz. The digitalcolor sensor measures the irradiance of a photodiode array consisting of red, green, blue,and clear filters. Figure 4.2a on page 44 is an image of the prototype sensor node. Theprototype is on two layer 1.65mm (0.065 in) thick circuit board and measures 178×119mm(7 × 4.7 in). Near-future modifications include drastically reducing board size, adding anaccelerometer, and reducing power consumption.

4.3 LED Controller

A TMS320F28027 from Texas Instruments runs custom firmware to control the LEDs andread the sensors. The system requires the use of a 24V supply and is powered by a laptoppower supply rated at 24V, 2.2A. Five BuckPlus DC-DC buck converters from LEDdynam-ics drive the LEDs. The intensity is controlled by five independent 16-bit PWM signalswith a fundamental frequency nominally set at 120Hz. A single pixel digital color sensor(TCS3414CS, Taos) samples the LED array from within the fixture in certain cases, andupdates color data at 20Hz. This is useful for closed loop control of color at the fixture. Op-tionally, an internal linear analog lux sensor (ISL29006, Intersil) is also sampled at 30Khz.These sensors are used for closed loop control of the LEDs. Full duplex communicationis handled by the system’s asynchronous receiver and is interfaced using either a wired orwireless link. Measuring the on-board color sensor requires precise timing. The integrationof the color sensor is controlled via hardware, which is synchronized to the LED fundamental

42

4.3. LED Controller

Sensor

Typ

eOutpu

tBan

dwidth

Ran

geADC

Resolution

mlx

/mV

a

ISL2

9006

,Intersil

analog

,linear

50Hz(externa

lfilte

r)20

00lux

21260

6mlx/m

VISL2

9006

,Intersil

analog

,linear

600Hz(nofilter)

b20

00lux

21260

6mlx/m

VISL2

9006

,Intersil

analog

,linear

50Hz(externa

lfilte

r)10

000lux

21230

30mlx/m

VTCS3

414C

S,Ta

osdigital(I2C),lin

ear

50Hz(in

ternal

filter)

5000

lux

21615

15mlx/m

V

Table4.1:

Illum

inan

cesensor

tableforthesensor

node

inSe

ction4.2.

The

samplingfreque

ncyof

thean

alog

sensorsis

30kH

z.The

samplingfrequency(in

tegrationtim

e)of

thedigitalsensorisc

ontrolledby

thesensor

node

.It

isno

minally

sett

o12

0Hz,

with

varia

ble

oversamplingsetat

2×(upto

10×).

a (Vrefh−Vrefl

)=

3.3V

b Based

onman

ufacturerda

taat

1000

lux.

43

4.3. LED Controller

(a) Sensor Node.

(b) LED Control Board.

Figure 4.2: The prototype sensing and control hardware used in the experiments.

44

4.4. LED Array

frequency. This ensures color measurement precision. Figure 4.2b on page 44 is an imageof the prototype controller with its input protection circuitry removed. The prototype is ona two layer 1.65mm (0.065 in) thick circuit board and measures 178× 119mm (7× 4.7 in).This board can also be reduced in size.

4.4 LED Array

The primary LED array used in this research consists of five unique wavelengths, four ofwhich are active emitters and one that is phosphor-converted (Table 4.2 on page 46). The fivecolors employed are royal blue, cyan, green, phosphor-converted amber, and red. Selecting aphosphor-converted amber allows for a higher efficiency. The intensity of the red, orange, andamber wavelengths (AlGaInP compounds) are quite temperature sensitive. The efficiencyof the device (i.e., conversion of electrons to photons) decreases in these compounds withshorter wavelengths. This is an unfortunate phenomenon, as our eye’s sensitivity peaksat 555 nm. The blue, cyan, and green wavelengths (InGaN) compounds do not exhibitthe same temperature dependence as the AlGaInP devices, however they too suffer fromdecreased quantum efficiency near our eye’s peak sensitivity. Krames et al. [33] detail therecent breakthroughs in high efficiency optoelectronics and discuss the status and futureof these devices for lighting applications. Moreover, the phosphor-converted (PC) amberuses an ultra-violet pumped blue LED to excite the organic phosphor which, re-radiates atlonger wavelengths. This fixes the problem of systems with decreasing efficiency utilizingwavelengths in the neighborhood of 580nm. The LED array also has on board temperaturesensors (to estimate LED junction temperature) and can measure, using filters, the red,green, blue, and clear values of the irradiance using a TCS3414 color sensor (Table 4.1 onpage 43). Figure 4.3 shows the actual array used in the research.

Informally, using discrete emitters creates a light source in which there is no scatteringof the discrete wavelengths. This leads to “fringing” on the outer edge of light incident toa surface. This effect is not desireable, as it produces individual shadows cast by the LEDsnear the edges of the light field. Ramer et al. [55] and Rains et al. [54] utilize constructiveinterference to reduce these effects. This technique requires the use of a reflective anddiffuse hemisphere to scatter the photons before they emerge from the LED array aperture.However, this improvement comes at significant cost, as scattering loss lowers the totalcandelas of the lamp. In this research, this technique is tested and presented in Chapter 5but ultimately, a direct view method (i.e., the LED array is facing downwards) is selectedfor efficiency. Figure 4.4 on page 48 shows the dome mounted on the LED array, which inturn is mounted to aluminum heatsink.

45

4.4. LED Array

Color

Dom

inan

tWavelen

gth

FW

HM

alm

/W

bLum

iledPart

Royal

Blue

440n

m-4

60nm

30nm

350m

W/W

cLX

ML-PR

01-035

0Cyan

490n

m-5

20nm

30nm

70lm

/WLX

ML-PE

01-007

0Green

520n

m-5

50nm

24nm

70lm

/WLX

ML-PM

01-007

0Ambe

r(P

hospho

r)58

7.8n

m-5

95.4nm

80nm

70lm

/WLX

M2-PL

01-000

0Red

620.5n

m-6

45nm

20nm

40lm

/WLX

ML-PD

01-004

0

Table4.2:

Descriptio

nof

thecolor,

dominan

twavelen

gth,

fullwidth

halfmax

imum

,efficacy

andfullpa

rtnu

mbe

rof

theLE

Dsused

totest

thesystem

.a F

ullW

idth

HalfM

axim

um(F

WHM)d

escribes

theba

ndwidth

ofthespectral

output.Itisrelatedto

thestan

dard

deviation(σ)o

faGau

ssiandistribu

tion

such

that:FW

HM

=2√

2ln

2σ.

b forwardcurrentis

350mA

c pub

lishedas

radiom

etricpo

wer

(mW

)

46

4.4. LED Array

(a) LED Ring comprised of Luxeon Rebel emitters

(b) Close up of LED array

Figure 4.3: The LED ring comprised of five wavelengths with on-board color and tempera-ture sensors (mounted on a concentric heat sink).

47

4.4. LED Array

(a) Rendering of dome and heatsink [62, 73].

(b) LED array with dome mounted and heatsink.

Figure 4.4: Interference dome mounted on the LED array with aluminum heatsink. Theinterference technique is adapted from [55, 54].

48

Chapter Five

Observations

The experimental results are presented in this section. Where needed, theoretical predictionsare provided for comparison.

5.1 Experiment Setting

For two weeks, a “dark room” conveniently located in building E14 served as the testing areato acquire the LED spectra. The generic setup utilized an adjustable light source, a fixedsensor node (with illuminance sensors), and a cosine-corrected spectrometer head. A laptopinterfaced the electronics and stored the data using MATLAB. The testing environment wasnot necessarily ideal due to reflected light (there were reflective aluminum objects in theroom) and lack of baffling, but sufficed to take the data. The wide field of view for both thesensor node and spectrometer are somewhat sensitive to reflected light from the the source.These data in the experiments were downsampled to 1 nm resolution for processing.

5.2 LED Measurement

A series of experiments to measure the linearity of the system were conducted. Theseexperiments measured absolute irradiance and the power of each channel of the system. Acosine-corrected USB4000 spectrometer radiometrically calibrated against a NIST traceableradiometric standard measured system irradiance (see Appendix D). The spectrometer dataare sampled at 0.1nm resolution and downsampled to 1.0nm for all subsequent processing.The spectrometer data were used to create a model of the light source for the computer-optimized control of the system.

Observing the Inverse Square Law

The light source (light source A) was set to maximum intensity and the distance of thelight from the from the spectrometer and sensor node was varied. The objective of this

49

5.2. LED Measurement

(a) Testing with the dome.

(b) Testing without the dome (notice fringing effects).

Figure 5.1: The makeshift “dark room” for acquiring LED spectra. In most experiments,the distance was fixed at 30 cm and the illuminance and power of the setting is measured.

50

5.3. Theoretical Limits of Operation

experiment is to confirm the assumption that the light source acts like a point source (seeFig. 5.2).

Linearity and Superposition

Experiment #1

In this experiment, the distance was fixed at 30 cm and the intensity of the individualLEDs, as well as, the “all on point” of light source A was varied. The illuminance wasmeasured using the TCS3414 sensor (clear channel) and a HP 34401A multimeter measuredthe current. The supply voltage was 24V. The spectrometer data from this experimentwere obtained incorrectly, so a second set of experiments were conducted (Fig 5.5 is agraph of this error). The problem was determined to be an electronic dark correctionsetting on the spectrometer. In effect, this setting subtracted too much irradiance fromthe individual LED measurements such that there was a large error between the whitepointand the superposition of the individual spectra. This problem was fixed in experiment2. However, data acquired using the sensor node’s digital color sensor are still valid (seeFig 5.3 and Fig 5.4). Although the sensor does not report an absolute measurement, it stillcaptures the effects of temperature (measured in sensor counts). The measured quiescentpower consumption, Pq = 2W, is subtracted from the measurements in Fig 5.3.

Experiment #2

To revise the errors in the spectrometer data from the previous experiment, the maximumintensity for each wavelength was acquired, this time disabling the electric dark correctionand calibrating the spectrometer integration time for each wavelength. Measurements weretaken at 30 cm from the source. The experiment was conducted for both the dome and nodome versions of light A (see Fig .5.6). Additionally, the optical efficiency of the dome isgiven in Table 5.1.

5.3 Theoretical Limits of Operation

To prepare a set of test points to prove the effectiveness of the proposed direct search algo-rithm, a series of experiments were devised. First, a Monte Carlo simulation was performedusing the data taken in Sec. 5.2 to understand the relationship between these variables be-fore applying the direct search algorithm. The resulting contours highlight the difficultyencountered when randomly searching in this space, namely getting stuck in local minima.Next, the individual LED spectra acquired in Experiment #2 were used to create a modelsimilar to that in the Monte Carlo experiments. Using this model, a standard set of testpoints for comparing and contrasting the linear equality, pseudoinverse, and direct searchwith linear inequality constraints methods was created. In short, the goal was to determineif direct search was a feasible nonlinear method for controlling a single fixture.

51


0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

10

100

200

300

400

500

600

700

800

900

1000

Dis

tanc

e (m

)

Lux (lm ⋅ m−2

)

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

101000

2000

3000

4000

5000

6000

7000

8000

9000

1000

0

Dis

tanc

e (m

)Clear Channel Sensor Counts

Illum

inan

ce v

s. D

ista

nce

Mea

sure

d Lu

xE

xpec

ted

Lux

Mea

sure

d S

enso

r C

ount

sE

xpec

ted

Sen

sor

Cou

nts

(a)A

linearplot

oftheexpe

riment.

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

100

200

300

400

500

600

700

800

900

1000

Dis

tanc

e (m

)

Lux (lm ⋅ m−2

)

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

11000

2000

3000

4000

5000

6000

7000

8000

9000

1000

0

Dis

tanc

e (m

)

Clear Channel Sensor Counts

Illum

inan

ce v

s. D

ista

nce

(log−

log

plot

)

Mea

sure

d Lu

xE

xpec

ted

Lux

Mea

sure

d S

enso

r C

ount

sE

xpec

ted

Sen

sor

Cou

nts

(b)A

log-logplot

oftheexpe

riment.

Figu

re5.2:

Luxan

dTCS3

414clearchan

nelm

easurements.The

distan

cebe

tweenthesensorsan

dlig

htsource

was

increm

entedat

10cm

stepsforatotalo

fseven

measurements.The

expe

cted

measurements,o

fthe

form

E v=

Iv d2,a

rethederiv

edfrom

thelast

data

point(the

luxan

dsensor

coun

tsat

90cm

).

52


0 1 2 3 4 5 6 7

x 104

0

400

800

1200

1600

2000

Pulse Width Modulation Setting (0−64000)

Cle

ar C

hann

el S

enso

r C

ount

s

0 1 2 3 4 5 6 7

x 104

0

0.8

1.6

2.4

3.2

4Sensor Counts and Power vs. Intensity (Blue Channel)

Pow

er (

W)

Sensor CountsPower (W)

(a) Results for royal blue LEDs.

0 1 2 3 4 5 6 7

x 104

0

400

800

1200

1600

2000


Cle

ar C

hann

el S

enso

r C

ount

s

Sensor Counts and Power vs. Intensity (Cyan Channel)

0 1 2 3 4 5 6 7

x 104

0

0.8

1.6

2.4

3.2

4

Pow

er (

W)


(b) Results for cyan LEDs.

0 1 2 3 4 5 6 7

x 104

0

400

800

1200

1600

2000


Cle

ar C

hann

el S

enso

r C

ount

s

0 1 2 3 4 5 6 7

x 104

0

0.8

1.6

2.4

3.2

4Sensor Counts and Power vs. Intensity (Green Channel)

Pow

er (

W)


(c) Results for green LEDs.

0 1 2 3 4 5 6 7

x 104

0

400

800

1200

1600

2000


Cle

ar C

hann

el S

enso

r C

ount

sSensor Counts and Power vs. Intensity (Amber Channel)

0 1 2 3 4 5 6 7

x 104

0

0.8

1.6

2.4

3.2

4

Pow

er (

W)


(d) Results for amber LEDs.

0 1 2 3 4 5 6 7

x 104

0

400

800

1200

1600

2000


Cle

ar C

hann

el S

enso

r C

ount

s

0 1 2 3 4 5 6 7

x 104

0

0.8

1.6

2.4

3.2

4

Pow

er (

W)

Sensor Counts and Power vs. Intensity (Red Channel)


(e) Results for red LEDs.

Figure 5.3: Results of measuring each of the five wavelengths for light source A as sensedby the digital color sensor and multimeter.

53


0 1 2 3 4 5 6 7

x 104

0

2000

4000

6000

8000

10000


Cle

ar C

hann

el S

enso

r C

ount

s

Sensor Counts and Power vs. Intensity (Equal Intensity)

0 1 2 3 4 5 6 7

x 104

0

4

8

12

16

20

Pow

er (

W)

Measured Equal IntensitySuperposition Equal IntensityMeasured Power (W)Superposition Power (W)

Figure 5.4: In an experiment designed to test the linearity and superposition of the individualwavelengths, all LEDs were set to the same intensity then progressively increased. Theobserved sensor counts and observed current were recorded. In this graph, the data acquiredpreviously (figure 5.3) are compared to the equal intensity test. These experiments are usefulto quantify the errors of the underlying linear assumption.

54


300 400 500 600 700 800−0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

Wavelength (nm)

Abs

olut

e Ir

radi

ance

(µW

⋅ cm

−2 )

Absolute Irradiance for Light A

Royal BlueCyanGreenAmberRed

Figure 5.5: This graph shows the individual acquired spectra from experiment #1. It waslater determined that an incorrect electronic dark setting in spectrometer led to incorrectreadings of the individual spectral components, thus these measurements were not used tocreate the model of the system. Yet, the spectra adequately depict the visible energy ofthe system. In subsequent experiments, the electronic dark setting was disabled and thespectrometer integration time was specified for each wavelength measurement.

55


300

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)

Absolute Irradiance (µW ⋅ cm−2

)

Abs

olut

e Ir

radi

ance

for

Max

imum

Set

ting

Ligh

t A (

Dom

e) (

dist

ance

= 3

0cm

)

M

easu

red

Irra

dian

ceS

uper

posi

tion

of Ir

rand

ianc

e

(a)Results

ofmeasured

white

pointan

dsum

ofindividu

alwavelength

spectrawithinterference

domean

ddistan

ce=30

cm

300

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Abs

olut

e Ir

radi

ance

for

Max

imum

Set

ting

Ligh

t A (

No

Dom

e) (

dist

ance

= 3

0cm

)

M

easu

red

Irra

dian

ceS

uper

posi

tion

of Ir

rand

ianc

e

(b)Results

ofmeasured

white

pointan

dsum

ofindividu

alwavelength

spectrawitho

utinterference

domean

ddistan

ce=30

cm

Figu

re5.6:

Inthisexpe

riment,thespectrom

eter

integrationtim

ewas

setfor

each

wavelen

gthto

measure

theerrorb

etweenestim

ating

awhite

pointusingsupe

rposition

.The

grap

hsalso

show

thediffe

renc

ein

radian

ten

ergy

betw

eenthesamelig

htsource

with

ado

me

andwith

out.

The

data

acqu

iredin

theseexpe

riments

areused

foralls

ubsequ

entmod

eling.

56


TestMetho

da

MeasuredLux

Estim

ated

Lux

b%

Differen

ceEfficacy

Dom

e55

5lux

572lux

3%30

(lux/W

)NoDom

e23

23lux

2443

lux

5%12

6(lu

x/W

)

Table5.1:

Summaryof

results

forthesecond

absolute

irrad

ianc

emeasurement.

a distance=

30cm

b Fou

ndusingthesupe

rpositionof

thefiv

ewavelengths

57


Dataset Mean CRI Mean lux Mean Efficacy Mean ∆uv

linear:500 lux 65 500 lux 136 (lux·W−1) 0



overdetermined:500 lux 76 500 lux 141 (lux·W−1) 0



Table 5.2: Summary of linear equality testpoints. Each data set is comprised of 11 whitepoints ranging from warm (2800 K) to cool white (10000 K). These datasets were generatedusing irradiance measurements with no dome at a distance of 30 cm. The measurement,mean ∆uv, is the Euclidean distance to the actual chromaticity coordinate of the colortemperature in CIE 1964 color space. In this test dataset, we supply the precise coordinatescorresponding to the color temperature and seek an exact solution of the form A−1b = x(exact) or A−†b = x (overdetermined), thus we expect that ∆uv is zero.

Predicted Results using Monte Carlo Simulations

Figure 5.7 and Figure 5.8 show the resulting contours for randomly picked setpoints basedupon the absolute irradiance experiments using the interference dome. In these graphs,the illuminance of the system is fixed at an arbitrary point and the resulting Ra is givenas a function of Tc and efficacy. In this experiment the illuminance, Ev is “constrained”by limiting which lux values are within the range we want to graph. This is analogous tothe linear inequality constaints given in Sec 3.1. In these experiments the system does notemploy dynamic efficacy. Instead based on our assumptions of the wavelengths present inthe system, two sets of simulations are performed. One set uses five wavelengths and theother set uses three wavelengths. Figures 5.7 and 5.8 are the contour plots representingpossible solutions.

Predicted Results using Linear Methods

Using the techniques discussed in Section 3.2, we proceed to generate the setpoints usinglinear methods. These setpoints correspond to 11 different color temperatures (warm tocool) for 500 lx, 1000 lx, and 1500 lx. These data were generated using both the exactsolution and pseudoinverse methods (see Figure A.1 in Appendix A for the test spectra).They are summarized in Table 5.2.

Predicted Results using Direct Search

The Global Optimization Toolbox in Matlab [71] contains the generalized pattern searchand mesh adaptive direct search optimization algorithms. As discussed Chapter 3, the Mesh

58


Correlated Color Temperature (CCT)

Effi

cacy

(Lu

x ⋅ W

−1 )

Contour Plot of Color Rendering Index using 5 Wavelengths (215 Lux, 30cm)

3000 4000 5000 6000 7000 8000 9000 1000026

28

30

32

34

36

38

40

425 Wavelength Monte Carlo Data

Col

or R

ende

ring

Inde

x0

10

20

30

40

50

60

70

80

90

100

Figure 5.7: Results of Monte Carlo simulation using five wavelengths for approximately215 lux. These results are from the irradiance data using the interference dome anddistance=30 cm. For any given color temperature, there exists plenty of local minima tomake maximizing the color rendering index a difficult problem. 1,000,000 data points weregenerated, and then filtered to ANSI white point standards and limited to approximately215 lux. The simulation took eight hours to complete on a Intel Core 2 Quad.

59


Correlated Color Temperature (CCT)

Effi

cacy

(Lu

x ⋅ W

−1 )

Contour Plot of Color Rendering Index using 3 Wavelengths (215 Lux, 30cm)

3000 4000 5000 6000 7000 8000 9000 1000026

28

30

32

34

36

38

40

423 Wavelength Monte Carlo Data

Col

or R

ende

ring

Inde

x

82

83

84

85

86

87

88

89

90

Figure 5.8: Results of Monte Carlo simulation using three wavelengths for approximately215 lux. These results are from the irradiance data using the interference dome anddistance=30 cm. For any given color temperature, there exists plenty of local minima tomake maximizing the color rendering index a difficult problem. 1,000,000 data points weregenerated, and then filtered to ANSI white point standards and limited to approximately215 lux. The simulation took seven hours to complete on a Intel Core 2 Quad.

Adaptive Direct Search algorithm generated the testpoints summarized in this section (seeSec. 3.3, as well as, [8] and [74]). Using the objective function defined in Chapter 3, theoptimal setpoints for the five wavelengths of light A were found. Six data sets were generatedfor a range of color temperatures (Figure A.3 on page 88). These are: optimized efficacyat 500 lux, optimized efficacy at 1000 lux, optimized efficacy at 1500 lux, optimized colorrendering index at 500 lux, optimized color rendering at 1000 lux, optimized color renderingat 1500 lux (see Table 5.3 on page 61). Each data set completed in just under three minutesusing an Intel Core Duo Quad. The settings used for the solver are given in Table 5.4.

60

5.4. Measured Results


Efficacy:500 lux 67 490 lux 191 (lux·W−1) 0.0472



CARI:500 lux 94 500 lux 155 (lux·W−1) 0.0070

CRI:1000 lux 94 996 lux 155 (lux·W−1) 0.0076

CRI:1500 lux 93 1523 lux 157 (lux·W−1) 0.0199

Table 5.3: Summary of testpoints obtained from the nonlinear solver (MADS). Each data setis comprised of 11 white points ranging from warm (2800 K) to cool white (10000 K). Thesetests points were determined using the irradiance measurements in Sec. 5.2 with no dome at30 cm. The measurement, mean ∆uv, is the Euclidean distance to the actual chromaticitycoordinate of the color temperature in CIE 1964 color space. It is a quantitative way ofdescribing the error in chromaticity for a particular color temperature.

Setting Value

Search Step MADSpositivebasisNp1Poll Step MADSpositivebasis2N

Max Iterations 4000Complete Poll On

Complete Search On

Table 5.4: Specific settings used in the configuration MATLAB function patternsearch(),a nonlinear solver (capable of handling linear and nonlinear constraints, as well as, bounds),that is part of the Global Optimization Toolbox. Specifically, the MADS algorithm [8] wasused to obtain the direct search testpoints.

5.4 Measured Results

Here we present the measured results of of the linear and direct search test methods forsetting light fixture A, with no dome.

Linear Methods

The results of the setpoint found using linear methods is summarized in Table 5.5. Theacquired spectra are found in Figure A.2 on page 87.

61

5.5. Sensor Data


linear:500 lux 55 504 lux 130 (lux·W−1) 0.0100



overdetermined:500 lux 67 499 lux 132 (lux·W−1) 0.0085



Table 5.5: Summary of the measured results of the testing data sets. Each data set iscomprised of 11 white points ranging from warm (2800 K) to cool white (10000 K). Thesedatasets were generated using irradiance measurements with no dome at a distance of 30 cm.





CRI:500 lux 92 490 lux 143 (lux·W−1) 0.0021

CRI:1000 lux 92 933 lux 139 (lux·W−1) 0.0047

CRI:1500 lux 90 1378 lux 138 (lux·W−1) 0.0174

Table 5.6: Summary of the measured results of the testing data sets found using the directsearch method. Each data set is comprised of 11 white points ranging from warm (2800 K)to cool white (10000 K). These datasets were generated using irradiance measurements withno dome at a distance of 30 cm.

Direct Search

The results of the setpoint found using the MADS algorithm is summarized in Table 5.6.The acquired spectra are found in Figure A.4 on page 89.

5.5 Sensor Data

For one week, a mobile cart (the same cart in Figure 5.1b on page 50) was used to collectdaylight and artificial light indoors at the Media Lab. Additionally, direct sunlight wasalso collected, but resulted in sensor saturation. The experiment was designed to collect adataset of different color temperatures and lux values used for optical calibration. Usingthese data, we calibrate a sensor node able to measure the lux and color temperature ofdaylight, incandescent, fluorescent, and LED sources (Figure 5.9 on page 63).

62

5.5. Sensor Data

4000

6000

8000

10000

2

4

6

x 104

60

70

80

90

100

Color Temperature (Kelvin)

Lighting Data Collected for Sensor Calibration

Lux (lm ⋅ m−2)

Col

or R

ende

ring

Inde

x

Interpolated SurfaceMeasured Data Points

Figure 5.9: Using a spectrometer, absolute irradiance was collected and then processed toestimate the color temperature, lux, and color rendering index of multiple light sources.

63

5.6. Estimating the Ambient Light

0 2000 4000 6000 8000 100000

1.6

3.2

4.8

6.4

8x 10

4

TC

S34

14 C

ount

s (M

ax 6

5535

)

Dynamic Range of Lux Sensors

0 2000 4000 6000 8000 100000

1000

2000

3000

4000

5000

Spectrometer Calculated Lux (lm ⋅ m−2)

ISL2

9006

Cou

nts

(max

409

5)

TCS3414 Clear ChannelISL29006, 2000 luxISL29006, 6000 lux

Figure 5.10: Dynamic range of the TCS3414 and two ISL29006 optical sensors set to differentsensitivities. The TCS3414 has an on board analog to digital converter. The ISL29006 aremeasured by the on board analog to digital converter. In this experiment, the TCS3414was set to oversample 10 times (the highest sensitivity) in order to estimate the level ofilluminance required to saturate the sensor. Nominally, the range is set to approximately10, 000 lx.

The lux measured by the optical sensors (TCS3414 clear channel, and ISL29006) is givenin Figure 5.10. The corresponding color-filtered data are given in Figure 5.11.

5.6 Estimating the Ambient Light

Several tests were conducted with light source A, with no dome, mounted at 30 cm from thesensor node to study the time it takes to calibrate a single light and measure the ambientlighting (see Table 5.7 on page 67). In this study, the unfiltered analog lux sensor, set

64


02

46

8

x 10

4

01234567

x 10

4

Lux

(lm ⋅

m−

2 )

Sensor Counts (max 65535)

TC

S34

14 R

ed C

hann

el

R

ed D

ata

(a)

02

46

8

x 10

4

01234567

x 10

4

Lux

(lm ⋅

m−

2 )


TC

S34

14 G

reen

Cha

nnel

G

reen

Dat

a

(b)

02

46

8

x 10

4

01234567

x 10

4

Lux

(lm ⋅

m−

2 )


TC

S34

14 B

lue

Cha

nnel

B

lue

Dat

a

(c)

Figu

re5.11:Red

,green

,and

blue

chan

nelr

eading

sfrom

theexpe

rimentin

Figu

re5.9.

65


to a sensitivity of approximately 5000 lx, was tested. This purpose of this experimentwas to measure the total system update rate (including read and writes from MATLAB)and to choose an integration time for the digital color sensor depending on the results ofthe experiments. In each experiment, 200 data points were gathered by toggling the lightsource and querying the sensor board in MATLAB. No oversampling of the sensors wasused. Instead, these data were averaged post-experiment. In each experiment, the delayswere set such that the light source was on and off for an equal amount of time. The polltime in the table is the total period of each experiment step. Therefore, 150 ms means thelight was on for 75 ms and off for 75 ms.

66


Con

dition

Lux

Ambient

0lx

OverheadLigh

ts22

1lx

LEDs

2044

lx(a)

Test

µoff(C

ounts)

σoff(C

ounts)

µon(C

ounts)

σon(C

ounts)

µpoll(m

s)σpoll(m

s)

Ligh

tsOn15

0ms

574

1709

715

6.3

2.1

Ligh

tsOff15

0ms

03

1646

315

6.2

2.1

Ligh

tsOn90

ms

482

1706

393

.93.0

Ligh

tsOff90

ms

00

1641

393

.83.0

Ligh

tsOn(non

e)10

2582

988

583

031

.32.1

Ligh

tsOff(non

e)89

082

581

582

431

.32.1

(b)

Table5.7:

In(a),theam

bientan

dtestingcond

ition

saregiven.

The

seweremeasuredusingthespectrom

eter.The

LED

light

source

was

placed

30cm

away

from

thehead

.In

(b),themeanan

dstan

dard

deviationaregivenford

ifferentp

ollin

gfrequenciesa

nddiffe

rent

cond

ition

s.

67

Chapter Six

Analysis

6.1 Sources of Error

This section begins by reviewing the sources of error in both the test setup and the spec-trometer. Park et al. [50] provide a complete discussion of uncertainty and error propagationin spectrometer-based LED measurement. Hwang et al. [30] provide an analogous treatmentof uncertainty in color measurement using a spectrometer.

Measurement Error

All subsequent calculations in this thesis are derived from the primary step of measuring theabsolute irradiance of the test fixture (see Figure 5.1 on page 50). There are four primarysources of error in the test setup. The first source of error is the estimated distance betweenthe LED array and the spectrometer head. The second source of error is the accuracy ofthe LED forward current. The third source of error is the specific junction temperatureof the LEDs on the array. The fourth source of error relates to the actual spectrometerirradiance measurement itself. In [50], the major components of uncertainty governingthe irradiance measurement error are: spectral irradiance reference data, calibration lampdetector distance, calibration lamp current, repeatability, wavelength accuracy, linearity,and spectral stray light.

Spectrometer Induced Error

All experiments in this work are preformed using a USB4000 spectrometer from Ocean Op-tics. The monochromator is an Czerny-Turner (see Wyszecki and Stiles [79]) with a 25µmentrance slit and a diffraction grating blazed at 500 nm (Ocean Optics grating numberH3). The spectral bandwidth (Full Width Half Maximum) of the monochromator is ap-proximately 1.5 nm with stray light estimates of <0.05% at 600 nm; 0.10% at 435 nm. Thedetector is a Toshiba TCD1304AP (200-1100 nm) with a corrected linearity of >99.8%. The

68

6.1. Sources of Error

sensitivity of the CCD is approximately 130 photons/count at 400 nm; 60 photons/countat 600 nm. The calibration source is Ocean Optics HL-2000 Tungsten Halogen (traceableto NIST S/N: F-211), with a drift of approximately .3% per hour of use. The radiometriccalibration source used in this research has an initial uncertainty of approximately 2.36%.The supply is regulated to within 0.5% of rated specifications.

Assuming a 3% certainty on mechanical setup and a 4% uncertainty regarding reflectedand scattered light, a back of the envelope calculation of the major sources of error allow usto estimate the expanded uncertainty (k = 2) of 14%, with the individual uncertainties ofstray light, linearity, etc., accounting for only 2% to 4% of the uncertainty. However, withoutformal testing of the individual sources above, these are only estimates. In a similar setup,Park et al. estimated the expanded uncertainty (k = 2) of their calculations to be: 4.33% forred wavelengths, 4.22% for green wavelengths and 5.54% for blue wavelengths. As a generalrule, 10% uncertainty is considered very good for spectrometer measurements. Uncertaintiesless than 2% are achievable only by NIST [57]. Additionally, the transferred uncertaintyin the calibration source was initially estimated to be close to 1-2% in the visible range.In order to properly validate the test setup, the prototype light source should be sent to athird party, such as Luminaire Testing Laboratory (LTL).

Error in Linearity Assumptions

Although understanding the sources of absolute error are important for comparison to othersystems (i.e., how this calibrated light source may appear at another optical instrument) itis also important to quantify the error introduced into the system due to assumptions oflinearity. There are two major sources of error. The first is exemplified in Figure 5.3 onpage 53. In a strictly linear system, the only information needed to estimate the intensityof any given light wavelength is the maximum intensity and the PWM setting. With nolinear fit to minimize square error, when the wavelength is dimmed, the effect on the systemis that the light source is brighter than the model anticipated. Accordingly, this manifestsitself as chromaticity error and intensity error in setting the color temperature of the lightsource. Warmer colors, which use more red intensity, are strongly affected.

The second major source of error is estimating the white point, or the “all-on” setting,of the system. This is exemplified in Figure 5.6 on page 56 where the graph depicts themeasured white point and the estimated white point, which is found by summing the totalspectra for the individual wavelengths. Not surprisingly, the peaks of the five wavelengthsin the system are lower for the measured white point. This effect, like the deviation fromlinearity in Fig 5.3, is due to temperature. The effect here is that the overall lux of the lightsource will be less than expected, particularly at the sensor node some distance from thelight source. Additionally, this also affects the color temperature of the system.

Previous work attempted to compensate for these temperature effects by directly mea-suring the temperature and following the intensity derating curve for the LEDs [6, 43, 3].

69

6.2. Performance of the Sensor Node

Yet, it is difficult to properly estimate the junction temperature with only a single temper-ature sensor on board. One reasonable approach is closing the loop (e.g., using feedback totune the individual wavelengths as discussed in Sec. 3.5). Additionally, we may also wantto remove as much open loop error as possible. It may be desireable to begin with a correctmodel of the system, therefore we present a technique to correct the individual wavelengthsbased on a white point measurement, or series of white point measurements. We beginby correcting the acquired spectra to account for the worst case operating temperature(assumed to be the white point) such that

minxf(x) = ‖Ee,wp(λ)−

m∑i=1

xiEe,i(λ)‖2, (6.1)

subject to x ∈ R : 0 ≤ x ≤ 1. In this case, we seek a solution to minimize the squareerror between the measured whitepoint and the weighted sum of the individual intensities.Yet, because this only captures a single white point, we go one step further and write

minxf(x) =

n∑j=1‖Ee,wp(λ)−

m∑i=1

xiEe,i(λ)‖2, (6.2)

subject to x ∈ R : 0 ≤ x ≤ 1. In this case, n represents the total number of whitepoints taken, for example with increasing intensity. The reasoning behind Eq. 6.2 is thatin Eq. 6.1, we assume that this relationship is fixed for all the operating points, when infact it is dynamic. By taking into account multiple white points, we achieve an open loopaccuracy theoretically higher than any line fitting method of a single LED. If we applieda linear fit to the data in Fig. 5.3, it will be incorrect because it does include the heatingeffect of the other LEDs. If the system is going to be running open loop color control, it iscritical to perform these steps. Yet, even with many acquired white points, there are toomany cases to take into account, and closed loop control may be required in certain cases.

Figure 6.1 shows the results of minimizing Eq. 6.1 for the data collected in Sec. 5.2. Thecorresponding correction factors are blue: 0.9373; cyan: 0.9255; green: 1.0000; amber: 0.9120;red: 0.8901. Then, using these corrected spectra, the optimal setpoints using either thedirect search or linear methods are computed. To see how these methods would improvethe results obtained in Sec. 5.4, we can measure the difference between measured resultsobtained when using the temperature-compensated superposition and the measured resultsusing just superposition.

6.2 Performance of the Sensor Node

The primary goal of collecting the data in Figure 5.9 on page 63 is to calibrate the digitalcolor sensor so that the color temperature of ambient lighting can be inferred. Early modelsof performance predicted that a calibration source with a sufficiently high color renderingindex would make a suitable adjustable calibration source for the digital color sensor. Using

70


300 400 500 600 700 8000

1

2

3

4

5

6

7

8

9

10

Wavelength (nm)

Abs

olut

e Ir

radi

ance

(µW

× c

m−

2 )

Absolute Irradiance for Maximum Setting Light A (No Dome) (distance 30cm)

Corrected SuperpositionMeasured IrradianceSuperposition of Irradiance

Figure 6.1: Results of minimizing the square error between the measured white point andthe sum of weighted spectra taken during a calibration step. If this step is performed beforeusing the linear and direct search methods to find the color temperature and intensity, theerror between the model and the actual system will be decreased. The setpoints foundduring these steps will better reflect the worst case operating temperature.

71


the calibration methods described in Sec. 3.5, a single sensor was calibrated using 31 whitepoints. This served as the LED calibration source. Additionally, daylight was collected(presented in Sec. 5.5) which served as the daylight calibration source (incandescent bulbswith multiple filters can function similarly). In Figure 6.2, the performance of the differentcalibration methods is given.

The results suggest that the method of calibration (either a five wavelength source ordaylight) is the biggest predictor of error. For example, in Figure 6.2b we see that an LEDsource, even with an average color rendering index of 92, is not an adequate calibrationsource to correctly measure daylight. Of interest is the result in Figure 6.2a, in which thedaylight calibration of the sensor produced a near linear output, but with a different slopethan the ideal response. With an appropriate correction factor, the error between measuringdaylight and artificial light sources could be minimized. More important here is the intendeduse of the sensor. Its purpose is not to measure with absolute certainty the correct colortemperature (it would be nice if that were the case), but instead provide an output thatcould be used to synchronize the indoor lighting with daylight environment automatically.

The secondary function of the sensor node is measuring the illuminance of the ambientlight and the adjustable light sources in the network to optimize the lighting at the point ofmeasurement. From testing, it is clear that a query and response command for the analoglux sensors may not be the best solution for control via MATLAB. The reason is due to thelack of resolution in the MATLAB pause function.

In Section. 5.6, performance metrics were given for the unfiltered lux sensor. It wasdetermined that because of the poor resolution of MATLAB’s pause function, strict timingwas not achievable. Therefore, the limitations of update rates (at present 11Hz) are tooslow in order to perform a room calibration without noticing the flickering of the lights.There exists the possibility of continued use of MATLAB with an augmented pause functionthat utilizes Microsoft Window’s high resolution clock. Thus, it may still be possible toachieve flicker-free calibration with only a minor software change. More testing needs to beperformed to firmly comment if this is possible using MATLAB. If not, another solution mustbe found. Additionally, the RC time constant of the 50 Hz low pass filters (approximately3 ms) theoretically would not hinder operation, but the present limitations of the timingresolution in MATLAB prevented any significant exploration.

The data presented in Figure 5.10 on page 64 were fit to the linear equation y = mx+ b.From Fig. 5.10, it is clear that the analog sensors saturate close to 2000 lx and 6000 lx. Thedigital sensor was set to 10× integration time for maximum sensitivity, and it saturated closeto 5000 lx. The ambient light levels on the fifth floor of the Media Lab average between250 lx to 2000 lx near a window. Reporting the answer in lux rather then sensor counts maynot be an issue, as the user is concerned only about the lighting on the surface. It is thegoal of the system to minimize the other parameters to match what the user expects to bepresent. The main issue here is not sensitivity, but update rate. Increasing the integrationtime of the digital sensor may lead to poor response times as transitions across the lighting

72


2000 4000 6000 8000 10000 120002000

4000

6000

8000

10000

12000

True Color Temperature (Kelvin)

Est

imat

ed C

olor

Tem

pera

ture

(K

elvi

n)

Color Temperature Calculation of Test Data Ra1500 (LED Source)

LED CalibrationDaylight CalibrationIdeal Response

(a) Results of calibration methods when measuring an LED source.

2000 4000 6000 8000 100002000

4000

6000

8000

10000

12000

14000

16000

18000

True Color Temperature (Kelvin)

Est

imat

ed C

olor

Tem

pera

ture

(K

elvi

n)

Color Temperature Calculation of Sunlight

LED CalibrationDaylight CalibrationIdeal Response

(b) Results of calibration methods when measuring daylight.

Figure 6.2: Comparison of calibration results. In (a) the color sensor is calibrated usingeither an LED source or daylight and its response to LED irradiance is given. In (b), thecolor sensor is calibrated the same way, but its response to daylight irradiance is given.Clearly, the results correlate well with the calibration method employed. However, in (a)the linearity of the daylight calibration method may make it applicable in measuring a broadrange of sources.

73

6.3. Performance of Control Methods

network will be averaged by the sensor. More testing is required to settle on the correctintegration time.

6.3 Performance of Control Methods

The experiments conducted in Sec. 5.4 are considered proof of concept measurements. Theyare intended to validate direct search as a new method of polychromatic intensity controlfor both a single fixture and a network of lights. Additionally, as discussed in Sec. 6.1 themeasurements were conducted with datasets that were not temperature compensated, soin some cases there are significant chromaticity errors. It was decided, for brevity, thatthe results in Table 5.5 and Table 5.6 be given as an average so that a general sense ofperformance could be quickly ascertained. Because the average results are presented, specificcolor comparisons (11 colors were tested) are not possible. These tables of results are onlyappropriate once the model used to generate the datasets is temperature compensated sothat the true open loop response can be measured. In Figure 6.4, the mean CRI and efficacyare presented for the tested light source.

Indeed, as suggested in Chapter 3, the direct search method finds the intensities for theindividual wavelengths such that CRI and Efficacy are maximized (see Fig. Figure 6.3 onpage 75). In theory, the variation across the three intensity groups (500 lx, 1000 lx, and1500 lx) occurs due to errors in the linearity assumptions discussed in Sec 6.1. Neither CRInor efficacy should vary if the model is perfect. In the worst case estimate, using directsearch results in a 45% improvement of the color rendering ability of the light source overthe linear method. Similarly, maximizing the efficacy results in a light source 28% moreefficient than its linear counter part. It is clear that these methods have great potential forcontrolling the spectral composition of the source.

On the other hand, in Figure 6.4b, we see the effect of maximizing efficacy. The objectivefunction will attempt to use the most efficient and fewest wavelengths possible to create thecorrect color temperature. Depending on the intensity and color, this could be two or threewavelengths. The penalty incurred for such efficiency is that we lose the quality of the whitepoint. The ∆uv reported in the figure is a measure of how close the measured white pointis to the blackbody curve. As a rule of thumb, a ∆uv < .01 is considered acceptable. Giventhe uncertainty of the spectrometer measurements to begin with it is difficult to commentabout the true error regarding the white point (there is much room for improvement here).But qualitatively, we would notice a significant change of color if these white points wereto be used. This could be improved by adding a coefficient in the evaluation function ofthe optimization problem, or by defining ∆uv as a nonlinear inequality contraint in theoptimization routine, or even by fixing the minimum number of wavelengths to be three,and not two (redefining the lower and upper bound of the optimization problem). This is arelatively small problem and is easily fixed (potentially at the cost of run-time performance).

74


500 lx 1000 lx 1500 lx0

20

40

60

80

100

Intensity

Col

or R

ende

ring

Inde

x (R

a) Results of Measured Color Rendering Index

Linear ExactLinear PseudoinverseDirect Search EfficacyDirect Search CRI

(a) Results of average CRI.

500 lx 1000 lx 1500 lx0

20

40

60

80

100

120

140

160

180

Intensity

Mea

n E

ffica

cy (

lx ⋅

W−

1 )

Mean Measured Efficacy (lx ⋅ W−1) at 30cm


(b) Results of average efficacy.

Figure 6.3: Results obtained from study of linear and nonlinear methods of color control.Here, we see the effectiveness of the direct search algorithm. It is capable of outperformingthe exact and overdetermined methods of control. A CRI score of 100 is considered perfect.In (b), the algorithm is able to create colors 28% more efficiently then the other methods.

75


500 lx 1000 lx 1500 lx0

250

500

750

1000

1250

1500

Intensity

Mea

n Lu

x (lm

⋅ m

−2 )

at 3

0cm

Mean Measured Lux (lm ⋅ m−2) at 30cm


(a)

500 lx 1000 lx 1500 lx0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

Intensity

Mea

n ∆u

v

Mean Chromaticity Error (∆uv)


(b)

Figure 6.4: Results obtained from study of linear and nonlinear methods of color control.In (a), the variation in illuminance is due to open loop error. In (b), the large ∆uv errormeans that the resulting color temperature is not exactly on the blackbody curve. Maximingefficacy resulted in lower quality white light.

76

6.4. Comparison to Commercial Systems

6.4 Comparison to Commercial Systems

The Caliper program, sponsored by the US Department of Energy, is responsible for docu-menting the progress and developments with SSL-related technology. In a recent publication(2009), it was reported that:

“The general trend of increasing efficacy continues, with Round 9 products aver-aging 46 lm/W and ranging from 17 to 79 lm/W. Unfortunately, wide disparitiesare still observed in the accuracy of manufacturer specifications. SSL recesseddownlights are now capable of meeting or exceeding light output levels and lightdistribution characteristics of downlights equipped with 45-75W incandescentand halogen lamps, and the SSL products are meeting or exceeding color qualityof CFL recessed downlights [48].”

These fixtures are all phosphor-based designs. The color rendering index of the fixtures inthe Caliper report range from 63 to 93, with a majority of the tested fixtures having a CCTranging from 2700 K to 6500 K, none of them being color adjustable.

We can estimate the performance of the prototype presented in this thesis for a range ofcolor temperatures as well (see Table 6.1). In this case, we consider the results obtained inTable 5.6 to serve as our comparison. In Figure 5.2 on page 52, we measured the test sourceto study if the relationship Ev = Iv/d

2 holds. Using this assumption, we can estimate thetotal candelas of the test fixture at a desired operating point (we assume we measured thelux directly under the source). First we estimate the total lux at a distance of 1 m from thesource by applying Eq. 3.2 rewritten to estimate the illuminance at two different distancesmeasured from the same source. The illuminance, now estimated at 1 m, can be convertedinto candelas. The next step involves calculating the solid angle of lamp (in steradians)which is 2π(1− cos(α/2)), where α is the beam angle of our source, which is estimated bybe roughly 110 degrees. We obtain the lumens by multiplying the intensity with the solidangle, Iv2π(1− cos(α/2)).

Several Color Kinetics iW Blast Powercore lamps were obtained and characterized in aset of experiments outlined in Chapter 5 as a commercial test fixture most analogous to theone developed. The iW Blast Powercore is a phosphor-converted adjustable white fixture. Itconsists of LEDs at two color temperatures, 2700 K and 6500 K, and allows the channels tobe mixed. The fixture is rated at 50 W at maximum intensity, and consists of two banks ofwarm white LEDs and one bank of cool white LEDs. The range of achievable output is fixedbetween these two color temperatures. The user controllable resolution of three channels islimited to eight bits where the input/output relationship (PWM versus intensity) followsthe power law, y = xk. The measured CRI of the light source varies with temperature andhas a range of 72 to 83 and an efficacy of 35 lm/W.

Phosphor-based designs already dominate the market and the chronological Caliper datasuggests that improvements across the board are happening. Phillips et al. [52] discuss the

77

6.5. Energy Saving Potential

Tc (kelvin) Lux (lm·m−2) Lumens (lm) power (W) efficacy (lm/W)

2748 98 263 8 332978 101 271 8 343538 108 289 8 364085 117 313 9 354538 124 332 10 334930 132 354 10 355382 135 362 11 335830 137 367 11 336335 138 370 12 317768 137 367 12 319613 137 367 12 31

Table 6.1: Estimated luminous efficacy of prototype light source for the measured lux ob-tained in Sec. 5.4.

future of phosphor and active emitter designs. The LED systems in the latest Caliperreport surpass most compact fluorescent designs, both in terms of efficacy and quality. Yetwe are still left with the problem of intelligent control and to some degree, the mixing ofcolor. There more than likely will still be a need in niche markets for adjustable systems,especially ones in which the reproduction of color temperature and color rendering areimportant. Over time, commercially available systems with the color adjustable range andcolor quality of the prototype presented in this thesis will begin to surface. From a designstandpoint, especially with regard to the lighting industry, color-adjustable systems arecomplex, both mathematically and physically. Above all, lower efficacy and higher costmake them less appealing to the mass consumer market. Certainly, if the goal is to maximizethe $/Klm ratio, it is not achieved using active emitters. In order for active emitters tobecome dominant in a large cross section of the market, the quantum efficiency of the greenemitters will need to improve to the levels of near ultra-violet emitters. From a engineeringperspective, the techniques presented here are just as applicable to color tunable whitesystems as they are to polychromatic active emitter systems.

6.5 Energy Saving Potential

The affordances offered by color-adjustable systems are broad, yet a great barrier in thebroad adoption of these devices is due in part to cost, lower efficacy and complexity. Oneidea explored in this work is dynamic efficacy, or the ability of the lighting system to optimizethe number of wavelengths used to create the white point. From an engineering standpoint,

78


the technique works quite well, however one issue remains: if given the choice between ahigh quality light and a poor quality one, which one do you choose? In order to maximizethe full potential of adopting a polychromatic system, the lighting must not burden theuser. A probable answer is that either directly, or indirectly, we do not really care as longas we are satisfied.

We turn our attention to estimating the performance of the network considered in Fig-ure 4.1. We consider a hypothetical illuminance profile of ambient light (e.g, daylight)collected in an office in MIT building E14 (Figure 6.5). In this example, the peak illu-minance, estimated to be 500 lx is measured during midday. The offices in building E14currently use compact fluorescent bulbs and the illuminance, 1.5 m from the plenum, wasestimated at 325 lx. From Figure 6.5, we see that for 1/3 of the day, the lighting level is300 lx or greater, which, to guarantee 325 lx at the work surface, would require 25 lx orless from the fixtures. By midday, according to the figure, the lights could be turned off.Although this is a completely a hypothetical situation, the implications are that the fixtureswould require less than 3% of their total output during the course of the normal wakinghours.

But, let us consider a case where the lighting network runs continuously for 24 hourseach day and is configured to maintain 325 lx at the work surface. By compensating fordaylight, we can infer that average illuminance in the room from the overhead lighting,over a period of a day, is 180 lx/h. This constitutes a 45% reduction if the lights were leftrunning at full intensity, 24 hours a day. In theory, another 1/6 of that time no illuminationis required, and in this case, the reduction is 59%.1 The example given was extreme, as inmost cases, the total ambient lux may never exceed the capabilities of the lighting system.Also of consequence is that the fact that the distribution will not be Gaussian, but erraticand most certainly, not smooth.

The idea behind dynamic efficacy is to augment closed loop intensity control by modu-lating the spectral composition of the source and automatically determining at which timesa more efficient light, rather than higher quality light, satisfies the user’s requirements. Theimportant distinction between dynamic efficacy and closed loop intensity control is thata reduction in overall energy usage is achieved without reducing the intensity and outputcolor. For example, let us consider the system whose results are described in table 5.6.Here, we see that the higher efficacy modes of operation are 23% more efficient than theirhigh CRI counterparts. This means that for the same intensity, the high CRI mode uses 1.2times more energy then the high efficacy mode.

If we reconsider the example above, we can imagine an operational scenario in which thenetwork determined situations and times where, although it could not adjust the intensity,it could tune the efficacy instead. In the example above, we assume that maintaining 325 lx

1Here we computer the average of maintaining 325 lx 5/6 of the time, and 0 lx the other 1/6, whichcorresponds to an average of 244 lx/h. For the adjustable intensity it becomes 0 lx for 50% of the day whichcorresponds to an average of 100 lx/h.

79


0 0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200 24000

50

100

150

200

250

300

350

400

450

500

Time of Day (Hours)

Mea

sure

d Lu

x

Hypothetical Daylight and Lighting in an E14 Office

Ambient LuxOffice LightingAdjusted Lighting

Figure 6.5: Hypothetical distribution of illuminance in an office in E14, with a maximumreading of 500 lx during midday. In this scenario, there are two lighting models used. Thefirst model assumes the lights remain on at a constant intensity during the day. The secondmethod assumes the intensity of the sources are adjustable.

80


on the surface of interest requires 40 W of electrical power and that decreasing the intensityof the system corresponds linearly with power, that is, 180 lx consumes 22 W of power. A40 W light, left on for 18 hours a day uses 720W·h. In the hypothetical 58% scenario, theequivalent power consumption of the light source is 16W each hour (16W·h). This leadsto a total consumption of 400W·h per day. By using feedback, we reduce the total energyconsumed by the system by 44%.

By adjusting the intensity throughout the day, we changed the equivalent power con-sumption of the light source, that is, a 40 W lamp “became” a 16 W lamp. Earlier, it wasstated that high CRI mode of operation uses 1.2 times the power of the high efficacy mode.If we assume that the equivalent 16 W source is running in high CRI mode, then for samelight level, the rated power becomes 13 W, a change of 18%. If the fixture ran in highefficacy mode for 24 hours, this is equivalent to 312W·h per day. Depending on the timespent using high efficacy or high CRI, we could save up to an additional 88W·h. We couldassume the efficacy mode is used 50% of time and our corresponding consumption becomes356W·h. Over a year, this saves 16 kW·h for just a single light source.

Compensating for ambient lighting and detecting occupancy in an office is not new; weare all familiar with the technology such as the building sensor that controls outdoor lightingwhen the sun sets or timer based lighting in the library stacks. However, these systems,lacking any cohesive infrastructure, cannot dynamically raise and lower the intensity in re-sponse to changing ambient conditions. Instead, for simplicity they are binary; either offor on (almost always off at the wrong time). For many systems this is adequate, but thereare still plenty opportunities left to pursue. Larger systems, designed to control the lightingin office type environments, typically have the ability to zone the lighting and adjust theintensity of individual lamps on the network. By far, the most prevalent method of con-trolling these networks is human input, using either change on demand or pre-programmedsequences. One example is the Lutron Grafik Eye, installed in E14. Yet, this system, for allof its capabilities, lacks the “off the shelf” components to realize the dynamic adjustment ofintensity to compensate for daylight. However, these systems are capable of being controlledusing a computer – it is not unreasonable to think that that the hardware and techniquespresented here could one day control the overhead fluorescent lighting at the Media Lab.

81

Chapter Seven

Conclusions

This thesis focused on the development, modeling, and testing of dynamic solid state light-ing. In this work, state of the art methods were presented to optimize the optical and energyrelated parameters of an intelligent lighting system. A comparison of control techniques,both linear and nonlinear, provided quantitative evidence that the methods described in thisthesis are optimal for the control and design of future polychromatic as well as phosphor-based lighting systems. In addition, the feasibility and current performance of room calibra-tion and the performance of a novel method of measuring the correlated color temperatureof daylight were also given. In short, the broad foundation for future research on the op-timal control of lighting has been both quantitatively and qualitatively discussed. Withthe proper techniques, subsystems, and methodologies in hand, it is now time to considerdynamic interaction with the network. The next immediate portion of the research willstudy the effects of the computer-controlled network of multiple light sources.

7.1 Overview of Performance

The network, as designed right now, has several bottlenecks. These are manifested as querydelays to sensor nodes and light fixtures, and timing expense incurred from roundtrip fullduplex communication. A majority of the proposed control algorithms utilize MATLAB,which when combined with Windows, is hardly a real-time OS. In the short term, minorfirmware revisions to correct expensive and consecutive read/write operations are critical.Because the optimal control of a single fixture and sensor node was considered, there maybe unexpected delays as the network scales. Now, after testing and reporting on the perfor-mance of the sensor nodes, the gain of the analog and digital sensors can be locked down.The main hindrance in transparent operation of the network is going to be subtly measuringthe ambient and current intensity of the uncontrolled and controlled lighting in the room.Currently, this update rate is 11 Hz.

In Chapter 3, the use of direct search was presented as an efficient nonlinear solver for thelighting network. It is not an issue of being able to accurately determine the correct intensity

82

7.2. Future Work

for the total number of wavelengths present in a lighting network; this was demonstrated.It is an issue of speed. For example, one possible solution is use to the direct search methodto calculate a table of intensities to be stored in each fixture, then the correct intensity ofthe network is determined via the control computer, and the intensity setting is transmittedto the fixtures. The lights themselves preform the linear adjustment to then correct theirintensity.

The techniques discussed in this work can be generalized to the control of traditionallighting technologies. One goal of designing the polychromatic system was to develop a lightsource capable of best-in-class photometric and color related properties. It was also theorizedthat existing methods of measuring and sensing color lead to decreased performance of thesesystems. The development of a custom light source enabled a quantitative comparison ofthe methods and techniques proposed for control. Additionally, it extends the work on theoptimal control of lighting in [64, 49] by considering spectrally tunable sources capable ofdynamic energy consumption. These extra layers of difficulty prompted the developmentand testing of new models and techniques to control the network. Yet, LED lighting is notwidespread. For example, in MIT building E14, a Lutron Grafik Eye controls the dimmableoverhead fluorescent fixtures. Future work will focus on incorporating more diverse sourcesinto the control of the system.

The benefits of using an adaptive approach to selecting the wavelengths of the source werementioned, yet this extra degree of control presents some major concerns from a usabilitystandpoint. At present, a single button press on the control node instructs the systemto perform a calibration step. The user is then able to set the desired illuminance leveland color temperature by adjusting a slider on the control node. These operations aremore complicated than turning on or off a switch. Certainly, adding a fourth operation –the ability to switch between efficient and high quality light is overkill. So how can this besimplified? First, an experiment to derive the users’ lighting preferences must be performed.In these experiments, it can be tested which light sources the user prefers under differentsituations. Ultimately, one conclusion may be that, although the dynamic efficacy offersadditional energy savings, the quality of the light is not preferred by any users. Long termtesting incorporating a large number of LED and other fixtures are required to report onthe true energy savings employed by these techniques. In the short term, testing is plannedto occur with the iW Blast Powercore from Color Kinetics.

7.2 Future Work

The work presented here leads to the “event horizon”, so to speak, of researching the broaderimplications of sensor-enabled lighting networks. For every question answered here, severalmore exist. Missing from this discussion are the effects of reflected light (i.e., headmountedgaze tracking) on the control and performance of the network. Yet using a sensor boarddesigned here, the preliminary effects of this type of control can be quickly studied. Also

83

7.2. Future Work

missing are the quantitative results of the control of at least two fixtures and a full study ofthe actual energy saved by adjusting the intensity in response to changing ambient light. Toget to this point, several illuminance studies of building E14 (concentrated primarily on the5th floor) are required. These studies will provide actual illuminance data of the area, whichwe will use to build models of the expected energy savings. These will be performed in theshort term and once complete, can be used to contrast the measured results of several tests inwhich the intensity is regulated according to ambient light. Going one step further, we planto measure the energy savings when the user is able to adjust the color and illuminance aswell. Also planned for the near-future is the redesign of the sensor node to make it compactand more energy efficient. Broader research ambitions include incorporating gesture-basedcontrol of the sensor node (i.e., create a light surface of this intensity and color quality here,similar to Bolt’s classic “Put That There”) [13]. In this manner a zone of lighting can bespecified, for example, by moving the sensor node around the desired area (an accelerometerin the sensor node can detect motion) or video could enable gesture recognition and tracking.

Looking towards the future, the next phase of work will aim to utilize more complicatedsensors and activity-recognition algorithms to better infer user context and light their en-vironment accordingly. Room-mounted and wearable cameras, for example, will be used totrack users and determine their zone of activity, lighting the area accordingly. Hands will betracked, and gestural control (e.g., pointing, sweeping, framing, etc.) will be used to directthe room lighting when it is necessary to go beyond the assumptions made by the contextengine [39]. Users can be tracked to within several meters by the wireless signal from theirwearable system – correlating motion detected by their wearable accelerometers and localcameras can precisely identify and locate them [70, 69]. In the final phase of the work,the system will be installed into a suite of working offices to examine the interaction andcontrol of this intelligent lighting system with other utilities (e.g., closed loop heating andair conditioning [22]) to create environments that automatically meet users’ utility goalsand comfort while minimizing energy. Also of interest are the persuasive interfaces to theseenvironments, through which users will see the system working, explore their energy con-sumption, compare their energy usage with their past activity and to that of other subjects,and adjust the assumptions that the system makes about them. This phase of activity willalso integrate with a ubiquitous interactive display system that we have installed throughoutour building that identifies users when they approach and facilitates a variety of contact andnon-contact interaction. Users of our system can employ these “portals” to invoke and in-teract with their energy-related information anywhere in the building. Although researchershave been exploring aspects of interactive lighting for several years, the agile control andrich sensor data afforded by these systems will inspire new means of interfacing with andcontrolling fine-grained utility infrastructure to maximize function while minimizing energy,illustrating a key application that will drive cyber-physical systems in the next decade.

84

Appendix A

Expanded Results

The Figures A.2, A.1, A.3, and A.4 contain the predicted and measured spectra used todescribe the experiments in Chapter 5.

85

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Line

ar R

atio

s, 5

00 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(a)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Line

ar R

atio

s, 1

000

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(b)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Line

ar R

atio

s, 1

500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(c)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Ove

rdet

erm

ined

, 500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(d)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Ove

rdet

erm

ined

, 100

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(e)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Ove

rdet

erm

ined

, 150

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(f)

Figu

reA.1:Pr

edictedspectral

respon

seforbo

thlin

earan

doverde

term

ined

metho

dsforthreediffe

rent

luxsettings

and11

color

tempe

ratures.

86

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Mea

sure

d Li

near

, 500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(a)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Mea

sure

d Li

near

, 100

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(b)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Mea

sure

d Li

near

, 150

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(c)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Mea

sure

d O

verd

eter

min

ed, 5

00 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(d)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Mea

sure

d O

verd

eter

min

ed, 1

000

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(e)

400

500

600

700

800

012345678910

Wav

elen

gth

(nm

)


)

Mea

sure

d O

verd

eter

min

ed, 1

500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(f)

Figu

reA.2:Measuredspectral

respon

seforbo

thlin

earan

doverde

term

ined

metho

dsforthreediffe

rent

luxsettings

and11

color

tempe

ratures.

87

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Max

imiz

ed E

ffica

cy, 5

00 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(a)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Max

imiz

ed E

ffica

cy, 1

000

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(b)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Max

imiz

ed E

ffica

cy, 1

500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(c)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Max

imiz

ed C

olor

Ren

derin

g In

dex,

500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(d)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Max

imiz

ed C

olor

Ren

derin

g In

dex,

100

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(e)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Max

imiz

ed C

olor

Ren

derin

g In

dex,

150

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(f)

Figu

reA.3:Pr

edictedspectral

respon

seusingtheMADSalgo

rithm

form

axim

izingeffi

cacy

orcolorr

enderin

gindexfort

hree

diffe

rent

luxsettings

and11

colortempe

ratures.

88

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Mea

sure

d M

axim

ized

Effi

cacy

, 500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(a)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Mea

sure

d M

axim

ized

Effi

cacy

, 100

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(b)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Mea

sure

d M

axim

ized

Effi

cacy

, 150

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(c)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Mea

sure

d C

olor

Ren

derin

g In

dex,

500

lux

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(d)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Mea

sure

d C

olor

Ren

derin

g In

dex,

100

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(e)

400

500

600

700

800

012345678

Wav

elen

gth

(nm

)


)

Mea

sure

d C

olor

Ren

derin

g In

dex,

150

0 lu

x

28

00K

3000

K35

00K

4000

K45

00K

5000

K55

00K

6000

K65

00K

8000

K10

000K

(f)

Figu

reA.4:Measuredspectral

respon

seusingtheMADSalgo

rithm

form

axim

izingeffi

cacy

orcolorr

enderin

gindexfort

hree

diffe

rent

luxsettings

and11

colortempe

ratures.

Insubfi

gure

b,thespectrom

eter

saturatedfortestpo

ints

8000

Kan

d10

000K.

89

Appendix B

Hardware Design

B.1 LED Controller Hardware Schematics

[follows on next page]

90

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

Title

NumberRevision

SizeBDate:4/29/2010

Sheet ofFile:

C:\Documents and Settings\..\Sheet1.SchDoc

Drawn By:

GPIO29/SCITXD

A/SCLA/TZ3

1

TRST2

XRS3

ADCINA6/A

IO64

ADCINA4/CO

MP2A

/AIO4

5

ADCINA7

6

ADCINA3

7

ADCINA1

8

ADCINA2/CO

MP1A

/AIO2

9

ADCINA0/VREFHI

10

VDDA

11

VSSA/VREFLO

12

ADCINB1

13

ADCINB2/COM

P1B/AIO

1014

ASCINB3

15

ADCINB4/COM

P2B/AIO

1216

ADCINB6/A

IO1417

ADCINB7

18

GPIO34/CO

MP2OUT

19

GPIO35/TD

I20

GPIO36/TM

S21

GPIO37/TD

O22

GPIO38/XCLKIN/TCK

23

GPIO18/SPICLK

A/SCITXDA/XCLKO

UT24

GPIO19/XCLKIN/SPISTEA/SCIRXD

A/ECAP125

GPIO17/SPISOM

IA/TZ326

GPIO16/SPISIM

OA/TZ2

27

GPIO1/EPW

M1B/CO

MP1O

UT28

GPIO0/EPW

M1A

29

TEST30

GPIO32/SD

AA/EPWM

SYNCI/A

DCSOCAO

31

VDD

32

VSS33

VREGENZ

34

VDDIO

35

GPIO33/SCLA/EPW

MSYNCO/A

DCSOCBO36

GPIO2/EPW

M2A

37

GPIO3/EPW

M2B/CO

MP2O

UT38

GPIO4/EPW

M3A

39

GPIO5/EPM

W3B/ECA

P140

GPIO6/EPW

M4A/EPW

MSYN

CI/EPWM

SYNCO

41

GPIO7/EPW

M4B/SCIRXD

A42

VDD

43

VSS44

X145

X246

GPIO12/TZ1/SCITXDA

47

GPIO28/SCIRXD

A/SDAA/TZ2

48

U1TMX320F28027

12

34

56

78

910

1112

1314

P1Header 7X2

CTMS

CTDI

CPDCTD

OCTCK

-RETCTCKCEM

U0CEM

U1CG

NDCG

NDCG

ND

CGND

CTRSTn

123456789101112

P2Header 12

123456789101112

P3Header 12

123456789101112

P4Header 12

123456789101112

P5Header 12

CTRSTn

CTDO

CTDI

CTMS

CTCK

CGND

C22.2uF

L1EMI BEA

D

L2EMI BEA

D

C12.2uF

C101.0uF

C91.0uF

C7.47uF C3.47uF

L4EMI BEA

DC111.0uF

CGND

CGND

CGND

CGND

CGND

CGND

CGND

C3V3INC3V3D

C3V3A

CRSTR70

CGND

TI JTAG

HEA

DER

LED A

RRAY PIN

OUT

MCU FA

NO

UT

CRX-SLAVE

CTX-SLAVE

CSDA

CSCL

CTAOSINPU

T

11

22

33

44

55

66

77

88

99

1010

1111

1212

1313

1414

1515

1616

1717

1818

1919

2020

2121

2222

2323

2424

2525

2626

2727

2828

2929

3030

3131

3232

3333

3434

3535

3636

3737

3838

3939

4040

4141

4242

4343

4444

4545

4646

4747

4848

4949

5050

H1N2550-6002-RB

CREDRTN

CRED

CAM

B

CREDCRED

CREDRTNCREDRTN

CAM

BCA

MB

CAM

BRTNCA

MBRTN

CAM

BRTN

CGRN

CGRN

CGRN

CGRNRTN

CGRNRTN

CGRNRTN

CBLUCBLU

CBLU

CBLURTNCBLURTNCBLURTN

TAOSOE

CS1CS0

CS3CS2

CSCL

CSDA

CSYNC

CINT

CFUNC

CINCTEM

P

CCYN

CCYN

CCYN

CCYNRTN

CCYNRTN

CCYNRTN

CTAOSINPU

T

CREDPWM

CAM

BERPWM

CGREEN

PWM

CBLUEPWM

CCYAN

PWM

CIRTXBRSTCG

PIO7

CO

NTR

OLLER

P1

CSYNC

CINT

CS3CS2

CGND

TAOSOE

R2TBD 0402

R3TBD 0402

CS0CS1

C3V3INC3V3IN

TAOS PU

LLUPSA

1

B2

Y3

GND

4

Y5

A/B6

G7

VCC8

U2SN74LVC2G157C24

100nFCG

ND

C3V3IN

CTEMP

CMSO

L3EMI BEA

D

C3V3IN

CADCLU

XL

CAGND

C12TBD

CGND

R14TBD

0402

CRSTS1B3S-1000

CGND

BTN RST

CFUNC

R1TBD 0402

CGND

TMP05 SETTING

S

CMSO

OPEN

OPEN

CADCLU

XH

CTEMP

R8TBD 0603

D1HSMH-C190

CGND

CSTATUS

1.8 0603 RED

CSTATUS

Float: Cont / GND: O

ne-Shot

CSDA

CSCL

R510K

C3V3IN

R610K

C3V3IN

R410K

C3V3IN

CIRTXBRST

D3TSAL6200

CGND R10TBD

0805

IR TXQ2NDS355AN

C3V3IN

Q1NDS355AN

C3V3IN

1243

NO

S2218-2LPST

R20TBD

0402R17TBD

04023.3K

820R12TBD

0402R11TBD

0402

CGND

CGND

C3V3INC3V3IN

CGPIO

34

CGPIO

34CTD

O

MCU BO

OT SETTIN

GS

LED

R13TBD

0402

CGND

OPTIO

N

R15TBD

0402

CGND

OPTIO

N

MUX_TA

OS_TEMP

CTX-SLAVE

CTRSTnCRST

ADCINA6

ADCINA4

ADCINA0

ADCINA7

ADCINB1

ADCINB3

ADCINB7

ADCINA6

ADCINA4

ADCINA7

CADCLU

XHCA

DCLUXL

CADCLU

XNOFLT

ADCINA0

C3V3ACAGND

ADCINB1

CSYNC

CSTATUS

ADCINB3

CMSO

ADCINB7

CGPIO

34

CTDO

CTDI

CTMS

CTCKCS3

MUX_TA

OS_TEMP

CS2TAO

SOE

CAM

BERPWM

CREDPWM

CSDA

CGND

C3V3D

CSCL

R9TBD 0603

D2HSMH-C190

CGND

C3V3IN

CGREEN

PWM

CBLUEPWM

CCYAN

PWM

CIRTXBRSTCG

PIO7

CGND

CTEMP

CRX-SLAVE

COLO

R / TEMP M

UX

CGND C810nF CG

ND C410nFC52.2pF

CGND

C61.0uF

CGND

CADCLU

XNOFLT OPEN

OPEN

OPEN

OPEN

GPIO3

C3V3IN

CINT

PIC101 PIC102COC1

PIC201 PIC202COC2

PIC301 PIC302COC3

PIC401 PIC402COC4

PIC501 PIC502COC5

PIC601 PIC602COC6PIC701 PIC702

COC7PIC801 PIC802COC8PIC901 PIC902

COC9PIC1001 PIC1002

COC10

PIC1101 PIC1102COC11


PIC2401PIC2402

COC24

PID101PID102COD1

PID201PID202COD2

PID301 PID302COD3

PIH101

PIH102

PIH103

PIH104

PIH105PIH106

PIH107PIH108

PIH109

PIH1010

PIH1011PIH1012

PIH1013PIH1014

PIH1015PIH1016

PIH1017PIH1018

PIH1019PIH1020

PIH1021PIH1022

PIH1023PIH1024

PIH1025PIH1026

PIH1027PIH1028

PIH1029PIH1030

PIH1031PIH1032

PIH1033PIH1034

PIH1035PIH1036

PIH1037PIH1038

PIH1039PIH1040

PIH1041PIH1042

PIH1043PIH1044

PIH1045PIH1046

PIH1047PIH1048

PIH1049PIH1050

COH1

PIL101PIL102

COL1

PIL201PIL202

COL2

PIL301PIL302

COL3

PIL401PIL402

COL4

PIP101

PIP102

PIP103

PIP104

PIP105

PIP106

PIP107

PIP108

PIP109

PIP1010

PIP1011PIP1012

PIP1013

PIP1014

COP1

PIP201

PIP202

PIP203

PIP204

PIP205

PIP206

PIP207

PIP208

PIP209

PIP2010

PIP2011

PIP2012 COP2

PIP301

PIP302

PIP303

PIP304

PIP305

PIP306

PIP307

PIP308

PIP309

PIP3010

PIP3011

PIP3012 COP3

PIP401

PIP402

PIP403

PIP404

PIP405

PIP406

PIP407

PIP408

PIP409

PIP4010

PIP4011

PIP4012 COP4

PIP501

PIP502

PIP503

PIP504

PIP505

PIP506

PIP507

PIP508

PIP509

PIP5010

PIP5011

PIP5012 COP5

PIQ101

PIQ102 PIQ103COQ1PIQ201

PIQ202 PIQ203COQ2

PIR101PIR102 COR1

PIR201 PIR202COR2

PIR301 PIR302COR3

PIR401 PIR402COR4

PIR501 PIR502COR5

PIR601 PIR602COR6

PIR701 PIR702COR7

PIR801PIR802 COR8PIR901PIR902 COR9

PIR1001 PIR1002COR10








PIS101

PIS102

PIS103

PIS104

COS1PIS201

PIS202

PIS203

PIS204

COS2

PIU101

PIU102

PIU103

PIU104

PIU105

PIU106

PIU107

PIU108

PIU109

PIU1010

PIU1011

PIU1012

PIU1013

PIU1014

PIU1015

PIU1016

PIU1017

PIU1018

PIU1019

PIU1020

PIU1021

PIU1022

PIU1023

PIU1024

PIU1025

PIU1026

PIU1027

PIU1028

PIU1029

PIU1030

PIU1031

PIU1032

PIU1033

PIU1034

PIU1035

PIU1036

PIU1037

PIU1038

PIU1039

PIU1040

PIU1041

PIU1042

PIU1043

PIU1044

PIU1045

PIU1046

PIU1047

PIU1048

COU1

PIU201

PIU202

PIU203

PIU204

PIU205

PIU206

PIU207

PIU208 COU2

PIP2010

PIU1010NLADCINA0

PIP205

PIU105

NLADCINA4

PIP204

PIU104

NLADCINA6

PIP206

PIU106

NLADCINA7

PIP3012

PIU1013NLADCINB1

PIP3010

PIU1015

NLADCINB3

PIP307

PIU1018

NLADCINB7

PIC602PIC702

PIC802PIL202

PIP2011

PIU1011

PIC302PIC402

PIC502PIL102

PIP402

PIU1035

PIC102PIC202

PIC2401

PID101PID201

PID302

PIL101

PIL201

PIL302

PIR202PIR302

PIR402PIR502

PIR602

PIR1102PIR1202

PIR1402

PIU208

PIP209

PIU109

NLCADCLUXH

PIP208

PIU108

NLCADCLUXL

PIP207

PIU107

NLCADCLUXNOFLT

PIC1102PIL402

PIP2012

PIU1012

PIH107PIH108

PIH109

NLCAMB

PIP504

PIU1040

NLCAMBERPWM

PIH1010

PIH1011PIH1012

NLCAMBRTN

PIH1013PIH1014

PIH1015

NLCBLU

PIP408

PIU1029NLCBLUEPWM

PIH1016

PIH1017PIH1018

NLCBLURTN

PIP409

PIU1028NLCCYANPWM

PIH1025PIH1026

PIH1027

NLCCYNPIH1028

PIH1029PIH1030

NLCCYNRTN

PIP1013

NLCEMU0PIP1014

NLCEMU1

PIH1046

PIR102

NLCFUNC

PIC101PIC201

PIC301PIC401

PIC501

PIC601PIC701

PIC801

PIC901PIC1001

PIC1101

PIC1201

PIC2402

PIH1031

PIH1041

PIH1045

PIL401

PIP104

PIP108

PIP1010

PIP1012

PIP404

PIP508

PIQ102PIQ202

PIR101

PIR701

PIR902

PIR1301

PIR1501

PIR1701PIR2001

PIS102

PIU1033

PIU1044

PIU204

PIU207

NLCGND

PIP506

PIU1042

NLCGPIO7PIP306

PIR1702

PIS201

PIU1019

NLCGPIO34

PIP501

PIU1037

NLCGREENPWM

PIH1019PIH1020

PIH1021

NLCGRN

PIH1022

PIH1023PIH1024

NLCGRNRTN

PIH1043NLCIN

PIH1040

PIR401

PIU1038

NLCINT

PIP505

PIQ201

PIU1041

NLCIRTXBRST

PIP308

PIU1017

PIU206 NLCMSO

PIP105

NLCPD

PIH101

PIH102

PIH103

NLCRED

PIP503

PIU1039

NLCREDPWM

PIH104

PIH105PIH106

NLCREDRTN

PIC1202

PIP203

PIR1401PIS104

PIU103

NLCRST

PIP5012

PIU1048

NLCRX0SLAVE

PIH1034

PIR201

NLCS0PIH1033

PIR301

NLCS1

PIH1036

PIP4011

PIU1026

NLCS2PIH1035

PIP301

PIU1024

NLCS3

PIH1047

PIP401

PIR601

PIU1036

NLCSCL

PIH1039

PIP406

PIR501

PIU1031

NLCSDA

PIP3011

PIQ101

PIU1014NLCSTATUS

PIH1048

PIP309

PIU1016

NLCSYNC

PIH1037

PIU201

NLCTAOSINPUT

PIP1011

PIP302 PIU1023

NLCTCKPIP109

NLCTCK0RET PIP103

PIP305 PIU1020

NLCTDI

PIP107

PIP303

PIR1502

PIR2002

PIS202

PIU1022

NLCTDO

PIH1044

PIP5011

PIU1047

PIU202

NLCTEMP

PIP101

PIP304 PIU1021

NLCTMSPIP102

PIP202

PIR1302

PIU102

NLCTRSTn

PIP201

PIU101

NLCTX0SLAVE

PIP502

NLGPIO3

PIP4012

PIU1025PIU205

NLMUX0TAOS0TEMP

PIS103

PIS101

PIC902

PIU1032

PIC1002

PIU1043

PID102PIR801PID202PIR901

PID301PIR1002

PIH1038PIL301

PIH1042

PIH1049PIH1050

PIP106

PIP403

PIP405

PIP407

PIP507

PIP509

PIP5010

PIQ103 PIR802PIQ203 PIR1001

PIR702PIU1034

PIR1101PIS204

PIR1201PIS203

PIU1030

PIU1045

PIU1046

PIU203

PIH1032

PIP4010

PIU1027

NLTAOSOE

B.1. LED Controller Hardware Schematics

91

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

Title

NumberRevision

SizeBDate:4/29/2010

Sheet ofFile:


Drawn By:

CO

NTR

OLLER

P2

4 31

52

U3AD8605ARTZ

VDD

1

GND

2

NC3

EN4

NC5

IOUT6

U6ISL29009IROZ-T7

C19100nF

R24

TBD 0402

C15100nF

CAGND

LUX SEN

SOR LO

W RA

NG

E

LUX SEN

SOR H

IGH

RAN

GE

CAGND

4 31

52

U4AD8605ARTZ

C17

100nF

CAGND

CADCLU

XLR23

TBD 0402

C221nF

CAGND

L5EMI BEA

DC14.47uF

CAGND

CAGND

C3V3A

4 31

52

U5AD8605ARTZ

C16

.1uF

C20TBD

C21TBD

R21

TBD 0402

R22

TBD 0402

CAGND

C3V3OP

C3V3OP

CAGND

C3V3OP

4 31

52

U7AD8605ARTZ

VDD

1

GND

2

NC3

EN4

NC5

IOUT6

U10

ISL29009IROZ-T7

C29100nF

R28

TBD 0402

CAGND

CAGND

4 31

52

U8AD8605ARTZ

C27

100nF

CAGND

CADCLU

XHR27

TBD 0402

C321nF

CAGND

C23

100nF

CAGND

4 31

52

U9AD8605ARTZ

C30TBD

C31TBD

R25

TBD 0402

R26

TBD 0402

CAGND

C3V3OP

C3V3OP

CAGND

C3V3OP

C13

100nF

CAGND

C18TBD

0603

C28TBD

0603

CAGND

CAGND

CAGND

CAGND

CAGND

CAGND

CAGND

CAGND

C3V3A

C3V3A

OPTIO

N

OPTIO

N

VDD

1

GND

2

NC3

EN4

NC5

IOUT6

U12

ISL29009IROZ-T7

C35100nF

C34TBD

0603

CAGND

CAGND

CAGND

C3V3A

R30TBD

0402

CAGND

CAGND

4 31

52

U11AD

8605ARTZ

C33

100nF

CAGND

CADCLU

XNOFLT

R29

TBD 0402

C361nF

CAGND

C3V3OP

OPTIO

N

HI SPEED

LUX

C26

.1uF

PIC1301PIC1302

COC13



PIC1601PIC1602

COC16

PIC1701PIC1702

COC17



PIC2001PIC2002COC20

PIC2101PIC2102COC21


PIC2301PIC2302

COC23

PIC2601PIC2602

COC26

PIC2701PIC2702

COC27



PIC3001PIC3002COC30

PIC3101PIC3102COC31


PIC3301PIC3302

COC33




PIL501PIL502

COL5

PIR2101PIR2102 C

OR21

PIR2201PIR2202 C

OR22

PIR2301PIR2302

COR23

PIR2401PIR2402 COR24

PIR2501PIR2502 CO

R25

PIR2601PIR2602 CO

R26

PIR2701PIR2702

COR27


PIR2901PIR2902

COR29


PIU301

PIU302

PIU303

PIU304

PIU305COU3

PIU401

PIU402

PIU403

PIU404

PIU405COU4

PIU501

PIU502

PIU503

PIU504

PIU505COU5

PIU601

PIU602

PIU603

PIU604

PIU605

PIU606

COU6

PIU701

PIU702

PIU703

PIU704

PIU705COU7

PIU801

PIU802

PIU803

PIU804

PIU805COU8

PIU901

PIU902

PIU903

PIU904

PIU905COU9

PIU1001

PIU1002

PIU1003PIU1004

PIU1005

PIU1006

COU10

PIU1101

PIU1102

PIU1103

PIU1104

PIU1105COU11

PIU1201

PIU1202

PIU1203PIU1204

PIU1205

PIU1206

COU12 PIC1402

PIC1802PIC1902

PIC2802PIC2902

PIC3402PIC3502

PIL501

PIU601

PIU1001

PIU1201

PIC1301

PIC1502

PIC1601

PIC1701

PIC2301

PIC2601

PIC2701

PIC3301

PIL502

PIU303

PIU305PIU405

PIU505

PIU703

PIU705PIU805

PIU905

PIU1105

PIC3202PIR2702

NLCADCLUXH

PIC2202PIR2302

NLCADCLUXL

PIC3602PIR2902

NLCADCLUXNOFLT

PIC1302PIC1401

PIC1501

PIC1602

PIC1702

PIC1801PIC1901

PIC2102PIC2201

PIC2302

PIC2602

PIC2702

PIC2801PIC2901

PIC3102PIC3201

PIC3302

PIC3401PIC3501

PIC3601PIR3002

PIU302PIU402

PIU502PIU602

PIU604

PIU702PIU802

PIU902PIU1002

PIU1004

PIU1102PIU1202

PIU1204

PIC2001

PIR2101PIR2202

PIC2002

PIU403

PIU501

PIU504

PIC2101

PIR2201PIU503

PIC3001

PIR2501PIR2602

PIC3002

PIU803

PIU901

PIU904

PIC3101

PIR2601PIU903

PIR2102

PIR2401

PIU301

PIR2301PIU401

PIU404

PIR2402

PIU304

PIU606

PIR2502

PIR2801

PIU701

PIR2701PIU801

PIU804

PIR2802

PIU704

PIU1006

PIR2901PIU1101

PIU1104PIR3001

PIU1103

PIU1206

PIU603

PIU605

PIU1003

PIU1005

PIU1203

PIU1205


92

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

Title

NumberRevision

SizeBDate:4/29/2010

Sheet ofFile:


Drawn By:

TXD1

DTR2

RTS3

VCCIO4

RXD5

RI6

GND

7

NC8

DSR9

DCD10

CTS11

CBUS4

12

CBUS2

13

CBUS3

14

USBDP15

USBDM16

3V3OUT

17

GND

18

RST19

VCC20

GND

21

CBUS1

22CBU

S023

NC24

AGND

25

TEST26

OSCI27

OSCO28

U13

FT232RL

1234

SH5

J161729-1010BLFC37

10nF

CGND

L7 EMI BEA

D

C25100nF

CGND

CGND

CGND

R35

TBD 0402

CVCCIO

CVCCIO

R320

R340

CRX-SLAVE

CTX-SLAVE

D10HSM

G-C190

2.2V 0603 GRN

D11HSM

H-C190

1.8 0603 RED

R18TBD

0603R19TBD

0603

CVCCIOCVCCIO

USB / SERIAL

11

22

33

44

55

66

77

88

99

1010

1111

1212

1313

1414

1515

1616

1717

1818

1919

2020

P691596-120LF

11

22

33

44

55

66

77

88

99

1010

1111

1212

1313

1414

1515

1616

1717

1818

1919

2020

P791596-120LF

ZIGBEE IN

TERFACE

CRX-SLAVE

CTX-SLAVE

CDIO

12CRSTCRSSICD

IO11

CDTR*

CDIO

4CCTS*CSLEEP*CG

NDCA

SSOCCRTS*CD

IO3

CDIO

2CD

IO1

CCOM

MIS

CGND

C3V3INR31

0

R330

R36TBD

0603

D4HSMH-C190

CGND

CASSOC

1.8 0603 RED

CO

NTR

OLLER

P3

C3830pF

C3930pF

CGND

CGND

S3B3S-1000CCO

MM

IS

CGND

COM

MISSIO

N / A

SSOCIA

TEPIC2501 PIC2502

COC25

PIC3701PIC3702 C

OC37


COC39

PID401PID402COD4

PID1001PID1002COD10

PID1101PID1102COD11

PIJ101

PIJ102

PIJ103

PIJ104

PIJ105

COJ1PIL701

PIL702

COL7PIP601

PIP602

PIP603

PIP604

PIP605

PIP606

PIP607

PIP608

PIP609

PIP6010

PIP6011

PIP6012

PIP6013

PIP6014

PIP6015PIP6016

PIP6017PIP6018

PIP6019

PIP6020 COP6

PIP701

PIP702

PIP703

PIP704

PIP705

PIP706

PIP707

PIP708

PIP709

PIP7010

PIP7011

PIP7012

PIP7013

PIP7014

PIP7015PIP7016

PIP7017PIP7018

PIP7019

PIP7020 COP7





PIR3301PIR3302 CO

R33




PIS301

PIS302

PIS303

PIS304

COS3

PIU1301

PIU1302

PIU1303

PIU1304

PIU1305

PIU1306

PIU1307

PIU1308

PIU1309

PIU13010

PIU13011

PIU13012

PIU13013

PIU13014

PIU13015

PIU13016

PIU13017

PIU13018

PIU13019

PIU13020

PIU13021

PIU13022

PIU13023

PIU13024

PIU13025

PIU13026

PIU13027

PIU13028 COU13

PIP601

PIP7012

PIR3601

NLCASSOC

PIP702

PIS304

NLCCOMMISPIP7018NLCCTS0

PIP704

NLCDIO1

PIP706

NLCDIO2

PIP708

NLCDIO3

PIP7020

NLCDIO4

PIP6013 NLCDIO11

PIP607 NLCDIO12

PIP6017 NLCDTR0

PIC2501

PIC3702

PIC3801PIC3901

PID402

PIJ104

PIP6019

PIP7014

PIS302

PIU1307

PIU13018

PIU13021

PIU13025

PIU13026

NLCGNDPIP6011 NLCRSSI

PIP609 NLCRST

PIP7010

NLCRTS0

PIR3101

PIR3201

NLCRX0SLAVE

PIP7016NLCSLEEP0

PIR3301

PIR3401

NLCTX0SLAVE

PIC2502

PID1001PID1101

PIR3501

PIU1304

PIU13017

NLCVCCIO

PIS303

PIS301

PIC3701

PIJ101

PIL701PIC3802

PIJ102PIU13016

PIC3902

PIJ103

PIU13015

PID401 PIR3602


PIJ105

PIL702PIU13020

PIP602

PIP603

PIR3102

PIP604

PIP605

PIR3302PIP606

PIP608

PIP6010

PIP6012

PIP6014

PIP6015PIP6016

PIP6018

PIP6020

PIP701

PIP703

PIP705

PIP707

PIP709

PIP7011

PIP7013

PIP7015

PIP7017

PIP7019

PIR1801

PIU13023

PIR1901

PIU13022

PIR3202

PIU1301

PIR3402PIU1305

PIR3502PIU13014

PIU1302

PIU1303

PIU1306

PIU1308

PIU1309

PIU13010

PIU13011

PIU13012

PIU13013

PIU13019

PIU13024

PIU13027

PIU13028


93

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

Title

NumberRevision

SizeBDate:4/29/2010

Sheet ofFile:


Drawn By:

VIN-1

VIN+3

5VREF4

CTL5

LED-

6

LED+

7U14

7021-D-E-500

Q3MM

BT3906

R37

TBD 0603

R40

TBD 0603

VIN-1

VIN+3

5VREF4

CTL5

LED-

6

LED+

7U15

7021-D-E-500

Q4MM

BT3906

R38

TBD 0603

R41

TBD 0603

VIN-1

VIN+3

5VREF4

CTL5

LED-

6

LED+

7U19

7021-D-E-500

Q7MM

BT3906

R44

TBD 0603

R46

TBD 0603

VIN-1

VIN+3

5VREF4

CTL5

LED-

6

LED+

7U18

7021-D-E-500

Q6MM

BT3906

R43

TBD 0603

R45

TBD 0603

VIN-1

VIN+3

5VREF4

CTL5

LED-

6

LED+

7U16

7021-D-E-500

Q5MM

BT3906

R39

TBD 0603

R42

TBD 0603

VLED

VLEDVLED VLED

VLED

LEDRTN

LEDRTN

LEDRTN

LEDRTN

LEDRTN

CRED

CREDRTN

CAM

B

CAM

BRTN

CBLU

CBLURTN

CGRN

CGRNRTN

CCYN

CCYNRTN

CREDPWM

CAM

BERPWM

CGREEN

PWM

CBLUEPWM

CCYAN

PWM

CO

NTR

OLLER

P4

VLEDSUPPLY

LEDRTN C48

22UF 25V X5R 1210

L8TBD

C4922UF 25V X5R 1210

LEDRTN

VLED

VIN1

2 VOUT

3

GND

U17R-78B5.0-1.0

24VREG

C4022UF 25V X5R 1210

CGND

C4122UF 25V X5R 1210

C4222UF 25V X5R 1210

CGND

CGND

5VREF

CGND

EN1

IN2

GND

3

OUT

4

NR/FB5

GND

6

U20

TPS79633

CGND

5VREFC432.2uF 6.3V X7R

C442.2uF 6.3V X7R

C452.2uF 6.3V X7R

CGND

C4610nF

CGND

C3V3IN

R48

0LED

RTNCG

ND

GN

D N

ET TIE

R47

0

CTNT M

ES

1 23J22.1mm ID x 5mm OD

LEDRTN

VLEDSUPPLY

6 - 30V IN

/ 5V O

UT BUCK

5V / 3.3V

LINEA

R

5 CHA

NN

EL LED D

RIVER

DESK

TOP PO

WER EN

TRY

F133V / 1.85A@25C

24VREG

t° RT1

FILT TBD

OPTIO

N

C4747uF 35V




PIC4301PIC4302COC43




PIC4701PIC4702COC47



PIF101

PIF102

COF1PIJ201

PIJ202

PIJ203

COJ2

PIL801PIL802

COL8

PIQ301 PIQ302PIQ303 COQ3





PIR3701PIR3702

COR37PIR3801

PIR3802COR38

PIR3901PIR3902

COR39

PIR4001PIR4002

COR40PIR4101

PIR4102COR41

PIR4201PIR4202

COR42

PIR4301PIR4302

COR43

PIR4401PIR4402

COR44

PIR4501PIR4502

COR45

PIR4601PIR4602

COR46

PIR4701PIR4702

COR47

PIR4801PIR4802

COR48

PIRT101PIRT102

CORT1

PIU1401

PIU1403

PIU1404

PIU1405

PIU1406

PIU1407

COU14

PIU1501

PIU1503

PIU1504

PIU1505

PIU1506

PIU1507

COU15

PIU1601

PIU1603

PIU1604

PIU1605

PIU1606

PIU1607

COU16

PIU1701

PIU1702

PIU1703

COU17

PIU1801

PIU1803

PIU1804

PIU1805

PIU1806

PIU1807

COU18

PIU1901

PIU1903

PIU1904

PIU1905PIU1906

PIU1907

COU19

PIU2001

PIU2002

PIU2003

PIU2004

PIU2005

PIU2006 COU20

PIC4102PIC4202

PIC4301

PIU1703

PIU2001

PIU2002

PIC4701PIC4802

PIF102

PIL801

PIRT101

PIR4702

PIU1907

NLCAMB

PIR4601NLCAMBERPWM

PIU1906NLCAMBRTN

PIU1507

NLCBLU

PIR4101NLCBLUEPWM

PIU1506

NLCBLURTN

PIR4201NLCCYANPWM

PIU1607NLCCYN

PIU1606

NLCCYNRTN

PIC4001PIC4101

PIC4201

PIC4302PIC4401

PIC4501PIC4601

PIR4802

PIU1702

PIU2003

PIU2006

PIR4501NLCGREENPWM

PIU1807NLCGRN

PIU1806

NLCGRNRTN

PIU1407

NLCRED

PIR4001NLCREDPWM

PIU1406

NLCREDRTN

PIC4702PIC4801

PIC4901

PIJ202

PIJ203

PIR4801

PIU1401PIU1501

PIU1601

PIU1801

PIU1901

PIC4002

PIRT102PIU1701

PIC4402PIC4502

PIR4701PIU2004

PIC4602PIU2005

PIQ301

PIR3701

PIR4002

PIQ302

PIR3702PIU1404

PIQ303

PIU1405

PIQ401

PIR3801

PIR4102

PIQ402

PIR3802PIU1504

PIQ403

PIU1505

PIQ501

PIR3901

PIR4202

PIQ502

PIR3902PIU1604

PIQ503

PIU1605

PIQ601

PIR4301

PIR4502

PIQ602

PIR4302PIU1804

PIQ603

PIU1805

PIQ701

PIR4401

PIR4602

PIQ702

PIR4402PIU1904

PIQ703

PIU1905

PIC4902

PIL802

PIU1403

PIU1503

PIU1603

PIU1803

PIU1903

PIF101

PIJ201


94

B.2. LED Controller PCB

B.2 LED Controller PCB


95

PAC101PAC102

COC1PAC201PAC202

COC2PAC301

PAC302COC3

PAC401PAC402

COC4PAC501

PAC502COC5

PAC601PAC602COC6

PAC701PAC702COC7

PAC801PAC802COC8

PAC901PAC902COC9

PAC1001

PAC1002

COC10

PAC1101PAC1102

COC11

PAC1201PAC1202

COC12

PAC1301PAC1302

COC13

PAC1402

PAC1401COC14PAC1501PAC1502

COC15

PAC1601PAC1602

COC16PAC1701PAC1702

COC17

PAC1802PAC1801

COC18

PAC1901PAC1902 COC19

PAC2001PAC2002

COC20PAC2101

PAC2102COC21

PAC2201PAC2202

COC22

PAC2301PAC2302

COC23

PAC2401PAC2402

COC24

PAC2501PAC2502

COC25

PAC2601

PAC2602COC26PAC2701

PAC2702COC27

PAC2802PAC2801

COC28

PAC2901PAC2902

COC29

PAC3001PAC3002COC30

PAC3101PAC3102

COC31

PAC3201PAC3202

COC32

PAC3301 PAC3302COC33

PAC3402PAC3401COC34

PAC3501PAC3502

COC35

PAC3601PAC3602

COC36

PAC3701PAC3702

COC37PAC3801PAC3802

COC38PAC3901

PAC3902COC39

PAC4001

PAC4002

COC40

PAC4102

PAC4101

COC41

PAC4202

PAC4201

COC42

PAC4301PAC4302

COC43

PAC4401PAC4402

COC44

PAC4501PAC4502

COC45

PAC4601PAC4602

COC46

PAC4701

PAC4702

COC47PAC4801

PAC4802

COC48PAC4901

PAC4902

COC49

PAD101

PAD102

COD1

PAD201

PAD202

COD2

PAD302

PAD301

COD3

PAD401

PAD402

COD4

PAD1001PAD1002

COD10

PAD1101PAD1102

COD11PAF101

PAF102

COF1PAH101

PAH103

PAH105PAH107PAH109PAH1011

PAH1013

PAH1015

PAH1017PAH1019PAH1021

PAH1023

PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049 PAH102

PAH104


PAH1014

PAH1016


PAH1024

PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAJ105PAJ101PAJ102

PAJ103PAJ104

COJ1PAJ203

PAJ201PAJ202

COJ2

PAL102PAL101

COL1

PAL202PAL201

COL2

PAL302PAL301COL3

PAL402

PAL401COL4

PAL502PAL501

COL5

PAL702PAL701

COL7

PAL801PAL802 COL8

PAP1014

PAP1013

PAP1012

PAP1011

PAP1010

PAP109

PAP108

PAP107

PAP106

PAP105

PAP104

PAP103

PAP102

PAP101

COP1

PAP2012

PAP2011

PAP2010

PAP209

PAP208

PAP207

PAP206

PAP205

PAP204

PAP203

PAP202

PAP201

COP2

PAP3012PAP3011

PAP3010PAP309

PAP308PAP307

PAP306PAP305

PAP304PAP303

PAP302PAP301

COP3PAP4012

PAP4011

PAP4010

PAP409

PAP408

PAP407

PAP406

PAP405

PAP404

PAP403

PAP402

PAP401

COP4PAP5012

PAP5011PAP5010

PAP509PAP508

PAP507PAP506

PAP505PAP504

PAP503PAP502

PAP501

COP5

PAP60U1

PAP6020

PAP6018

PAP6016

PAP6014

PAP6012

PAP6010

PAP608

PAP606

PAP604

PAP602

PAP6019

PAP6017

PAP6015

PAP6013

PAP6011

PAP609

PAP607

PAP605

PAP603

PAP601

COP6

PAP70U1

PAP7020

PAP7018

PAP7016

PAP7014

PAP7012

PAP7010

PAP708

PAP706

PAP704

PAP702

PAP7019

PAP7017

PAP7015

PAP7013

PAP7011

PAP709

PAP707

PAP705

PAP703

PAP701

COP7

PAQ101PAQ102PAQ103

COQ1

PAQ203PAQ202

PAQ201

COQ2

PAQ301PAQ302

PAQ303

COQ3

PAQ403PAQ402

PAQ401

COQ4

PAQ501

PAQ502PAQ503COQ5

PAQ601PAQ602

PAQ603

COQ6

PAQ701PAQ702

PAQ703

COQ7

PAR102PAR101

COR1

PAR202PAR201

COR2

PAR302PAR301

COR3

PAR402PAR401COR4

PAR502PAR501

COR5

PAR602PAR601

COR6

PAR702PAR701

COR7

PAR802

PAR801

COR8

PAR902

PAR901

COR9

PAR1001PAR1002

COR10

PAR1102PAR1101

COR11PAR1202

PAR1201

COR12

PAR1302PAR1301

COR13

PAR1402PAR1401COR14

PAR1502PAR1501

COR15

PAR1702PAR1701

COR17

PAR1802PAR1801

COR18PAR1902PAR1901COR19

PAR2002PAR2001COR20

PAR2102

PAR2101COR21

PAR2202PAR2201

COR22

PAR2302PAR2301

COR23

PAR2401

PAR2402

COR24

PAR2502PAR2501

COR25

PAR2602PAR2601COR26PAR2702

PAR2701COR27

PAR2802PAR2801

COR28

PAR2902

PAR2901COR29 PAR3002

PAR3001COR30

PAR3101PAR3102

COR31

PAR3202PAR3201

COR32

PAR3301PAR3302

COR33

PAR3402PAR3401

COR34PAR3502PAR3501

COR35

PAR3602

PAR3601

COR36

PAR3702PAR3701COR37

PAR3802

PAR3801COR38PAR3902

PAR3901COR39

PAR4002

PAR4001COR40

PAR4102PAR4101 COR41PAR4202PAR4201

COR42

PAR4302PAR4301

COR43PAR4402

PAR4401

COR44

PAR4502

PAR4501COR45

PAR4602

PAR4601COR46

PAR4702

PAR4701COR47

PAR4802PAR4801

COR48

PART102

PART101

CORT1

PAS101

PAS103PAS104

PAS102 COS1

PAS203PAS202

PAS204PAS201

COS2

PAS301

PAS303PAS304

PAS302 COS3

PAU101PAU102

PAU103PAU104

PAU105

PAU106PAU107

PAU108PAU109

PAU1010PAU1011

PAU1012PAU1013PAU1014PAU1015PAU1016PAU1017PAU1018PAU1019PAU1020PAU1021PAU1022PAU1023PAU1024

PAU1025

PAU1026

PAU1027PAU1028

PAU1029PAU1030

PAU1031PAU1032

PAU1033

PAU1034PAU1035

PAU1036

PAU1037PAU1038

PAU1039PAU1040PAU1041

PAU1042PAU1043

PAU1044PAU1045PAU1046PAU1047

PAU1048COU1



COU2

PAU305PAU304PAU303

PAU302PAU301

COU3

PAU405PAU404PAU403

PAU402PAU401

COU4

PAU505PAU504PAU503

PAU502PAU501

COU5

PAU601

PAU606

PAU602PAU603

PAU605PAU604 PAU60HSCOU6

PAU705PAU704PAU703

PAU702PAU701

COU7

PAU805PAU804PAU803

PAU802PAU801

COU8

PAU905PAU904PAU903

PAU902PAU901

COU9

PAU1001

PAU1006

PAU1002PAU1003

PAU1005PAU1004

PAU100HSCOU10


PAU1102PAU1101

COU11

PAU1201


PAU1205PAU1204PAU120HS COU12

PAU1301PAU1302PAU1303PAU1304PAU1305PAU1306PAU1307PAU1308PAU1309PAU13010PAU13011PAU13012PAU13013PAU13014


COU13

PAU1407PAU1406PAU1405PAU1404PAU1403PAU1401

COU14

PAU1501PAU1503

PAU1504PAU1505

PAU1506PAU1507

COU15

PAU1601PAU1603

PAU1604PAU1605

PAU1606PAU1607

COU16 PAU1701PAU1702PAU1703

COU17

PAU1801PAU1803

PAU1804PAU1805

PAU1806PAU1807

COU18

PAU1901PAU1903

PAU1904PAU1905

PAU1906PAU1907

COU19

PAU2006

PAU2005PAU2004PAU2003PAU2002PAU2001COU20

PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAC4802

PAF102

PAL801

PART101

PAP2010

PAU1010

PAP205

PAU105

PAP204

PAU104

PAP206PAU106

PAP3012

PAU1013

PAP3010

PAU1015PAP307 PAU1018PAC602

PAC702PAC802

PAC1402

PAC1802

PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501

PAP2011

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102PAP402

PAU1035

PAC102PAC202

PAC2401

PAD101PAD201

PAD302

PAL101

PAL201

PAL302

PAP601

PAR202

PAR302

PAR402PAR502

PAR602PAR1102

PAR1202

PAR1402

PAR4702

PAU208

PAC1301

PAC1502

PAC1601

PAC1701

PAC2301

PAC2601

PAC2701PAC3301

PAL502

PAU303

PAU305

PAU405 PAU505

PAU703PAU705

PAU805 PAU905

PAU1105

PAC3202PAP209

PAR2702

PAU109

PAC2202

PAP208PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602

PAC1702

PAC1801PAC1901PAC2102

PAC2201

PAC2302

PAC2602

PAC2702

PAC2801

PAC2901

PAC3102

PAC3201

PAC3302

PAC3401

PAC3501

PAC3601

PAL402

PAP2012

PAR3002

PAU1012

PAU302

PAU402 PAU502 PAU602

PAU604

PAU702

PAU802 PAU902 PAU1002 PAU1004

PAU1102 PAU1202 PAU1204

PAH107PAH108PAH109

PAU1907

PAP504

PAR4601

PAU1040


PAU1906

PAP7012

PAR3601

PAH1013PAH1014

PAH1015

PAU1507PAP408

PAR4101PAU1029

PAH1016

PAH1017PAH1018

PAU1506

PAP702

PAS304

PAP7018

PAP409PAR4201

PAU1028

PAH1025PAH1026

PAH1027PAU1607

PAH1028

PAH1029PAH1030PAU1606

PAP704

PAP706

PAP708

PAP7020

PAP6013

PAP607

PAP6017

PAP1013

PAP1014

PAH1046PAR102

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAD402

PAH1031

PAH1041

PAH1045

PAJ104PAJ105

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6019

PAP7014

PAQ102

PAQ202

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701

PAR2001

PAR4802PAS102

PAS302

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1702

PAU2003

PAU2006

PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAP501

PAR4501

PAU1037


PAU1807

PAH1022

PAH1023PAH1024

PAU1806

PAH1043 PAH1040PAR401

PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAP105

PAH101PAH102

PAH103

PAU1407

PAP503

PAR4001

PAU1039

PAH104

PAH105PAH106

PAU1406

PAP6011

PAC1202

PAP203

PAP609

PAR1401PAS104

PAU103

PAP7010

PAP5012

PAR3101

PAR3201

PAU1048

PAH1034PAR201

PAH1033PAR301

PAH1036PAP4011

PAU1026

PAH1035

PAP301

PAU1024

PAH1047

PAP401

PAR601

PAU1036

PAH1039

PAP406

PAR501

PAU1031

PAP7016

PAP3011

PAQ101

PAU1014

PAH1048

PAP309

PAU1016

PAH1037

PAU201

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAH1044

PAP5011

PAU1047

PAU202

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502

PAU1038

PAC4702

PAC4801

PAC4901

PAJ202PAJ203

PAR4801

PAU1401

PAU1501PAU1601

PAU1801

PAU1901

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501PAU504PAC2101

PAR2201

PAU503PAC3001

PAR2501

PAR2602PAC3002PAU803

PAU901PAU904PAC3101

PAR2601

PAU903

PAC3701

PAJ101

PAL701

PAC3802

PAJ102PAU13016 PAC3902

PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002 PAR1802

PAD1102

PAR1902

PAH1038PAL301

PAL702

PAU13020

PAP603PAR3102

PAP605PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ301PAR3701

PAR4002PAQ302PAR3702

PAU1404

PAQ303

PAU1405

PAQ401PAR3801 PAR4102

PAQ402PAR3802

PAU1504PAQ403

PAU1505


PAQ502PAR3902

PAU1604

PAQ503

PAU1605

PAQ601PAR4301

PAR4502PAQ602

PAR4302

PAU1804

PAQ603

PAU1805

PAQ701PAR4401

PAR4602PAQ702

PAR4402PAU1904

PAQ703

PAU1905

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1801PAU13023

PAR1901PAU13022

PAR2102

PAR2401

PAU301

PAR2301

PAU401

PAU404

PAR2402PAU304

PAU606

PAR2502

PAR2801PAU701

PAR2701

PAU801PAU804

PAR2802PAU704

PAU1006

PAR2901

PAU1101PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301

PAR3402 PAU1305

PAR3502

PAU13014

PAH1032

PAP4010

PAU1027

PAC4902 PAL802

PAU1403

PAU1503PAU1603

PAU1803

PAU1903

PAF101

PAJ201


96

PAC101PAC102

COC1PAC201PAC202

COC2PAC301

PAC302COC3

PAC401PAC402

COC4PAC501

PAC502COC5

PAC601PAC602COC6

PAC701PAC702COC7

PAC801PAC802COC8

PAC901PAC902COC9

PAC1001

PAC1002

COC10

PAC1101PAC1102

COC11

PAC1201PAC1202

COC12

PAC1301PAC1302

COC13

PAC1402


COC15

PAC1601PAC1602

COC16PAC1701PAC1702

COC17

PAC1802PAC1801

COC18


PAC2001PAC2002

COC20PAC2101

PAC2102COC21

PAC2201PAC2202

COC22

PAC2301PAC2302

COC23

PAC2401PAC2402

COC24

PAC2501PAC2502

COC25

PAC2601

PAC2602COC26PAC2701

PAC2702COC27

PAC2802PAC2801

COC28

PAC2901PAC2902

COC29

PAC3001PAC3002COC30

PAC3101PAC3102

COC31

PAC3201PAC3202

COC32


PAC3402PAC3401COC34

PAC3501PAC3502

COC35

PAC3601PAC3602

COC36

PAC3701PAC3702

COC37PAC3801PAC3802

COC38PAC3901

PAC3902COC39

PAC4001

PAC4002

COC40

PAC4102

PAC4101

COC41

PAC4202

PAC4201

COC42

PAC4301PAC4302

COC43

PAC4401PAC4402

COC44

PAC4501PAC4502

COC45

PAC4601PAC4602

COC46

PAC4701

PAC4702

COC47PAC4801

PAC4802

COC48PAC4901

PAC4902

COC49

PAD101

PAD102

COD1

PAD201

PAD202

COD2

PAD302

PAD301

COD3

PAD401

PAD402

COD4

PAD1001PAD1002

COD10

PAD1101PAD1102

COD11PAF101

PAF102

COF1PAH101

PAH103


PAH1013

PAH1015


PAH1023

PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049 PAH102

PAH104


PAH1014

PAH1016


PAH1024

PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAJ105PAJ101PAJ102

PAJ103PAJ104

COJ1PAJ203

PAJ201PAJ202

COJ2

PAL102PAL101

COL1

PAL202PAL201

COL2

PAL302PAL301COL3

PAL402

PAL401COL4

PAL502PAL501

COL5

PAL702PAL701

COL7

PAL801PAL802 COL8

PAP1014

PAP1013

PAP1012

PAP1011

PAP1010

PAP109

PAP108

PAP107

PAP106

PAP105

PAP104

PAP103

PAP102

PAP101

COP1

PAP2012

PAP2011

PAP2010

PAP209

PAP208

PAP207

PAP206

PAP205

PAP204

PAP203

PAP202

PAP201

COP2

PAP3012PAP3011

PAP3010PAP309

PAP308PAP307

PAP306PAP305

PAP304PAP303

PAP302PAP301

COP3PAP4012

PAP4011

PAP4010

PAP409

PAP408

PAP407

PAP406

PAP405

PAP404

PAP403

PAP402

PAP401

COP4PAP5012

PAP5011PAP5010

PAP509PAP508

PAP507PAP506

PAP505PAP504

PAP503PAP502

PAP501

COP5

PAP60U1

PAP6020

PAP6018

PAP6016

PAP6014

PAP6012

PAP6010

PAP608

PAP606

PAP604

PAP602

PAP6019

PAP6017

PAP6015

PAP6013

PAP6011

PAP609

PAP607

PAP605

PAP603

PAP601

COP6

PAP70U1

PAP7020

PAP7018

PAP7016

PAP7014

PAP7012

PAP7010

PAP708

PAP706

PAP704

PAP702

PAP7019

PAP7017

PAP7015

PAP7013

PAP7011

PAP709

PAP707

PAP705

PAP703

PAP701

COP7

PAQ101PAQ102PAQ103

COQ1

PAQ203PAQ202

PAQ201

COQ2

PAQ301PAQ302

PAQ303

COQ3

PAQ403PAQ402

PAQ401

COQ4

PAQ501

PAQ502PAQ503COQ5

PAQ601PAQ602

PAQ603

COQ6

PAQ701PAQ702

PAQ703

COQ7

PAR102PAR101

COR1

PAR202PAR201

COR2

PAR302PAR301

COR3

PAR402PAR401COR4

PAR502PAR501

COR5

PAR602PAR601

COR6

PAR702PAR701

COR7

PAR802

PAR801

COR8

PAR902

PAR901

COR9

PAR1001PAR1002

COR10

PAR1102PAR1101

COR11PAR1202

PAR1201

COR12

PAR1302PAR1301

COR13

PAR1402PAR1401COR14

PAR1502PAR1501

COR15

PAR1702PAR1701

COR17

PAR1802PAR1801


PAR2002PAR2001COR20

PAR2102

PAR2101COR21

PAR2202PAR2201

COR22

PAR2302PAR2301

COR23

PAR2401

PAR2402

COR24

PAR2502PAR2501

COR25


PAR2701COR27

PAR2802PAR2801

COR28

PAR2902


PAR3001COR30

PAR3101PAR3102

COR31

PAR3202PAR3201

COR32

PAR3301PAR3302

COR33

PAR3402PAR3401

COR34PAR3502PAR3501

COR35

PAR3602

PAR3601

COR36

PAR3702PAR3701COR37

PAR3802

PAR3801COR38PAR3902

PAR3901COR39

PAR4002

PAR4001COR40


COR42

PAR4302PAR4301

COR43PAR4402

PAR4401

COR44

PAR4502

PAR4501COR45

PAR4602

PAR4601COR46

PAR4702

PAR4701COR47

PAR4802PAR4801

COR48

PART102

PART101

CORT1

PAS101

PAS103PAS104

PAS102 COS1

PAS203PAS202

PAS204PAS201

COS2

PAS301

PAS303PAS304

PAS302 COS3

PAU101PAU102

PAU103PAU104

PAU105

PAU106PAU107

PAU108PAU109

PAU1010PAU1011


PAU1025

PAU1026

PAU1027PAU1028

PAU1029PAU1030

PAU1031PAU1032

PAU1033

PAU1034PAU1035

PAU1036

PAU1037PAU1038


PAU1042PAU1043


PAU1048COU1



COU2

PAU305PAU304PAU303

PAU302PAU301

COU3

PAU405PAU404PAU403

PAU402PAU401

COU4

PAU505PAU504PAU503

PAU502PAU501

COU5

PAU601

PAU606

PAU602PAU603


PAU705PAU704PAU703

PAU702PAU701

COU7

PAU805PAU804PAU803

PAU802PAU801

COU8

PAU905PAU904PAU903

PAU902PAU901

COU9

PAU1001

PAU1006

PAU1002PAU1003

PAU1005PAU1004

PAU100HSCOU10


PAU1102PAU1101

COU11

PAU1201





COU13


COU14

PAU1501PAU1503

PAU1504PAU1505

PAU1506PAU1507

COU15

PAU1601PAU1603

PAU1604PAU1605

PAU1606PAU1607


COU17

PAU1801PAU1803

PAU1804PAU1805

PAU1806PAU1807

COU18

PAU1901PAU1903

PAU1904PAU1905

PAU1906PAU1907

COU19

PAU2006


PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAC4802

PAF102

PAL801

PART101

PAP2010

PAU1010

PAP205

PAU105

PAP204

PAU104

PAP206PAU106

PAP3012

PAU1013

PAP3010


PAC702PAC802

PAC1402

PAC1802

PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501

PAP2011

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102PAP402

PAU1035

PAC102PAC202

PAC2401

PAD101PAD201

PAD302

PAL101

PAL201

PAL302

PAP601

PAR202

PAR302

PAR402PAR502

PAR602PAR1102

PAR1202

PAR1402

PAR4702

PAU208

PAC1301

PAC1502

PAC1601

PAC1701

PAC2301

PAC2601

PAC2701PAC3301

PAL502

PAU303

PAU305

PAU405 PAU505

PAU703PAU705

PAU805 PAU905

PAU1105

PAC3202PAP209

PAR2702

PAU109

PAC2202

PAP208PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602

PAC1702


PAC2201

PAC2302

PAC2602

PAC2702

PAC2801

PAC2901

PAC3102

PAC3201

PAC3302

PAC3401

PAC3501

PAC3601

PAL402

PAP2012

PAR3002

PAU1012

PAU302


PAU604

PAU702


PAU1102 PAU1202 PAU1204

PAH107PAH108PAH109

PAU1907

PAP504

PAR4601

PAU1040


PAU1906

PAP7012

PAR3601

PAH1013PAH1014

PAH1015

PAU1507PAP408

PAR4101PAU1029

PAH1016

PAH1017PAH1018

PAU1506

PAP702

PAS304

PAP7018

PAP409PAR4201

PAU1028

PAH1025PAH1026

PAH1027PAU1607

PAH1028


PAP704

PAP706

PAP708

PAP7020

PAP6013

PAP607

PAP6017

PAP1013

PAP1014

PAH1046PAR102

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAD402

PAH1031

PAH1041

PAH1045

PAJ104PAJ105

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6019

PAP7014

PAQ102

PAQ202

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701

PAR2001

PAR4802PAS102

PAS302

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1702

PAU2003

PAU2006

PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAP501

PAR4501

PAU1037


PAU1807

PAH1022

PAH1023PAH1024

PAU1806


PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAP105

PAH101PAH102

PAH103

PAU1407

PAP503

PAR4001

PAU1039

PAH104

PAH105PAH106

PAU1406

PAP6011

PAC1202

PAP203

PAP609

PAR1401PAS104

PAU103

PAP7010

PAP5012

PAR3101

PAR3201

PAU1048

PAH1034PAR201

PAH1033PAR301

PAH1036PAP4011

PAU1026

PAH1035

PAP301

PAU1024

PAH1047

PAP401

PAR601

PAU1036

PAH1039

PAP406

PAR501

PAU1031

PAP7016

PAP3011

PAQ101

PAU1014

PAH1048

PAP309

PAU1016

PAH1037

PAU201

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAH1044

PAP5011

PAU1047

PAU202

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502

PAU1038

PAC4702

PAC4801

PAC4901

PAJ202PAJ203

PAR4801

PAU1401

PAU1501PAU1601

PAU1801

PAU1901

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501PAU504PAC2101

PAR2201

PAU503PAC3001

PAR2501


PAU901PAU904PAC3101

PAR2601

PAU903

PAC3701

PAJ101

PAL701

PAC3802


PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002 PAR1802

PAD1102

PAR1902

PAH1038PAL301

PAL702

PAU13020

PAP603PAR3102

PAP605PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ301PAR3701


PAU1404

PAQ303

PAU1405


PAQ402PAR3802

PAU1504PAQ403

PAU1505


PAQ502PAR3902

PAU1604

PAQ503

PAU1605

PAQ601PAR4301

PAR4502PAQ602

PAR4302

PAU1804

PAQ603

PAU1805

PAQ701PAR4401

PAR4602PAQ702

PAR4402PAU1904

PAQ703

PAU1905

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1801PAU13023

PAR1901PAU13022

PAR2102

PAR2401

PAU301

PAR2301

PAU401

PAU404

PAR2402PAU304

PAU606

PAR2502

PAR2801PAU701

PAR2701

PAU801PAU804

PAR2802PAU704

PAU1006

PAR2901

PAU1101PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301

PAR3402 PAU1305

PAR3502

PAU13014

PAH1032

PAP4010

PAU1027

PAC4902 PAL802

PAU1403

PAU1503PAU1603

PAU1803

PAU1903

PAF101

PAJ201


97

PAC101PAC102

COC1PAC201PAC202

COC2PAC301

PAC302COC3

PAC401PAC402

COC4PAC501

PAC502COC5

PAC601PAC602COC6

PAC701PAC702COC7

PAC801PAC802COC8

PAC901PAC902COC9

PAC1001

PAC1002

COC10

PAC1101PAC1102

COC11

PAC1201PAC1202

COC12

PAC1301PAC1302

COC13

PAC1402


COC15

PAC1601PAC1602

COC16PAC1701PAC1702

COC17

PAC1802PAC1801

COC18


PAC2001PAC2002

COC20PAC2101

PAC2102COC21

PAC2201PAC2202

COC22

PAC2301PAC2302

COC23

PAC2401PAC2402

COC24

PAC2501PAC2502

COC25

PAC2601

PAC2602COC26PAC2701

PAC2702COC27

PAC2802PAC2801

COC28

PAC2901PAC2902

COC29

PAC3001PAC3002COC30

PAC3101PAC3102

COC31

PAC3201PAC3202

COC32


PAC3402PAC3401COC34

PAC3501PAC3502

COC35

PAC3601PAC3602

COC36

PAC3701PAC3702

COC37PAC3801PAC3802

COC38PAC3901

PAC3902COC39

PAC4001

PAC4002

COC40

PAC4102

PAC4101

COC41

PAC4202

PAC4201

COC42

PAC4301PAC4302

COC43

PAC4401PAC4402

COC44

PAC4501PAC4502

COC45

PAC4601PAC4602

COC46

PAC4701

PAC4702

COC47PAC4801

PAC4802

COC48PAC4901

PAC4902

COC49

PAD101

PAD102

COD1

PAD201

PAD202

COD2

PAD302

PAD301

COD3

PAD401

PAD402

COD4

PAD1001PAD1002

COD10

PAD1101PAD1102

COD11PAF101

PAF102

COF1PAH101

PAH103


PAH1013

PAH1015


PAH1023

PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049 PAH102

PAH104


PAH1014

PAH1016


PAH1024

PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAJ105PAJ101PAJ102

PAJ103PAJ104

COJ1PAJ203

PAJ201PAJ202

COJ2

PAL102PAL101

COL1

PAL202PAL201

COL2

PAL302PAL301COL3

PAL402

PAL401COL4

PAL502PAL501

COL5

PAL702PAL701

COL7

PAL801PAL802 COL8

PAP1014

PAP1013

PAP1012

PAP1011

PAP1010

PAP109

PAP108

PAP107

PAP106

PAP105

PAP104

PAP103

PAP102

PAP101

COP1

PAP2012

PAP2011

PAP2010

PAP209

PAP208

PAP207

PAP206

PAP205

PAP204

PAP203

PAP202

PAP201

COP2

PAP3012PAP3011

PAP3010PAP309

PAP308PAP307

PAP306PAP305

PAP304PAP303

PAP302PAP301

COP3PAP4012

PAP4011

PAP4010

PAP409

PAP408

PAP407

PAP406

PAP405

PAP404

PAP403

PAP402

PAP401

COP4PAP5012

PAP5011PAP5010

PAP509PAP508

PAP507PAP506

PAP505PAP504

PAP503PAP502

PAP501

COP5

PAP60U1

PAP6020

PAP6018

PAP6016

PAP6014

PAP6012

PAP6010

PAP608

PAP606

PAP604

PAP602

PAP6019

PAP6017

PAP6015

PAP6013

PAP6011

PAP609

PAP607

PAP605

PAP603

PAP601

COP6

PAP70U1

PAP7020

PAP7018

PAP7016

PAP7014

PAP7012

PAP7010

PAP708

PAP706

PAP704

PAP702

PAP7019

PAP7017

PAP7015

PAP7013

PAP7011

PAP709

PAP707

PAP705

PAP703

PAP701

COP7

PAQ101PAQ102PAQ103

COQ1

PAQ203PAQ202

PAQ201

COQ2

PAQ301PAQ302

PAQ303

COQ3

PAQ403PAQ402

PAQ401

COQ4

PAQ501

PAQ502PAQ503COQ5

PAQ601PAQ602

PAQ603

COQ6

PAQ701PAQ702

PAQ703

COQ7

PAR102PAR101

COR1

PAR202PAR201

COR2

PAR302PAR301

COR3

PAR402PAR401COR4

PAR502PAR501

COR5

PAR602PAR601

COR6

PAR702PAR701

COR7

PAR802

PAR801

COR8

PAR902

PAR901

COR9

PAR1001PAR1002

COR10

PAR1102PAR1101

COR11PAR1202

PAR1201

COR12

PAR1302PAR1301

COR13

PAR1402PAR1401COR14

PAR1502PAR1501

COR15

PAR1702PAR1701

COR17

PAR1802PAR1801


PAR2002PAR2001COR20

PAR2102

PAR2101COR21

PAR2202PAR2201

COR22

PAR2302PAR2301

COR23

PAR2401

PAR2402

COR24

PAR2502PAR2501

COR25


PAR2701COR27

PAR2802PAR2801

COR28

PAR2902


PAR3001COR30

PAR3101PAR3102

COR31

PAR3202PAR3201

COR32

PAR3301PAR3302

COR33

PAR3402PAR3401

COR34PAR3502PAR3501

COR35

PAR3602

PAR3601

COR36

PAR3702PAR3701COR37

PAR3802

PAR3801COR38PAR3902

PAR3901COR39

PAR4002

PAR4001COR40


COR42

PAR4302PAR4301

COR43PAR4402

PAR4401

COR44

PAR4502

PAR4501COR45

PAR4602

PAR4601COR46

PAR4702

PAR4701COR47

PAR4802PAR4801

COR48

PART102

PART101

CORT1

PAS101

PAS103PAS104

PAS102 COS1

PAS203PAS202

PAS204PAS201

COS2

PAS301

PAS303PAS304

PAS302 COS3

PAU101PAU102

PAU103PAU104

PAU105

PAU106PAU107

PAU108PAU109

PAU1010PAU1011


PAU1025

PAU1026

PAU1027PAU1028

PAU1029PAU1030

PAU1031PAU1032

PAU1033

PAU1034PAU1035

PAU1036

PAU1037PAU1038


PAU1042PAU1043


PAU1048COU1



COU2

PAU305PAU304PAU303

PAU302PAU301

COU3

PAU405PAU404PAU403

PAU402PAU401

COU4

PAU505PAU504PAU503

PAU502PAU501

COU5

PAU601

PAU606

PAU602PAU603


PAU705PAU704PAU703

PAU702PAU701

COU7

PAU805PAU804PAU803

PAU802PAU801

COU8

PAU905PAU904PAU903

PAU902PAU901

COU9

PAU1001

PAU1006

PAU1002PAU1003

PAU1005PAU1004

PAU100HSCOU10


PAU1102PAU1101

COU11

PAU1201





COU13


COU14

PAU1501PAU1503

PAU1504PAU1505

PAU1506PAU1507

COU15

PAU1601PAU1603

PAU1604PAU1605

PAU1606PAU1607


COU17

PAU1801PAU1803

PAU1804PAU1805

PAU1806PAU1807

COU18

PAU1901PAU1903

PAU1904PAU1905

PAU1906PAU1907

COU19

PAU2006


PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAC4802

PAF102

PAL801

PART101

PAP2010

PAU1010

PAP205

PAU105

PAP204

PAU104

PAP206PAU106

PAP3012

PAU1013

PAP3010


PAC702PAC802

PAC1402

PAC1802

PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501

PAP2011

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102PAP402

PAU1035

PAC102PAC202

PAC2401

PAD101PAD201

PAD302

PAL101

PAL201

PAL302

PAP601

PAR202

PAR302

PAR402PAR502

PAR602PAR1102

PAR1202

PAR1402

PAR4702

PAU208

PAC1301

PAC1502

PAC1601

PAC1701

PAC2301

PAC2601

PAC2701PAC3301

PAL502

PAU303

PAU305

PAU405 PAU505

PAU703PAU705

PAU805 PAU905

PAU1105

PAC3202PAP209

PAR2702

PAU109

PAC2202

PAP208PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602

PAC1702


PAC2201

PAC2302

PAC2602

PAC2702

PAC2801

PAC2901

PAC3102

PAC3201

PAC3302

PAC3401

PAC3501

PAC3601

PAL402

PAP2012

PAR3002

PAU1012

PAU302


PAU604

PAU702


PAU1102 PAU1202 PAU1204

PAH107PAH108PAH109

PAU1907

PAP504

PAR4601

PAU1040


PAU1906

PAP7012

PAR3601

PAH1013PAH1014

PAH1015

PAU1507PAP408

PAR4101PAU1029

PAH1016

PAH1017PAH1018

PAU1506

PAP702

PAS304

PAP7018

PAP409PAR4201

PAU1028

PAH1025PAH1026

PAH1027PAU1607

PAH1028


PAP704

PAP706

PAP708

PAP7020

PAP6013

PAP607

PAP6017

PAP1013

PAP1014

PAH1046PAR102

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAD402

PAH1031

PAH1041

PAH1045

PAJ104PAJ105

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6019

PAP7014

PAQ102

PAQ202

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701

PAR2001

PAR4802PAS102

PAS302

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1702

PAU2003

PAU2006

PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAP501

PAR4501

PAU1037


PAU1807

PAH1022

PAH1023PAH1024

PAU1806


PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAP105

PAH101PAH102

PAH103

PAU1407

PAP503

PAR4001

PAU1039

PAH104

PAH105PAH106

PAU1406

PAP6011

PAC1202

PAP203

PAP609

PAR1401PAS104

PAU103

PAP7010

PAP5012

PAR3101

PAR3201

PAU1048

PAH1034PAR201

PAH1033PAR301

PAH1036PAP4011

PAU1026

PAH1035

PAP301

PAU1024

PAH1047

PAP401

PAR601

PAU1036

PAH1039

PAP406

PAR501

PAU1031

PAP7016

PAP3011

PAQ101

PAU1014

PAH1048

PAP309

PAU1016

PAH1037

PAU201

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAH1044

PAP5011

PAU1047

PAU202

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502

PAU1038

PAC4702

PAC4801

PAC4901

PAJ202PAJ203

PAR4801

PAU1401

PAU1501PAU1601

PAU1801

PAU1901

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501PAU504PAC2101

PAR2201

PAU503PAC3001

PAR2501


PAU901PAU904PAC3101

PAR2601

PAU903

PAC3701

PAJ101

PAL701

PAC3802


PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002 PAR1802

PAD1102

PAR1902

PAH1038PAL301

PAL702

PAU13020

PAP603PAR3102

PAP605PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ301PAR3701


PAU1404

PAQ303

PAU1405


PAQ402PAR3802

PAU1504PAQ403

PAU1505


PAQ502PAR3902

PAU1604

PAQ503

PAU1605

PAQ601PAR4301

PAR4502PAQ602

PAR4302

PAU1804

PAQ603

PAU1805

PAQ701PAR4401

PAR4602PAQ702

PAR4402PAU1904

PAQ703

PAU1905

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1801PAU13023

PAR1901PAU13022

PAR2102

PAR2401

PAU301

PAR2301

PAU401

PAU404

PAR2402PAU304

PAU606

PAR2502

PAR2801PAU701

PAR2701

PAU801PAU804

PAR2802PAU704

PAU1006

PAR2901

PAU1101PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301

PAR3402 PAU1305

PAR3502

PAU13014

PAH1032

PAP4010

PAU1027

PAC4902 PAL802

PAU1403

PAU1503PAU1603

PAU1803

PAU1903

PAF101

PAJ201


98

PAC101PAC102

COC1PAC201PAC202

COC2PAC301

PAC302COC3

PAC401PAC402

COC4PAC501

PAC502COC5

PAC601PAC602COC6

PAC701PAC702COC7

PAC801PAC802COC8

PAC901PAC902COC9

PAC1001

PAC1002

COC10

PAC1101PAC1102

COC11

PAC1201PAC1202

COC12

PAC1301PAC1302

COC13

PAC1402


COC15

PAC1601PAC1602

COC16PAC1701PAC1702

COC17

PAC1802PAC1801

COC18


PAC2001PAC2002

COC20PAC2101

PAC2102COC21

PAC2201PAC2202

COC22

PAC2301PAC2302

COC23

PAC2401PAC2402

COC24

PAC2501PAC2502

COC25

PAC2601

PAC2602COC26PAC2701

PAC2702COC27

PAC2802PAC2801

COC28

PAC2901PAC2902

COC29

PAC3001PAC3002COC30

PAC3101PAC3102

COC31

PAC3201PAC3202

COC32


PAC3402PAC3401COC34

PAC3501PAC3502

COC35

PAC3601PAC3602

COC36

PAC3701PAC3702

COC37PAC3801PAC3802

COC38PAC3901

PAC3902COC39

PAC4001

PAC4002

COC40

PAC4102

PAC4101

COC41

PAC4202

PAC4201

COC42

PAC4301PAC4302

COC43

PAC4401PAC4402

COC44

PAC4501PAC4502

COC45

PAC4601PAC4602

COC46

PAC4701

PAC4702

COC47PAC4801

PAC4802

COC48PAC4901

PAC4902

COC49

PAD101

PAD102

COD1

PAD201

PAD202

COD2

PAD302

PAD301

COD3

PAD401

PAD402

COD4

PAD1001PAD1002

COD10

PAD1101PAD1102

COD11PAF101

PAF102

COF1PAH101

PAH103


PAH1013

PAH1015


PAH1023

PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049 PAH102

PAH104


PAH1014

PAH1016


PAH1024

PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAJ105PAJ101PAJ102

PAJ103PAJ104

COJ1PAJ203

PAJ201PAJ202

COJ2

PAL102PAL101

COL1

PAL202PAL201

COL2

PAL302PAL301COL3

PAL402

PAL401COL4

PAL502PAL501

COL5

PAL702PAL701

COL7

PAL801PAL802 COL8

PAP1014

PAP1013

PAP1012

PAP1011

PAP1010

PAP109

PAP108

PAP107

PAP106

PAP105

PAP104

PAP103

PAP102

PAP101

COP1

PAP2012

PAP2011

PAP2010

PAP209

PAP208

PAP207

PAP206

PAP205

PAP204

PAP203

PAP202

PAP201

COP2

PAP3012PAP3011

PAP3010PAP309

PAP308PAP307

PAP306PAP305

PAP304PAP303

PAP302PAP301

COP3PAP4012

PAP4011

PAP4010

PAP409

PAP408

PAP407

PAP406

PAP405

PAP404

PAP403

PAP402

PAP401

COP4PAP5012

PAP5011PAP5010

PAP509PAP508

PAP507PAP506

PAP505PAP504

PAP503PAP502

PAP501

COP5

PAP60U1

PAP6020

PAP6018

PAP6016

PAP6014

PAP6012

PAP6010

PAP608

PAP606

PAP604

PAP602

PAP6019

PAP6017

PAP6015

PAP6013

PAP6011

PAP609

PAP607

PAP605

PAP603

PAP601

COP6

PAP70U1

PAP7020

PAP7018

PAP7016

PAP7014

PAP7012

PAP7010

PAP708

PAP706

PAP704

PAP702

PAP7019

PAP7017

PAP7015

PAP7013

PAP7011

PAP709

PAP707

PAP705

PAP703

PAP701

COP7

PAQ101PAQ102PAQ103

COQ1

PAQ203PAQ202

PAQ201

COQ2

PAQ301PAQ302

PAQ303

COQ3

PAQ403PAQ402

PAQ401

COQ4

PAQ501

PAQ502PAQ503COQ5

PAQ601PAQ602

PAQ603

COQ6

PAQ701PAQ702

PAQ703

COQ7

PAR102PAR101

COR1

PAR202PAR201

COR2

PAR302PAR301

COR3

PAR402PAR401COR4

PAR502PAR501

COR5

PAR602PAR601

COR6

PAR702PAR701

COR7

PAR802

PAR801

COR8

PAR902

PAR901

COR9

PAR1001PAR1002

COR10

PAR1102PAR1101

COR11PAR1202

PAR1201

COR12

PAR1302PAR1301

COR13

PAR1402PAR1401COR14

PAR1502PAR1501

COR15

PAR1702PAR1701

COR17

PAR1802PAR1801


PAR2002PAR2001COR20

PAR2102

PAR2101COR21

PAR2202PAR2201

COR22

PAR2302PAR2301

COR23

PAR2401

PAR2402

COR24

PAR2502PAR2501

COR25


PAR2701COR27

PAR2802PAR2801

COR28

PAR2902


PAR3001COR30

PAR3101PAR3102

COR31

PAR3202PAR3201

COR32

PAR3301PAR3302

COR33

PAR3402PAR3401

COR34PAR3502PAR3501

COR35

PAR3602

PAR3601

COR36

PAR3702PAR3701COR37

PAR3802

PAR3801COR38PAR3902

PAR3901COR39

PAR4002

PAR4001COR40


COR42

PAR4302PAR4301

COR43PAR4402

PAR4401

COR44

PAR4502

PAR4501COR45

PAR4602

PAR4601COR46

PAR4702

PAR4701COR47

PAR4802PAR4801

COR48

PART102

PART101

CORT1

PAS101

PAS103PAS104

PAS102 COS1

PAS203PAS202

PAS204PAS201

COS2

PAS301

PAS303PAS304

PAS302 COS3

PAU101PAU102

PAU103PAU104

PAU105

PAU106PAU107

PAU108PAU109

PAU1010PAU1011


PAU1025

PAU1026

PAU1027PAU1028

PAU1029PAU1030

PAU1031PAU1032

PAU1033

PAU1034PAU1035

PAU1036

PAU1037PAU1038


PAU1042PAU1043


PAU1048COU1



COU2

PAU305PAU304PAU303

PAU302PAU301

COU3

PAU405PAU404PAU403

PAU402PAU401

COU4

PAU505PAU504PAU503

PAU502PAU501

COU5

PAU601

PAU606

PAU602PAU603


PAU705PAU704PAU703

PAU702PAU701

COU7

PAU805PAU804PAU803

PAU802PAU801

COU8

PAU905PAU904PAU903

PAU902PAU901

COU9

PAU1001

PAU1006

PAU1002PAU1003

PAU1005PAU1004

PAU100HSCOU10


PAU1102PAU1101

COU11

PAU1201





COU13


COU14

PAU1501PAU1503

PAU1504PAU1505

PAU1506PAU1507

COU15

PAU1601PAU1603

PAU1604PAU1605

PAU1606PAU1607


COU17

PAU1801PAU1803

PAU1804PAU1805

PAU1806PAU1807

COU18

PAU1901PAU1903

PAU1904PAU1905

PAU1906PAU1907

COU19

PAU2006


PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAC4802

PAF102

PAL801

PART101

PAP2010

PAU1010

PAP205

PAU105

PAP204

PAU104

PAP206PAU106

PAP3012

PAU1013

PAP3010


PAC702PAC802

PAC1402

PAC1802

PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501

PAP2011

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102PAP402

PAU1035

PAC102PAC202

PAC2401

PAD101PAD201

PAD302

PAL101

PAL201

PAL302

PAP601

PAR202

PAR302

PAR402PAR502

PAR602PAR1102

PAR1202

PAR1402

PAR4702

PAU208

PAC1301

PAC1502

PAC1601

PAC1701

PAC2301

PAC2601

PAC2701PAC3301

PAL502

PAU303

PAU305

PAU405 PAU505

PAU703PAU705

PAU805 PAU905

PAU1105

PAC3202PAP209

PAR2702

PAU109

PAC2202

PAP208PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602

PAC1702


PAC2201

PAC2302

PAC2602

PAC2702

PAC2801

PAC2901

PAC3102

PAC3201

PAC3302

PAC3401

PAC3501

PAC3601

PAL402

PAP2012

PAR3002

PAU1012

PAU302


PAU604

PAU702


PAU1102 PAU1202 PAU1204

PAH107PAH108PAH109

PAU1907

PAP504

PAR4601

PAU1040


PAU1906

PAP7012

PAR3601

PAH1013PAH1014

PAH1015

PAU1507PAP408

PAR4101PAU1029

PAH1016

PAH1017PAH1018

PAU1506

PAP702

PAS304

PAP7018

PAP409PAR4201

PAU1028

PAH1025PAH1026

PAH1027PAU1607

PAH1028


PAP704

PAP706

PAP708

PAP7020

PAP6013

PAP607

PAP6017

PAP1013

PAP1014

PAH1046PAR102

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAD402

PAH1031

PAH1041

PAH1045

PAJ104PAJ105

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6019

PAP7014

PAQ102

PAQ202

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701

PAR2001

PAR4802PAS102

PAS302

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1702

PAU2003

PAU2006

PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAP501

PAR4501

PAU1037


PAU1807

PAH1022

PAH1023PAH1024

PAU1806


PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAP105

PAH101PAH102

PAH103

PAU1407

PAP503

PAR4001

PAU1039

PAH104

PAH105PAH106

PAU1406

PAP6011

PAC1202

PAP203

PAP609

PAR1401PAS104

PAU103

PAP7010

PAP5012

PAR3101

PAR3201

PAU1048

PAH1034PAR201

PAH1033PAR301

PAH1036PAP4011

PAU1026

PAH1035

PAP301

PAU1024

PAH1047

PAP401

PAR601

PAU1036

PAH1039

PAP406

PAR501

PAU1031

PAP7016

PAP3011

PAQ101

PAU1014

PAH1048

PAP309

PAU1016

PAH1037

PAU201

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAH1044

PAP5011

PAU1047

PAU202

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502

PAU1038

PAC4702

PAC4801

PAC4901

PAJ202PAJ203

PAR4801

PAU1401

PAU1501PAU1601

PAU1801

PAU1901

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501PAU504PAC2101

PAR2201

PAU503PAC3001

PAR2501


PAU901PAU904PAC3101

PAR2601

PAU903

PAC3701

PAJ101

PAL701

PAC3802


PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002 PAR1802

PAD1102

PAR1902

PAH1038PAL301

PAL702

PAU13020

PAP603PAR3102

PAP605PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ301PAR3701


PAU1404

PAQ303

PAU1405


PAQ402PAR3802

PAU1504PAQ403

PAU1505


PAQ502PAR3902

PAU1604

PAQ503

PAU1605

PAQ601PAR4301

PAR4502PAQ602

PAR4302

PAU1804

PAQ603

PAU1805

PAQ701PAR4401

PAR4602PAQ702

PAR4402PAU1904

PAQ703

PAU1905

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1801PAU13023

PAR1901PAU13022

PAR2102

PAR2401

PAU301

PAR2301

PAU401

PAU404

PAR2402PAU304

PAU606

PAR2502

PAR2801PAU701

PAR2701

PAU801PAU804

PAR2802PAU704

PAU1006

PAR2901

PAU1101PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301

PAR3402 PAU1305

PAR3502

PAU13014

PAH1032

PAP4010

PAU1027

PAC4902 PAL802

PAU1403

PAU1503PAU1603

PAU1803

PAU1903

PAF101

PAJ201


99

B.3. LED Controller Bill of Materials

B.3 LED Controller Bill of Materials


100

Bill o

f Ma

teria

ls

So

urc

e D

ata

Fro

m:

LE

D C

on

trolle

r, Re

v 1

Pro

jec

t:L

ED

Co

ntro

ller, R

ev 1

De

sig

n:

M. A

ldric

h

Report D

ate:1/12/2010

12:53:02 PM

Print D

ate:29-A

pr-106:39:26 P

M#

Desig

nato

rD

escrip

tion

Man

ufa

ctu

rer 1

Man

ufa

ctu

rer P

art N

um

ber 1

Fo

otp

rint

Su

pp

lier 1

Su

pp

lier P

art N

um

ber 1

Qu

an

tity

1N

/ALED

Controller, PC

BAdvanced C

ircuitsN

/AN

/AN

/AN

/A1

2C

1, C2

CAP C

ERAM

IC 2.2U

F 10V X5R 0805

Kemet

C0805C

225K8PACTU

CAPC

2012LD

igi-Key399-3125-1-N

D2

3C

3, C7, C

14C

AP CER

M 1U

F 6.3V X5R 0402

Kemet

C0402C

105K9PACTU

CAPC

1005LD

igi-Key399-4873-1-N

D3

4C

4, C8, C

37, C46

CAP C

ER 10000PF 25V 10%

X7R 0402

Murata Electronics N

orth America

GR

M155R

71E103KA01DC

APC1005L

Digi-Key

490-1312-1-ND

45

C5

CAP C

ER 2.2PF 50V C

0G 0402


orth America

GR

M1555C

1H2R

2CZ01D

CAPC

1005LD

igi-Key490-1267-1-N

D1

6C

6, C9, C

10, C11

CAP C

ERM

1UF 6.3V X5R

0402Kem

etC

0402C105K9PAC

TUC

APC1005L

Digi-Key

399-4873-1-ND

47

C12

CAP C

ER 1000PF 50V C

0G 0402,


orth America

GR

M155R

61A104KA01D,

CAPC

1005LD

igi-Key490-3244-1-N

D1

8C

20, C30

CAP C

ER .47U

F 6.3V X5R 0402


orth America

GR

M155R

60J474KE19DC

APC1005L

Digi-Key

490-3266-1-ND

29

C21, C

31C

AP CER

.22UF 6.3V X5R

0402M

urata Electronics North Am

ericaG

RM

155R60J224KE01D

CAPC

1005LD

igi-Key490-5407-1-N

D2

10C

20, C21, C

30, C31

CAP C

ER .1U

F 10V 10% X5R

0402M


ericaG

RM

155R61A104KA01D

CAPC

1005LD

igi-Key490-1318-1-N

D4

11C

13, C15, C

17, C19, C

23, C24, C

25, C27, C

29, C33, C

35C

AP CER

.1UF 10V 10%

X5R 0402


orth America

GR

M155R

61A104KA01DC

APC1005L

Digi-Key

490-1318-1-ND

1112

C16, C

26C

AP CER

.1UF 50V 10%

X7R 0603


orth America

GR

M188R

71H104KA93D

CAPC

1608ND

igi-Key490-1519-1-N

D

213

C18, C

28, C34

CAP C

ER 1.0U

F 10V 10% X5R

0603M


ericaG

RM

188R61A105KA61D

CAPC

1608ND

igi-Key490-1543-1-N

D3

14C

22, C32, C

36C

AP CER

3300PF 50V X7R 0402


orth America

GR

M155R

71H332KA01D

CAPC

1005LD

igi-Key490-3248-1-N

D3

15C

38, C39

CAP C

ER 30PF 50V 5%

C0G

0402M


ericaG

RM

1555C1H

300JZ01DC

APC1005L

Digi-Key

490-1285-1-ND

216

C40, C

41, C42, C

48, C49

CAP C

ER 22U

F 25V X5R 1210

Taiyo YudenTM

K325BJ226MM

-TC

APC3225N

Digi-Key

587-2086-1-ND

517

C43, C

44, C45

CAP C

ER 2.2U

F 6.3V 10% X7R

0805M


ericaG

RM

21BR70J225KA01L

CAPC

2012LD

igi-Key490-1698-1-N

D3

18C

47C

AP 47UF 35V ELEC

T NH

G R

ADIAL

Panasonic - ECG

ECA-1VH

G470

CAPPR

2-5x11D

igi-KeyP5550-N

D1

19D

2, D10, D

11LED

570NM

GR

EEN D

IFF 0603 SMD

Avago Technologies US Inc.

HSM

G-C

1901608

Digi-Key

516-1425-1-ND

320

D1, D

4LED

630NM

HE R

ED D

IFF 0603 SMD


HSM

S-C190

1608D

igi-Key516-1422-1-N

D2

21D

3EM

ITTER IR

5MM

HI EFF 940N

MVishay/Sem

iconductorsTSAL6200

HLM

P-5029D

igi-Key751-1204-N

D1

22(D

3)H

OU

SING

RA FO

R 5M

M H

IGH

DO

ME LED


HLM

P-5029N

/AD

igi-Key516-1395-N

D1

23F1

PTC R

ESET 33V 1.85A SMD

2920Littelfuse Inc

2920L185DR

2920L185DR

Digi-Key

F2871CT-N

D1

24H

1C

ON

N H

EADER

50 POS STR

GH

T GO

LD3M

N2550-6002-R

BN

2550-6002-RB

Digi-Key

MH

C50K-N

D1

25J1

USB R

CPT B-TYPE R

/A FULL BAC

KFC

I61729-1010BLF

61729-1010BLFD

igi-Key609-3656-N

D1

26J2

MALE BAR

REL C

ON

NEC

TOR

2.1x5.5Sw

itchcraft Inc.R

APC722X

RAPC

722XM

ouser502-R

APC722X

127

L1, L3FER

RITE C

HIP 60 O

HM

1700MA 0402


orth America

BLM15PD

600SN1D

1005D

igi-Key490-5201-1-N

D2

28L2, L4

FILTER C

HIP 220 O

HM

700MA 0402


orth America

BLM15EG

221SN1D

1005D

igi-Key490-4005-1-N

D2

29L5

IND

UC

TOR

MU

LTILAYER 330N

H 0402

TDK C

orporationM

LG1005SR

33J1005

Digi-Key

445-3077-1-ND

130

(L5)IN

DU

CTO

R M

ULTILAYER

2.2UH

0402TD

K Corporation

MLF1005A2R

2KT1005

Digi-Key

445-3518-1-ND

131

(L5)FER

RITE C

HIP 1000 O

HM

200MA 0402


orth America

BLM15H

G102SN

1D1005

Digi-Key

490-3999-1-ND

132

L7FER

RITE C

HIP 40 O

HM

550MA 0402

TDK C

orporationM

MZ1005Y400C

1005D

igi-Key445-2150-2-N

D1

33L8

IND

UC

TOR

22UH

1.8A SMD

SHIELD

EDM

urata Power Solutions Inc

46223C46223C

Digi-Key

811-1173-1-ND

134

P1C

ON

N H

DR

DU

AL 14POS .100 SR

T AUM

olex Connector C

orporation10-89-7142

THD

igi-KeyW

M26814-N

D1

35P2, P3, P4, P5

CO

NN

HEAD

ER 12PO

S .100 VERT TIN

Molex C

onnector Corporation

22-28-4120TH

Digi-Key

WM

6412-ND

436

P6, P7C

ON

N R

ECEPT 20PO

S 2MM

STR D

L SMD

FCI

91596-120LFSM

TD

igi-Key609-2730-N

D2

37Q

1, Q2

MO

SFET N-C

H 30V 1.7A SSO

T3Fairchild Sem

iconductorN

DS355AN

SOT-23

Digi-Key

ND

S355ANC

T-ND

238

Q3, Q

4, Q5, Q

6, Q7

TRAN

SISTOR

GP PN

P AMP SO

T-23Fairchild Sem

iconductorM

MBT3906

SO-G

3D

igi-KeyM

MBT3906FSC

T-ND

539

R10

RES 16.9 O

HM

1/8W 1%

0805 SMD

Rohm

Semiconductor

MC

R10EZH

F16R9

RESC

2012LD

igi-KeyR

HM

16.9CC

T-ND

140

(R10)

RES 4.32 O

HM

1/8W 1%

0805 SMD

Vishay/Dale

CR

CW

08054R32FKEA

RESC

2012LD

igi-Key541-4.32C

CC

T-ND

141

R37, R

38, R39, R

43, R44, R

40, R41, R

42, R45, R

46R

ES 5.1K OH

M 1/10W

5% 0603 SM

DVishay/D

aleC

RC

W06035K10JN

EAR

ESC1608N

Digi-Key

541-5.1KGC

T-ND

1042

R24, R

28, R30 (0-500 lx)

RES 820K O

HM

1/16W 1%

0402 SMD

Vishay/Dale

CR

CW

0402820KFKEDR

ESC1005L

Digi-Key

541-820KLCT-N

D3

43(R

24, R28, R

30) (0-2,000 lx)R

ES 402K OH

M 1/16W

1% 0402 SM

DVishay/D

aleC

RC

W0402402KFKED

RESC

1005LD

igi-Key541-402KLC

T-ND

344

(R24, R

28, R30) (0-10,000 lx)

RES 182K O

HM

1/16W 1%

0402 SMD

Vishay/Dale

CR

CW

0402182KFKEDR

ESC1005L

Digi-Key

541-182KLCT-N

D3

Altiu

m L

imite

d C

on

fide

ntia

l4/29/2010

Page 1


101

45(R

24, R28, R

30) (0-35,000 lx)R

ES 100K OH

M 1/16W

1% 0402 SM

DR

ohm Sem

iconductor M

CR

01MZPF1003

RESC

1005LD

igi-KeyR

HM

100KLCT-N

D3

46R

13, R15

RES 20.0K O

HM

1/16W 1%

0402 SMD

Rohm

Semiconductor

MC

R01M

ZPF2002R

ESC1005L

Digi-Key

RH

M20.0KLC

T-ND

247

R11, R

12R

ES 820 OH

M 1/16W

1% 0402 SM

DVishay/D

aleC

RC

W0402820R

FKEDR

ESC1005L

Digi-Key

541-820LCT-N

D2

48R

14R

ES 2.70K OH

M 1/16W

1% 0402 SM

DR

ohm Sem

iconductorM

CR

01MZPF2701

RESC

1005LD

igi-KeyR

HM

2.70KLCT-N

D1

49R

17, R20

RES 3.30K O

HM

1/16W 1%

0402 SMD

Rohm

Semiconductor

MC

R01M

ZPF3301R

ESC1005L

Digi-Key

RH

M3.30KLC

T-ND

250

R1, R

7, R31, R

32, R33, R

34, R35

RES 0.0 O

HM

1/16W 5%

0402 SMD

Rohm

Semiconductor

MC

R01M

ZPJ000R

ESC1005L

Digi-Key

RH

M0.0JC

T-ND

751

R2, R

3, R4, R

5, R6, R

21, R22, R

25, R26

RES 10.0K O

HM

1/16W 1%

0402 SMD

Rohm

Semiconductor

MC

R01M

ZPF1002R

ESC1005L

Digi-Key

RH

M10.0KLC

T-ND

952

R23, R

27, R29

RES 56.0 O

HM

1/16W 1%

0402 SMD

Rohm

Semiconductor

MC

R01M

ZPF56R0

RESC

1005LD

igi-KeyR

HM

56.0LCT-N

D3

53R

8, R9, R

18, R19, R

36R

ES 560 OH

M 1/10W

1% 0603 SM

DR

ohm Sem

iconductorM

CR

03EZPFX5600R

ESC1608N

Digi-Key

RH

M560H

CT-N

D

554

R47, R

48R

ES 0.0 OH

M 1W

5% 2512 SM

DVishay/D

aleC

RC

W25120000Z0EG

2512D

igi-Key541-0.0XC

T-ND

255

RT1

CU

RR

ENT LIM

ITER IN

RU

SH 5 0H

M 20%

Cantherm

MF72-005D

7M

F72-005D7

Digi-Key

317-1148-ND

156

S1, S3SW

ITCH

TACT 6M

M 160G

F H=4.3M

MO

mron Electronics

B3S-1000PTH

Digi-Key

SW836C

T-ND

257

S2SW

ITCH

DIP H

ALF PITCH

2POS

CTS Electrocom

ponents218-2LPST

218-2LPSTD

igi-KeyC

T2182LPST-ND

158

U1

IC M

CU

32-BIT W/FLASH

48-LQFP

Texas Instruments

TMX320F28027PTA

TSQFP50P900X900X145-48L

Digi-Key

296-24123-ND

159

U2

IC 2-1LIN

E DATA SELEC

T/MU

X SM8

Texas Instruments

SN74LVC

2G157D

CTR

TSOP65P400X130-8L

Digi-Key

296-13266-1-ND

160

U3, U

4, U5, U

7, U8, U

9, U11

IC O

PAMP SN

GL R

-R I/O

LN SO

T23-5Analog D

evicesAD

8605ARTZ

SOT23-5

Digi-Key

AD8605AR

TZREEL7C

T-ND

761

U6, U

10, U12

IC PH

OTO

DETEC

TOR

AMBIEN

T 6-OD

FNIntersil

ISL29009IRO

Z-T7ISL29009IR

OZ-T7

Digi-Key

ISL29009IRO

Z-T7CT-N

D3

62U

13IC

USB TO

SERIAL U

ART 28-SSO

PFTD

IFT232R

L RSO

P65P780X200-28LD

igi-Key768-1007-1-N

D1

63U

14, U15, U

16, U18, U

19LED

DVR

BUC

KPLUS 500M

A 7SIP DIM

LEDdynam

ics Inc7021-D

-E-5007021-D

-E-500D

igi-Key788-1009-N

D5

64U

17SW

ITCH

ING

REG

ULATO

R, 5V, 1.0 A

Recom

Power Incorporated

R-78B5.0-1.0

THAllied

394-00521

65U

20IC

LDO

REG

3.3V 1A SOT223-6

Texas Instruments

TPS79633DC

QR

SOT127P706X180-6L

Digi-Key

296-13766-1-ND

1

No

tes

169

Ap

pro

ve

d

Lin

e Ite

m (3

6) E

q. P

/N: 3

M, 9

53

22

0-2

00

0-A

R-P

R; 3

M, 9

56

22

0-2

00

0-A

R-P

R; S

ullin

s N

PP

N1

02

GH

NP

-RC

(DIG

IKE

Y)

Lin

e Ite

m (6

5) A

ltern

ate

Su

pp

lier: M

ou

se

r, PN

: 59

5-T

PS

79

63

3D

CQ

Altiu

m L

imite

d C

on

fide

ntia

l4/29/2010

Page 2


102

B.4. Sensor Node Hardware Schematics

B.4 Sensor Node Hardware Schematics


103

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

GPIO29/SCITXD

A/SCLA/TZ3

1

TRST2

XRS3

ADCINA6/A

IO64

ADCINA4/CO

MP2A

/AIO4

5

ADCINA7

6

ADCINA3

7

ADCINA1

8

ADCINA2/CO

MP1A

/AIO2

9

ADCINA0/VREFHI

10

VDDA

11

VSSA/VREFLO

12

ADCINB1

13

ADCINB2/COM

P1B/AIO

1014

ASCINB3

15

ADCINB4/COM

P2B/AIO

1216

ADCINB6/A

IO1417

ADCINB7

18

GPIO34/CO

MP2OUT

19

GPIO35/TD

I20

GPIO36/TM

S21

GPIO37/TD

O22

GPIO38/XCLKIN/TCK

23

GPIO18/SPICLK

A/SCITXDA/XCLKO

UT24

GPIO19/XCLKIN/SPISTEA/SCIRXD

A/ECAP125

GPIO17/SPISOM

IA/TZ326

GPIO16/SPISIM

OA/TZ2

27

GPIO1/EPW

M1B/CO

MP1O

UT28

GPIO0/EPW

M1A

29

TEST30

GPIO32/SD

AA/EPWM

SYNCI/A

DCSOCAO

31

VDD

32

VSS33

VREGENZ

34

VDDIO

35

GPIO33/SCLA/EPW

MSYNCO/A

DCSOCBO36

GPIO2/EPW

M2A

37

GPIO3/EPW

M2B/CO

MP2O

UT38

GPIO4/EPW

M3A

39

GPIO5/EPM

W3B/ECA

P140

GPIO6/EPW

M4A/EPW

MSYN

CI/EPWM

SYNCO

41

GPIO7/EPW

M4B/SCIRXD

A42

VDD

43

VSS44

X145

X246

GPIO12/TZ1/SCITXDA

47

GPIO28/SCIRXD

A/SDAA/TZ2

48

U1TMX320F28027

12

34

56

78

910

1112

1314

P1Header 7X2

CTMS

CTDI

CPDCTD

OCTCK

-RETCTCKCEM

U0CEM

U1CG

NDCG

NDCG

ND

CGND

CTRSTn

123456789101112

P2Header 12

123456789101112

P3Header 12

123456789101112

P4Header 12

123456789101112

P5Header 12

CTRSTn

CTDO

CTDI

CTMS

CTCK

CGND

C22.2uF

L1EMI BEA

D

L2EMI BEA

D

C12.2uF

C101.0uF

C91.0uF

C7.47uF C3.47uF

L4EMI BEA

DC111.0uF

CGND

CGND

CGND

CGND

CGND

CGND

CGND

C3V3INC3V3D

C3V3A

CRSTR70

CGND

TI JTAG

HEA

DER

MCU FA

NO

UT

CRX-SLAVE

CTX-SLAVE

CSDA

CSCL

CTAOSINPU

T

CGPIO

7

PUC

K P1

CSYNC

CINT

CS3CS2TAO

SOE

R2TBD 0402

R3TBD 0402

CS0CS1

C3V3INC3V3IN

TAOS PU

LLUPS

A1

B2

Y3

GND

4

Y5

A/B6

G7

VCC8

U2SN74LVC2G157C24

100nFCG

ND

C3V3IN

CTEMP

CMSO

CADCLU

XL

CAGND

C12TBD

CGND

R14TBD

0402

CRSTS1B3S-1000

CGND

BTN RST

CFUNC

R1TBD 0402

CGND

TMP05 SETTING

S

CMSO

OPEN

CADCLU

XH

CTEMP

R8TBD 0603

D1HSMH-C190

CGND

CSTATUS

1.8 0603 RED

CSTATUS

Float: Cont / GND: O

ne-Shot

CSDA

CSCL

R510K

C3V3IN

R610K

C3V3IN

R410K

C3V3IN

CIRTXBRST

D3TSAL6200

CGND R10TBD

0805

IR TX

C3V3D

Q2NDS355AN

C3V3IN

Q1NDS355AN

C3V3IN

1243

NO

S2218-2LPST

R20TBD

0402R17TBD

04023.3K

820R12TBD

0402R11TBD

0402

CGND

CGND

C3V3INC3V3IN

CGPIO

34

CGPIO

34CTD

O

MCU BO

OT SETTIN

GS

LED

R13TBD

0402

CGND

OPTIO

N

R15TBD

0402

CGND

OPTIO

N

MUX_TA

OS_TEMP

CTX-SLAVE

CTRSTnCRST

ADCINA4

ADCINA0

ADCINB1

ADCINB3

ADCINB7

ADCINA4

CADCLU

XHCA

DCLUXL

CADCLU

XNOFLT

ADCINA0

C3V3ACAGND

ADCINB1

CSYNC

CSTATUS

ADCINB3

CMSO

ADCINB7

CGPIO

34

CTDO

CTDI

CTMS

CTCKCS3

MUX_TA

OS_TEMP

CS2TAO

SOE

CSDA

CGND

C3V3D

CSCL

R9TBD 0603

D2HSMH-C190

CGND

C3V3IN

CIRTXBRSTCG

PIO7

CGND

CTEMP

CRX-SLAVE

COLO

R / TEMP M

UX

CGND C810nF CG

ND C410nFC52.2pF

CGND

C61.0uF

CGND

CADCLU

XNOFLT OPEN

OPEN

OPEN

OPEN

S01

S12

OE3

GND

4VD

D5

OUT

6

S27

S38

U14

TCS32X0

SCLA1

SYNC

A2

GND

A3

SDAB1

VDD

B2

INTB3

U15

TCS34X4CS

C55

TBD 0603

C56

TBD 0603

CGND

CGND

VDD

5GN

D4

IN2

FUNC

3TEM

P1

U16

TMP05AK

SZ

C48

TBD 0603

CGND

Color to Frequency

I2C Color Sensor

PWM

Temp Sensor

CS2CS3

TAOSOE

CSCLCSD

A

CSYNC

CINT

CFUNC

CINCTEM

P

CS1CS0

CTAOSINPU

T

C3V3IN

C3V3IN

C3V3IN

CL2

F4.63 F2.8 LENS

CIRTXBRST

LCRITCRI

ADCCO

LORAD

CINTNSTY

CIRRX

MOTION

ADCCO

LOR

ADCINTNSTY

LCRITCRI

MOTION

CIRRX

GPIO5

OPEN

OPEN

GPIO3

GPIO5

OPEN

CC1

OPAQUE D

OME

CL1

F4.63 F2.8 LENS

CC2

OPAQUE D

OME

CC3

OPAQUE D

OME

CINT

PIC101 PIC102COC1

PIC201 PIC202COC2

PIC301 PIC302COC3

PIC401 PIC402COC4

PIC501 PIC502COC5


COC7PIC801 PIC802COC8PIC901 PIC902

COC9PIC1001 PIC1002

COC10



PIC2401PIC2402

COC24

PIC4801PIC4802 C

OC48

PIC5501PIC5502 C

OC55

PIC5601PIC5602

COC56

COCC1

COCC2

COCC3

COCL1COCL2

PID101PID102COD1

PID201PID202COD2

PID301 PID302COD3

PIL101PIL102

COL1

PIL201PIL202

COL2

PIL401PIL402

COL4

PIP101

PIP102

PIP103

PIP104

PIP105

PIP106

PIP107

PIP108

PIP109

PIP1010

PIP1011PIP1012

PIP1013

PIP1014

COP1

PIP201

PIP202

PIP203

PIP204

PIP205

PIP206

PIP207

PIP208

PIP209

PIP2010

PIP2011

PIP2012 COP2

PIP301

PIP302

PIP303

PIP304

PIP305

PIP306

PIP307

PIP308

PIP309

PIP3010

PIP3011

PIP3012 COP3

PIP401

PIP402

PIP403

PIP404

PIP405

PIP406

PIP407

PIP408

PIP409

PIP4010

PIP4011

PIP4012 COP4

PIP501

PIP502

PIP503

PIP504

PIP505

PIP506

PIP507

PIP508

PIP509

PIP5010

PIP5011

PIP5012 COP5

PIQ101

PIQ102 PIQ103COQ1

PIQ201

PIQ202 PIQ203COQ2

PIR101PIR102 COR1

PIR201 PIR202COR2

PIR301 PIR302COR3

PIR401 PIR402COR4

PIR501 PIR502COR5

PIR601 PIR602COR6

PIR701 PIR702COR7

PIR801PIR802 COR8PIR901PIR902 COR9









PIS101

PIS102

PIS103

PIS104

COS1PIS201

PIS202

PIS203

PIS204

COS2

PIU101

PIU102

PIU103

PIU104

PIU105

PIU106

PIU107

PIU108

PIU109

PIU1010

PIU1011

PIU1012

PIU1013

PIU1014

PIU1015

PIU1016

PIU1017

PIU1018

PIU1019

PIU1020

PIU1021

PIU1022

PIU1023

PIU1024

PIU1025

PIU1026

PIU1027

PIU1028

PIU1029

PIU1030

PIU1031

PIU1032

PIU1033

PIU1034

PIU1035

PIU1036

PIU1037

PIU1038

PIU1039

PIU1040

PIU1041

PIU1042

PIU1043

PIU1044

PIU1045

PIU1046

PIU1047

PIU1048

COU1

PIU201

PIU202

PIU203

PIU204

PIU205

PIU206

PIU207

PIU208 COU2

PIU1401

PIU1402

PIU1403

PIU1404PIU1405

PIU1406

PIU1407

PIU1408 COU14

PIU150A1

PIU150A2

PIU150A3

PIU150B1

PIU150B2

PIU150B3

COU15

PIU1601PIU1602

PIU1603

PIU1604PIU1605 C

OU16

PIP204

PIU106

NLADCCOLOR

PIP2010

PIU1010NLADCINA0

PIP205

PIU105

NLADCINA4

PIP3012

PIU1013NLADCINB1

PIP3010

PIU1015

NLADCINB3

PIP307

PIU1018

NLADCINB7

PIP206

PIU104

NLADCINTNSTY

PIC602PIC702

PIC802PIL202

PIP2011

PIU1011

PIC302PIC402

PIC502PIL102

PIP402

PIR1402

PIU1035

PIC102PIC202

PIC2401

PIC4802

PIC5502

PIC5601

PID101PID201

PID302PIL101

PIL201

PIR202PIR302

PIR402PIR502

PIR602

PIR1102PIR1202

PIU208

PIU1405

PIU150B2

PIU1605

PIP209

PIU109

NLCADCLUXH

PIP208

PIU108

NLCADCLUXL

PIP207

PIU107

NLCADCLUXNOFLT

PIC1102PIL402

PIP2012

PIU1012

PIP1013

NLCEMU0PIP1014

NLCEMU1

PIR102

PIU1603NLCFUNC

PIC101PIC201

PIC301PIC401

PIC501

PIC601PIC701

PIC801

PIC901PIC1001

PIC1101

PIC1201

PIC2402

PIC4801

PIC5501

PIC5602

PIL401

PIP104

PIP108

PIP1010

PIP1012

PIP404

PIP508

PIQ102

PIQ202

PIR101

PIR701

PIR902

PIR1301

PIR1501

PIR1701PIR2001

PIS102

PIU1033

PIU1044

PIU204

PIU207

PIU1404PIU150A3

PIU1604NLCGND

PIP506

PIU1042

NLCGPIO7PIP306

PIR1702

PIS201

PIU1019

NLCGPIO34

PIU1602NLCIN

PIR401

PIU1038

PIU150B3NLCINT

PIP503

PIU1039

NLCIRRX

PIP505

PIQ201

PIU1041

NLCIRTXBRST

PIP308

PIU1017

PIU206 NLCMSO

PIP105

NLCPD

PIC1202

PIP203

PIR1401PIS104

PIU103

NLCRST

PIP5012

PIU1048

NLCRX0SLAVE

PIR201PIU1401

NLCS0PIR301

PIU1402NLCS1

PIP4011

PIU1026

PIU1407NLCS2

PIP301

PIU1024

PIU1408NLCS3

PIP401

PIR601

PIU1036

PIU150A1NLCSCL

PIP406

PIR501

PIU1031

PIU150B1NLCSDA

PIP3011

PIQ101

PIU1014NLCSTATUS

PIP309

PIU1016

PIU150A2NLCSYNC

PIU201

PIU1406NLCTAOSINPUT

PIP1011

PIP302 PIU1023

NLCTCKPIP109

NLCTCK0RET PIP103

PIP305 PIU1020

NLCTDI

PIP107

PIP303

PIR1502

PIR2002

PIS202

PIU1022

NLCTDO

PIP5011

PIU1047

PIU202

PIU1601NLCTEMP

PIP101

PIP304 PIU1021

NLCTMSPIP102

PIP202

PIR1302

PIU102

NLCTRSTn

PIP201

PIU101

NLCTX0SLAVE

PIP502

NLGPIO3

PIP504

PIU1040

NLGPIO5

PIP409

PIU1028NLLCRI

PIP501

PIU1037

NLMOTION

PIP4012

PIU1025PIU205

NLMUX0TAOS0TEMP

PIS103

PIS101

PIC902

PIU1032

PIC1002

PIU1043


PID301PIR1002

PIP106

PIP403

PIP405

PIP407

PIP507

PIP509

PIP5010

PIQ103 PIR802

PIQ203 PIR1001

PIR702PIU1034

PIR1101PIS204

PIR1201PIS203

PIU1030

PIU1045

PIU1046

PIU203

PIP4010

PIU1027

PIU1403NLTAOSOE

PIP408

PIU1029NLTCRI


104

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

PUC

K P2

4 31

52

U3AD8605ARTZ

VDD

1

GND

2

NC3

EN4

NC5

IOUT6

U6ISL29009IROZ-T7

C19100nF

R24

TBD 0402

C1580pF

CAGND

LUX SEN

SOR LO

W RA

NG

E

LUX SEN

SOR H

IGH

RAN

GE

CAGND

4 31

52

U4AD8605ARTZ

C17

100nF

CAGND

CADCLU

XLR23

TBD 0402

C221nF

CAGND

L5308nHC1480pF

CAGND

CAGND

C3V3A

4 31

52

U5AD8605ARTZ

C20TBD

C21TBD

R21

TBD 0402

R22

TBD 0402

CAGND

C3V3OP

C3V3OP

CAGND

C3V3OP

4 31

52

U7AD8605ARTZ

VDD

1

GND

2

NC3

EN4

NC5

IOUT6

U10

ISL29009IROZ-T7

C29100nF

R28

TBD 0402

CAGND

CAGND

4 31

52

U8AD8605ARTZ

C27

100nF

CAGND

CADCLU

XHR27

TBD 0402

C321nF

CAGND

C23

100nF

CAGND

4 31

52

U9AD8605ARTZ

C30TBD

C31TBD

R25

TBD 0402

R26

TBD 0402

CAGND

C3V3OP

C3V3OP

CAGND

C3V3OP

C13

100nF

CAGND

C18TBD

0603

C28TBD

0603

CAGND

CAGND

CAGND

CAGND

CAGND

CAGND

CAGND

CAGND

C3V3A

C3V3A

OPTIO

N

OPTIO

N

VDD

1

GND

2

NC3

EN4

NC5

IOUT6

U12

ISL29009IROZ-T7

C35100nF

C34TBD

0603

CAGND

CAGND

CAGND

C3V3A

R30TBD

0402

CAGND

CAGND

4 31

52

U11AD

8605ARTZ

C33

100nF

CAGND

CADCLU

XNOFLT

R29

TBD 0402

C361nF

CAGND

C3V3OP

OPTIO

N

HI SPEED

LUX

C16

.1uF

C26

.1uF

PI FILTER OPTION 1

25MHz

OPTIO

N

OR

L5: 100MHz220 O

hm

C14, C15: 80pFL5: 308nH

PI FILTER OPTION 2

3.5MHz

C14, C15: 572pFL5: 2.2uH

OR

PIC1301PIC1302

COC13



PIC1601PIC1602

COC16

PIC1701PIC1702

COC17



PIC2001PIC2002COC20

PIC2101PIC2102COC21


PIC2301PIC2302

COC23

PIC2601PIC2602

COC26

PIC2701PIC2702

COC27



PIC3001PIC3002COC30

PIC3101PIC3102COC31


PIC3301PIC3302

COC33




PIL501PIL502

COL5

PIR2101PIR2102 C

OR21

PIR2201PIR2202 C

OR22

PIR2301PIR2302

COR23


PIR2501PIR2502 CO

R25

PIR2601PIR2602 CO

R26

PIR2701PIR2702

COR27


PIR2901PIR2902

COR29


PIU301

PIU302

PIU303

PIU304

PIU305COU3

PIU401

PIU402

PIU403

PIU404

PIU405COU4

PIU501

PIU502

PIU503

PIU504

PIU505COU5

PIU601

PIU602

PIU603

PIU604

PIU605

PIU606

COU6

PIU701

PIU702

PIU703

PIU704

PIU705COU7

PIU801

PIU802

PIU803

PIU804

PIU805COU8

PIU901

PIU902

PIU903

PIU904

PIU905COU9

PIU1001

PIU1002

PIU1003PIU1004

PIU1005

PIU1006

COU10

PIU1101

PIU1102

PIU1103

PIU1104

PIU1105COU11

PIU1201

PIU1202

PIU1203PIU1204

PIU1205

PIU1206

COU12 PIC1402

PIC1802PIC1902

PIC2802PIC2902

PIC3402PIC3502

PIL501

PIU601

PIU1001

PIU1201

PIC1301

PIC1502

PIC1601

PIC1701

PIC2301

PIC2601

PIC2701

PIC3301

PIL502

PIU303

PIU305PIU405

PIU505

PIU703

PIU705PIU805

PIU905

PIU1105

PIC3202PIR2702

NLCADCLUXH

PIC2202PIR2302

NLCADCLUXL

PIC3602PIR2902

NLCADCLUXNOFLT

PIC1302PIC1401

PIC1501

PIC1602

PIC1702

PIC1801PIC1901

PIC2102PIC2201

PIC2302

PIC2602

PIC2702

PIC2801PIC2901

PIC3102PIC3201

PIC3302

PIC3401PIC3501

PIC3601PIR3002

PIU302PIU402

PIU502PIU602

PIU604

PIU702PIU802

PIU902PIU1002

PIU1004

PIU1102PIU1202

PIU1204

PIC2001

PIR2101PIR2202

PIC2002

PIU403

PIU501

PIU504

PIC2101

PIR2201PIU503

PIC3001

PIR2501PIR2602

PIC3002

PIU803

PIU901

PIU904

PIC3101

PIR2601PIU903

PIR2102

PIR2401

PIU301

PIR2301PIU401

PIU404

PIR2402

PIU304

PIU606

PIR2502

PIR2801

PIU701

PIR2701PIU801

PIU804

PIR2802

PIU704

PIU1006

PIR2901PIU1101

PIU1104PIR3001

PIU1103

PIU1206

PIU603

PIU605

PIU1003

PIU1005

PIU1203

PIU1205


105

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

TXD1

DTR2

RTS3

VCCIO4

RXD5

RI6

GND

7

NC8

DSR9

DCD10

CTS11

CBUS4

12

CBUS2

13

CBUS3

14

USBDP15

USBDM16

3V3OUT

17

GND

18

RST19

VCC20

GND

21

CBUS1

22CBU

S023

NC24

AGND

25

TEST26

OSCI27

OSCO28

U13

FT232RL

1234

SH5

J161729-1010BLFC37

10nF

CGND

L7 EMI BEA

D

C25100nF

CGND

CGND

CGND

R35

TBD 0402

CVCCIO

CVCCIO

R320

R340

CRX-SLAVE

CTX-SLAVE

D10HSM

G-C190

2.2V 0603 GRN

D11HSM

H-C190

1.8 0603 RED

R18TBD

0603R19TBD

0603

CVCCIOCVCCIO

USB / SERIAL

11

22

33

44

55

66

77

88

99

1010

1111

1212

1313

1414

1515

1616

1717

1818

1919

2020

P691596-120LF

11

22

33

44

55

66

77

88

99

1010

1111

1212

1313

1414

1515

1616

1717

1818

1919

2020

P791596-120LF

ZIGBEE IN

TERFACE

CRX-SLAVE

CTX-SLAVE

CDIO

12CRSTCRSSICD

IO11

CDTR*

CDIO

4CCTS*CSLEEP*CG

NDCA

SSOCCRTS*CD

IO3

CDIO

2CD

IO1

CCOM

MIS

CGND

C3V3INR31

0

R330

R36TBD

0603

D4HSMH-C190

CGND

CASSOC

1.8 0603 RED

PUC

K P3

C3830pF

C3930pF

CGND

CGND

S3B3S-1000CCO

MM

IS

CGND

COM

MISSIO

N / A

SSOCIA

TE

FIX PIN

S 2 -> 1


PIC3701PIC3702 C

OC37


COC39

PID401PID402COD4

PID1001PID1002COD10

PID1101PID1102COD11

PIJ101

PIJ102

PIJ103

PIJ104

PIJ105

COJ1PIL701

PIL702

COL7PIP601

PIP602

PIP603

PIP604

PIP605

PIP606

PIP607

PIP608

PIP609

PIP6010

PIP6011

PIP6012

PIP6013

PIP6014

PIP6015PIP6016

PIP6017PIP6018

PIP6019

PIP6020

COP6PIP701

PIP702

PIP703

PIP704

PIP705

PIP706

PIP707

PIP708

PIP709

PIP7010

PIP7011

PIP7012

PIP7013

PIP7014

PIP7015PIP7016

PIP7017PIP7018

PIP7019

PIP7020

COP7





PIR3301PIR3302 CO

R33




PIS301

PIS302

PIS303

PIS304

COS3

PIU1301

PIU1302

PIU1303

PIU1304

PIU1305

PIU1306

PIU1307

PIU1308

PIU1309

PIU13010

PIU13011

PIU13012

PIU13013

PIU13014

PIU13015

PIU13016

PIU13017

PIU13018

PIU13019

PIU13020

PIU13021

PIU13022

PIU13023

PIU13024

PIU13025

PIU13026

PIU13027

PIU13028 COU13

PIP602

PIP7011

PIR3601

NLCASSOC

PIP701

PIS304

NLCCOMMISPIP7017NLCCTS0

PIP703

NLCDIO1

PIP705

NLCDIO2

PIP707

NLCDIO3

PIP7019

NLCDIO4

PIP6014 NLCDIO11

PIP608 NLCDIO12

PIP6018 NLCDTR0

PIC2501

PIC3702

PIC3801PIC3901

PID402

PIJ104

PIP6020

PIP7013

PIS302

PIU1307

PIU13018

PIU13021

PIU13025

PIU13026

NLCGNDPIP6012 NLCRSSI

PIP6010 NLCRST

PIP709

NLCRTS0

PIR3101

PIR3201

NLCRX0SLAVE

PIP7015NLCSLEEP0

PIR3301

PIR3401

NLCTX0SLAVE

PIC2502

PID1001PID1101

PIR3501

PIU1304

PIU13017

NLCVCCIO

PIS303

PIS301

PIC3701

PIJ101

PIL701PIC3802

PIJ102PIU13016

PIC3902

PIJ103

PIU13015

PID401 PIR3602


PIJ105

PIL702PIU13020

PIP601

PIP603

PIP604

PIR3102

PIP605

PIP606

PIR3302

PIP607

PIP609

PIP6011

PIP6013

PIP6015PIP6016

PIP6017

PIP6019

PIP702

PIP704

PIP706

PIP708

PIP7010

PIP7012

PIP7014

PIP7016

PIP7018

PIP7020

PIR1801

PIU13023

PIR1901

PIU13022

PIR3202

PIU1301

PIR3402PIU1305

PIR3502PIU13014

PIU1302

PIU1303

PIU1306

PIU1308

PIU1309

PIU13010

PIU13011

PIU13012

PIU13013

PIU13019

PIU13024

PIU13027

PIU13028


106

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

PUC

K P4VLEDSUPPLY

LEDRTN

VIN1

2 VOUT

3

GND

U17R-78B5.0-1.0

24VREG

CGND

CGND

CGND

5VREF

CGND

EN1

IN2

GND

3

OUT

4

NR/FB5

GND

6

U20

TPS79633

CGND

5VREFC432.2uF 6.3V X7R

C442.2uF 6.3V X7R

C452.2uF 6.3V X7R

CGND

C4610nF

CGND

C3V3IN

R48

0LED

RTNCG

ND

GN

D N

ET TIE

R47

0

CTNT M

ES

1 23J22.1mm ID x 5mm OD

LEDRTN

VLEDSUPPLY

6 - 30V IN

/ 5V O

UT BUCK

5V / 3.3V

LINEA

R

DESK

TOP PO

WER EN

TRY

F133V / 1.85A@25C

24VREG

t° RT1

R40RS6011Y

A6009

S4

101-TS5511T26002R-EV

C3V3IN

C3V3A

CAGND

VS3

OUT

1

GND

2

IR RX

U18

TSOP38338C49TBD

0603 R38

TBD 0603

CGND

C3V3IN

CIRRX

IR RXUSER SLID

ER CON

TROL

CRI CON

TROL

CAGND

4 31

52U19AD

8605ARTZ

C50

100nF

CAGND

ADCCO

LORR41

56

C513.3nF

CAGND

C3V3OP

OPTIO

N

R43RS6011Y

A6009

C3V3A

CAGND

CAGND

4 31

52

U22AD

8605ARTZ

C52

100nF

CAGND

ADCINTNSTY

R44

56

C543.3nF

CAGND

C3V3OP

OPTIO

N

CGND

C3V3IN

Q3NDS355AN

CGND

LCRI

TCRI

C4747uF 35V

R16

TBD 0603

OUT

1

VDD

2

GND

3

U21

AMN11112

MO

TION

DETECTIO

N

C53TBD

0603

CGND

C3V3IN

MOTION

R42

TBD 0603

C4022UF 25V X5R 1210

C4122UF 25V X5R 1210

C4222UF 25V X5R 1210

R37TBD

0603

R39

TBD 0603

1uF

100 O




PIC4301PIC4302COC43




PIC4701PIC4702COC47


PIC5001PIC5002

COC50


PIC5201PIC5202

COC52



PIF101

PIF102

COF1PIJ201

PIJ202

PIJ203

COJ2

PIQ301

PIQ302 PIQ303COQ3

PIR1601PIR1602

COR16


PIR3801PIR3802

COR38


PIR4001PIR4002

PIR4003

COR40

PIR4101PIR4102

COR41

PIR4201PIR4202

COR42

PIR4301PIR4302

PIR4303

COR43

PIR4401PIR4402

COR44

PIR4701PIR4702

COR47

PIR4801PIR4802

COR48

PIRT101PIRT102

CORT1

PIS401

PIS402

PIS403

PIS404

PIS405

PIS406 COS4

PIU1701

PIU1702

PIU1703

COU17

PIU1801

PIU1802

PIU1803

COU18

PIU1901

PIU1902

PIU1903

PIU1904

PIU1905COU19

PIU2001

PIU2002

PIU2003

PIU2004

PIU2005

PIU2006 COU20

PIU2101

PIU2102

PIU2103

COU21

PIU2201

PIU2202

PIU2203

PIU2204

PIU2205COU22

PIC4102PIC4202

PIC4301

PIU1703

PIU2001

PIU2002

PIC4701

PIF102

PIRT101

PIC5102PIR4102

NLADCCOLOR

PIC5402PIR4402

NLADCINTNSTY

PIR4001PIR4301PIC5302

PIR3702

PIR3802

PIR4702

PIS406

PIU2102

PIC5001

PIC5201

PIU1905PIU2205

PIC5002

PIC5101

PIC5202

PIC5401

PIR4003PIR4303

PIU1902PIU2202PIC4001

PIC4101PIC4201

PIC4302PIC4401

PIC4501PIC4601

PIC4901

PIC5301

PIQ302

PIR4802

PIS403

PIS404

PIU1702

PIU1802

PIU2003

PIU2006

PIU2103

PIR1602NLCIRRX

PIQ301NLLCRI

PIC4702

PIJ202

PIJ203

PIR4801 PIR4202NLMOTION

PIC4002

PIRT102PIU1701

PIC4402PIC4502

PIR4701PIU2004

PIC4602PIU2005

PIC4902PIR3801

PIU1803

PIQ303

PIR3901

PIR1601

PIU1801

PIR3902PIS405

PIR4002PIU1903

PIR4101PIU1901

PIU1904

PIR4201

PIU2101

PIR4302PIU2203

PIR4401PIU2201

PIU2204

PIR3701PIS401

PIS402

NLTCRI

PIF101

PIJ201


107

B.5. Sensor Node PCB

B.5 Sensor Node PCB


108

PAC102PAC101

COC1PAC202PAC201

COC2

PAC302PAC301

COC3PAC402

PAC401COC4

PAC502PAC501

COC5

PAC602PAC601

COC6PAC702PAC701

COC7PAC802PAC801

COC8

PAC902PAC901

COC9

PAC1001PAC1002COC10

PAC1102

PAC1101

COC11

PAC1202PAC1201

COC12

PAC1302PAC1301

COC13

PAC1401

PAC1402COC14

PAC1502

PAC1501 COC15PAC1602

PAC1601COC16

PAC1702PAC1701

COC17

PAC1801

PAC1802

COC18PAC1902

PAC1901

COC19

PAC2002

PAC2001

COC20 PAC2102

PAC2101

COC21

PAC2202

PAC2201

COC22

PAC2302

PAC2301

COC23

PAC2402PAC2401

COC24

PAC2502PAC2501

COC25

PAC2602PAC2601

COC26

PAC2702PAC2701

COC27

PAC2801

PAC2802COC28

PAC2902 PAC2901

COC29

PAC3002

PAC3001

COC30 PAC3102

PAC3101

COC31

PAC3202

PAC3201

COC32 PAC3302PAC3301 COC33

PAC3401

PAC3402

COC34PAC3502 PAC3501

COC35

PAC3602PAC3601

COC36

PAC3702PAC3701

COC37PAC3802

PAC3801

COC38PAC3902PAC3901

COC39PAC4002

PAC4001

COC40

PAC4101

PAC4102

COC41

PAC4201

PAC4202

COC42

PAC4302PAC4301COC43

PAC4402PAC4401

COC44

PAC4502PAC4501

COC45

PAC4602PAC4601

COC46

PAC4702PAC4701

COC47

PAC4801PAC4802

COC48

PAC4901PAC4902

COC49

PAC5001PAC5002

COC50

PAC5101PAC5102

COC51

PAC5201PAC5202

COC52

PAC5302

PAC5301COC53

PAC5401

PAC5402

COC54

PAC5502

PAC5501 COC55

PAC5602

PAC5601

COC56

PACC10DH

COCC1

PACC20DH

COCC2

PACC30DH

COCC3

PACL10DHCOCL1

PACL20DHCOCL2

PAD102

PAD101

COD1

PAD202

PAD201

COD2

PAD301

PAD302

COD3

PAD402

PAD401

COD4

PAD1002PAD1001

COD10PAD1102PAD1101

COD11PAF102

PAF101COF1

PAJ105PAJ104

PAJ103PAJ102

PAJ101COJ1

PAJ202PAJ201 PAJ203

COJ2

PAL101PAL102

COL1

PAL201PAL202

COL2

PAL402

PAL401

COL4

PAL501PAL502

COL5 PAL701

PAL702COL7

PAP101

PAP102

PAP103

PAP104

PAP105

PAP106

PAP107

PAP108

PAP109

PAP1010

PAP1011

PAP1012

PAP1013

PAP1014COP1

PAP201

PAP202

PAP203

PAP204PAP205

PAP206

PAP207

PAP208PAP209

PAP2010

PAP2011

PAP2012

COP2

PAP301PAP302

PAP303PAP304

PAP305PAP306

PAP307PAP308

PAP309PAP3010

PAP3011PAP3012

COP3

PAP401

PAP402

PAP403

PAP404PAP405

PAP406

PAP407

PAP408PAP409

PAP4010

PAP4011

PAP4012

COP4PAP501

PAP502PAP503

PAP504PAP505

PAP506PAP507

PAP508PAP509

PAP5010PAP5011

PAP5012

COP5

PAP60U1PAP601

PAP603

PAP605

PAP607

PAP609

PAP6011

PAP6013

PAP6015

PAP6017

PAP6019

PAP602

PAP604

PAP606

PAP608

PAP6010

PAP6012

PAP6014

PAP6016

PAP6018

PAP6020

COP6

PAP70U1PAP701

PAP703

PAP705

PAP707

PAP709

PAP7011

PAP7013

PAP7015

PAP7017

PAP7019

PAP702

PAP704

PAP706

PAP708

PAP7010

PAP7012

PAP7014

PAP7016

PAP7018

PAP7020

COP7

PAQ103

PAQ102PAQ101

COQ1

PAQ201

PAQ202PAQ203

COQ2

PAQ303PAQ302

PAQ301COQ3

PAR101PAR102COR1

PAR201PAR202

COR2 PAR301PAR302

COR3

PAR401PAR402

COR4

PAR501PAR502

COR5

PAR601PAR602

COR6

PAR701PAR702COR7

PAR801

PAR802

COR8

PAR901

PAR902

COR9

PAR1002PAR1001

COR10

PAR1101PAR1102COR11

PAR1201PAR1202COR12

PAR1301PAR1302

COR13

PAR1401PAR1402

COR14

PAR1501PAR1502

COR15

PAR1602PAR1601COR16

PAR1701PAR1702

COR17

PAR1801PAR1802

COR18PAR1901

PAR1902

COR19

PAR2001PAR2002

COR20

PAR2101PAR2102

COR21

PAR2201

PAR2202

COR22PAR2301

PAR2302

COR23PAR2402

PAR2401COR24

PAR2501PAR2502

COR25

PAR2601

PAR2602COR26PAR2701

PAR2702

COR27PAR2801

PAR2802COR28

PAR2901PAR2902

COR29

PAR3001PAR3002COR30

PAR3102PAR3101COR31

PAR3201

PAR3202

COR32

PAR3302PAR3301COR33

PAR3401

PAR3402

COR34PAR3501

PAR3502

COR35

PAR3601

PAR3602

COR36

PAR3701

PAR3702

COR37

PAR3801PAR3802COR38

PAR3901PAR3902

COR39

PAR400LPAR4003 PAR4002

PAR400MH PAR4001

COR40

PAR4101PAR4102

COR41PAR4201

PAR4202COR42


PAR430MH PAR4301

COR43

PAR4401PAR4402

COR44

PAR4701

PAR4702

COR47

PAR4801PAR4802COR48

PART101

PART102

CORT1

PAS102

PAS104PAS103

PAS101

COS1

PAS201PAS204

PAS202PAS203

COS2

PAS302

PAS304PAS303

PAS301

COS3

PAS403

PAS406PAS402

PAS404PAS405

PAS401

COS4


PAU1035

PAU1034

PAU1033

PAU1032PAU1031

PAU1030

PAU1029

PAU1028

PAU1027PAU1026

PAU1025




PAU1013PAU1012

PAU1011

PAU1010PAU109

PAU108

PAU107

PAU106

PAU105PAU104

PAU103

PAU102

PAU101

COU1

PAU201

PAU202

PAU203PAU204

PAU208

PAU207

PAU206PAU205

COU2

PAU301

PAU302PAU303

PAU304

PAU305

COU3

PAU401

PAU402PAU403

PAU404

PAU405

COU4

PAU501

PAU502PAU503

PAU504

PAU505

COU5

PAU60HSPAU604PAU605

PAU603PAU602

PAU606PAU601 COU6

PAU701

PAU702PAU703

PAU704

PAU705

COU7

PAU801

PAU802PAU803

PAU804

PAU805

COU8

PAU901

PAU902PAU903

PAU904

PAU905

COU9

PAU100HSPAU1004PAU1005

PAU1003PAU1002

PAU1006PAU1001

COU10

PAU1101PAU1102PAU1103 PAU1104PAU1105

COU11


PAU1203PAU1202

PAU1206PAU1201

COU12

PAU13015PAU13016PAU13017

PAU13018PAU13019

PAU13020PAU13021

PAU13022PAU13023PAU13024

PAU13025PAU13026

PAU13027PAU13028

PAU13014PAU13013PAU13012

PAU13011PAU13010

PAU1309PAU1308


PAU1304PAU1303

PAU1302PAU1301

COU13



COU14

PAU150A1PAU150A2PAU150A3

PAU150B1PAU150B2PAU150B3

COU15

PAU1601PAU1602


COU16


COU17

PAU1801

PAU1803

PAU1802

COU18


COU19

PAU2006


PAU2101PAU2102

PAU2103

COU21


PAU2202PAU2201

COU22

PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAF102 PART101

PAC5102

PAP204

PAR4102

PAU106

PAP2010

PAU1010

PAP205

PAU105

PAP3012

PAU1013

PAP3010

PAU1015PAP307 PAU1018

PAC5402

PAP206

PAR4402

PAU104

PAC602PAC702

PAC802

PAC1402

PAC1802PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501PAP2011

PAR4001PAR4301

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102

PAP402

PAR1402

PAU1035

PAC102PAC202

PAC2401

PAC4802

PAC5302

PAC5502

PAC5601

PAD101PAD201

PAD302

PAL101

PAL201

PAP602

PAR202PAR302

PAR402PAR502

PAR602

PAR1102PAR1202

PAR3702

PAR3802

PAR4702

PAS406

PAU208

PAU1405

PAU150B2

PAU1605

PAU2102

PAC1301

PAC1502

PAC1601PAC1701

PAC2301

PAC2601PAC2701

PAC3301

PAC5001

PAC5201

PAL502

PAU303

PAU305PAU405

PAU505

PAU703

PAU705PAU805

PAU905

PAU1105 PAU1905PAU2205

PAC3202

PAP209

PAR2702

PAU109

PAC2202

PAP208

PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602PAC1702

PAC1801PAC1901

PAC2102

PAC2201

PAC2302

PAC2602PAC2702

PAC2801PAC2901

PAC3102

PAC3201 PAC3302

PAC3401PAC3501

PAC3601

PAC5002

PAC5101

PAC5202

PAC5401

PAL402

PAP2012

PAR3002

PAR4003PAR4303

PAU1012

PAU302PAU402

PAU502

PAU602PAU604

PAU702PAU802

PAU902

PAU1002PAU1004

PAU1102

PAU1202

PAU1204

PAU1902PAU2202

PAP7011

PAR3601

PAP701

PAS304

PAP7017

PAP703

PAP705

PAP707

PAP7019

PAP6014

PAP608

PAP6018

PAP1013

PAP1014

PAR102PAU1603

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAC4801 PAC4901

PAC5301

PAC5501

PAC5602

PAD402

PAJ104

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6020

PAP7013

PAQ102

PAQ202

PAQ302

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701PAR2001

PAR4802PAS102

PAS302

PAS403PAS404

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1404

PAU150A3

PAU1604

PAU1702

PAU1802

PAU2003

PAU2006

PAU2103PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAU1602

PAR401PAU1038

PAU150B3

PAP503

PAR1602

PAU1039

PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAS101PAS301

PAP105

PAP6012

PAC1202

PAP203

PAP6010

PAR1401

PAS104

PAU103

PAP709

PAP5012

PAR3101

PAR3201

PAU1048

PAR201

PAU1401PAR301

PAU1402

PAP4011

PAU1026

PAU1407PAP301

PAU1024

PAU1408

PAP401PAR601

PAU1036

PAU150A1

PAP406PAR501

PAU1031

PAU150B1

PAP7015

PAP3011

PAQ101

PAU1014

PAP309

PAU1016

PAU150A2

PAU201

PAU1406

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAP5011

PAU1047

PAU202

PAU1601

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502PAP504

PAU1040

PAP409

PAQ301

PAU1028

PAC4702PAJ202

PAJ203PAR4801

PAP501

PAR4202

PAU1037

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043

PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501

PAU504PAC2101

PAR2201PAU503

PAC3001

PAR2501

PAR2602

PAC3002

PAU803

PAU901

PAU904PAC3101

PAR2601PAU903

PAC3701

PAJ101

PAL701

PAC3802


PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAC4902

PAR3801PAU1803

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002

PAR1802

PAD1102

PAR1902

PAL702

PAU13020

PAP604PAR3102

PAP606PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ303PAR3901

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1601PAU1801

PAR1801PAU13023

PAR1901PAU13022

PAR2102PAR2401 PAU301

PAR2301PAU401

PAU404

PAR2402

PAU304

PAU606

PAR2502PAR2801

PAU701PAR2701

PAU801

PAU804

PAR2802

PAU704

PAU1006

PAR2901

PAU1101

PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301PAR3402

PAU1305PAR3502

PAU13014

PAR3902

PAS405

PAR4002

PAU1903PAR4101

PAU1901PAU1904

PAR4201PAU2101

PAR4302

PAU2203PAR4401

PAU2201PAU2204

PAS103PAS303

PAP4010

PAU1027

PAU1403

PAP408

PAR3701

PAS401PAS402

PAU1029

PAF101 PAJ201


109

PAC102PAC101

COC1PAC202PAC201

COC2

PAC302PAC301

COC3PAC402

PAC401COC4

PAC502PAC501

COC5

PAC602PAC601

COC6PAC702PAC701

COC7PAC802PAC801

COC8

PAC902PAC901

COC9

PAC1001PAC1002COC10

PAC1102

PAC1101

COC11

PAC1202PAC1201

COC12

PAC1302PAC1301

COC13

PAC1401

PAC1402COC14

PAC1502


PAC1601COC16

PAC1702PAC1701

COC17

PAC1801

PAC1802

COC18PAC1902

PAC1901

COC19

PAC2002

PAC2001

COC20 PAC2102

PAC2101

COC21

PAC2202

PAC2201

COC22

PAC2302

PAC2301

COC23

PAC2402PAC2401

COC24

PAC2502PAC2501

COC25

PAC2602PAC2601

COC26

PAC2702PAC2701

COC27

PAC2801

PAC2802COC28

PAC2902 PAC2901

COC29

PAC3002

PAC3001

COC30 PAC3102

PAC3101

COC31

PAC3202

PAC3201


PAC3401

PAC3402


COC35

PAC3602PAC3601

COC36

PAC3702PAC3701

COC37PAC3802

PAC3801

COC38PAC3902PAC3901

COC39PAC4002

PAC4001

COC40

PAC4101

PAC4102

COC41

PAC4201

PAC4202

COC42

PAC4302PAC4301COC43

PAC4402PAC4401

COC44

PAC4502PAC4501

COC45

PAC4602PAC4601

COC46

PAC4702PAC4701

COC47

PAC4801PAC4802

COC48

PAC4901PAC4902

COC49

PAC5001PAC5002

COC50

PAC5101PAC5102

COC51

PAC5201PAC5202

COC52

PAC5302

PAC5301COC53

PAC5401

PAC5402

COC54

PAC5502

PAC5501 COC55

PAC5602

PAC5601

COC56

PACC10DH

COCC1

PACC20DH

COCC2

PACC30DH

COCC3

PACL10DHCOCL1

PACL20DHCOCL2

PAD102

PAD101

COD1

PAD202

PAD201

COD2

PAD301

PAD302

COD3

PAD402

PAD401

COD4

PAD1002PAD1001

COD10PAD1102PAD1101

COD11PAF102

PAF101COF1

PAJ105PAJ104

PAJ103PAJ102

PAJ101COJ1

PAJ202PAJ201 PAJ203

COJ2

PAL101PAL102

COL1

PAL201PAL202

COL2

PAL402

PAL401

COL4

PAL501PAL502

COL5 PAL701

PAL702COL7

PAP101

PAP102

PAP103

PAP104

PAP105

PAP106

PAP107

PAP108

PAP109

PAP1010

PAP1011

PAP1012

PAP1013

PAP1014COP1

PAP201

PAP202

PAP203

PAP204PAP205

PAP206

PAP207

PAP208PAP209

PAP2010

PAP2011

PAP2012

COP2

PAP301PAP302

PAP303PAP304

PAP305PAP306

PAP307PAP308

PAP309PAP3010

PAP3011PAP3012

COP3

PAP401

PAP402

PAP403

PAP404PAP405

PAP406

PAP407

PAP408PAP409

PAP4010

PAP4011

PAP4012

COP4PAP501

PAP502PAP503

PAP504PAP505

PAP506PAP507

PAP508PAP509

PAP5010PAP5011

PAP5012

COP5

PAP60U1PAP601

PAP603

PAP605

PAP607

PAP609

PAP6011

PAP6013

PAP6015

PAP6017

PAP6019

PAP602

PAP604

PAP606

PAP608

PAP6010

PAP6012

PAP6014

PAP6016

PAP6018

PAP6020

COP6

PAP70U1PAP701

PAP703

PAP705

PAP707

PAP709

PAP7011

PAP7013

PAP7015

PAP7017

PAP7019

PAP702

PAP704

PAP706

PAP708

PAP7010

PAP7012

PAP7014

PAP7016

PAP7018

PAP7020

COP7

PAQ103

PAQ102PAQ101

COQ1

PAQ201

PAQ202PAQ203

COQ2

PAQ303PAQ302

PAQ301COQ3

PAR101PAR102COR1

PAR201PAR202

COR2 PAR301PAR302

COR3

PAR401PAR402

COR4

PAR501PAR502

COR5

PAR601PAR602

COR6

PAR701PAR702COR7

PAR801

PAR802

COR8

PAR901

PAR902

COR9

PAR1002PAR1001

COR10

PAR1101PAR1102COR11

PAR1201PAR1202COR12

PAR1301PAR1302

COR13

PAR1401PAR1402

COR14

PAR1501PAR1502

COR15

PAR1602PAR1601COR16

PAR1701PAR1702

COR17

PAR1801PAR1802

COR18PAR1901

PAR1902

COR19

PAR2001PAR2002

COR20

PAR2101PAR2102

COR21

PAR2201

PAR2202

COR22PAR2301

PAR2302

COR23PAR2402

PAR2401COR24

PAR2501PAR2502

COR25

PAR2601

PAR2602COR26PAR2701

PAR2702

COR27PAR2801

PAR2802COR28

PAR2901PAR2902

COR29

PAR3001PAR3002COR30

PAR3102PAR3101COR31

PAR3201

PAR3202

COR32

PAR3302PAR3301COR33

PAR3401

PAR3402

COR34PAR3501

PAR3502

COR35

PAR3601

PAR3602

COR36

PAR3701

PAR3702

COR37

PAR3801PAR3802COR38

PAR3901PAR3902

COR39


PAR400MH PAR4001

COR40

PAR4101PAR4102

COR41PAR4201

PAR4202COR42


PAR430MH PAR4301

COR43

PAR4401PAR4402

COR44

PAR4701

PAR4702

COR47

PAR4801PAR4802COR48

PART101

PART102

CORT1

PAS102

PAS104PAS103

PAS101

COS1

PAS201PAS204

PAS202PAS203

COS2

PAS302

PAS304PAS303

PAS301

COS3

PAS403

PAS406PAS402

PAS404PAS405

PAS401

COS4


PAU1035

PAU1034

PAU1033

PAU1032PAU1031

PAU1030

PAU1029

PAU1028

PAU1027PAU1026

PAU1025




PAU1013PAU1012

PAU1011

PAU1010PAU109

PAU108

PAU107

PAU106

PAU105PAU104

PAU103

PAU102

PAU101

COU1

PAU201

PAU202

PAU203PAU204

PAU208

PAU207

PAU206PAU205

COU2

PAU301

PAU302PAU303

PAU304

PAU305

COU3

PAU401

PAU402PAU403

PAU404

PAU405

COU4

PAU501

PAU502PAU503

PAU504

PAU505

COU5

PAU60HSPAU604PAU605

PAU603PAU602

PAU606PAU601 COU6

PAU701

PAU702PAU703

PAU704

PAU705

COU7

PAU801

PAU802PAU803

PAU804

PAU805

COU8

PAU901

PAU902PAU903

PAU904

PAU905

COU9


PAU1003PAU1002

PAU1006PAU1001

COU10


COU11


PAU1203PAU1202

PAU1206PAU1201

COU12

PAU13015PAU13016PAU13017

PAU13018PAU13019

PAU13020PAU13021

PAU13022PAU13023PAU13024

PAU13025PAU13026

PAU13027PAU13028

PAU13014PAU13013PAU13012

PAU13011PAU13010

PAU1309PAU1308


PAU1304PAU1303

PAU1302PAU1301

COU13



COU14



COU15

PAU1601PAU1602


COU16


COU17

PAU1801

PAU1803

PAU1802

COU18


COU19

PAU2006


PAU2101PAU2102

PAU2103

COU21


PAU2202PAU2201

COU22

PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAF102 PART101

PAC5102

PAP204

PAR4102

PAU106

PAP2010

PAU1010

PAP205

PAU105

PAP3012

PAU1013

PAP3010


PAC5402

PAP206

PAR4402

PAU104

PAC602PAC702

PAC802

PAC1402

PAC1802PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501PAP2011

PAR4001PAR4301

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102

PAP402

PAR1402

PAU1035

PAC102PAC202

PAC2401

PAC4802

PAC5302

PAC5502

PAC5601

PAD101PAD201

PAD302

PAL101

PAL201

PAP602

PAR202PAR302

PAR402PAR502

PAR602

PAR1102PAR1202

PAR3702

PAR3802

PAR4702

PAS406

PAU208

PAU1405

PAU150B2

PAU1605

PAU2102

PAC1301

PAC1502

PAC1601PAC1701

PAC2301

PAC2601PAC2701

PAC3301

PAC5001

PAC5201

PAL502

PAU303

PAU305PAU405

PAU505

PAU703

PAU705PAU805

PAU905


PAC3202

PAP209

PAR2702

PAU109

PAC2202

PAP208

PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602PAC1702

PAC1801PAC1901

PAC2102

PAC2201

PAC2302

PAC2602PAC2702

PAC2801PAC2901

PAC3102

PAC3201 PAC3302

PAC3401PAC3501

PAC3601

PAC5002

PAC5101

PAC5202

PAC5401

PAL402

PAP2012

PAR3002

PAR4003PAR4303

PAU1012

PAU302PAU402

PAU502

PAU602PAU604

PAU702PAU802

PAU902

PAU1002PAU1004

PAU1102

PAU1202

PAU1204

PAU1902PAU2202

PAP7011

PAR3601

PAP701

PAS304

PAP7017

PAP703

PAP705

PAP707

PAP7019

PAP6014

PAP608

PAP6018

PAP1013

PAP1014

PAR102PAU1603

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAC4801 PAC4901

PAC5301

PAC5501

PAC5602

PAD402

PAJ104

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6020

PAP7013

PAQ102

PAQ202

PAQ302

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701PAR2001

PAR4802PAS102

PAS302

PAS403PAS404

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1404

PAU150A3

PAU1604

PAU1702

PAU1802

PAU2003

PAU2006

PAU2103PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAU1602

PAR401PAU1038

PAU150B3

PAP503

PAR1602

PAU1039

PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAS101PAS301

PAP105

PAP6012

PAC1202

PAP203

PAP6010

PAR1401

PAS104

PAU103

PAP709

PAP5012

PAR3101

PAR3201

PAU1048

PAR201

PAU1401PAR301

PAU1402

PAP4011

PAU1026

PAU1407PAP301

PAU1024

PAU1408

PAP401PAR601

PAU1036

PAU150A1

PAP406PAR501

PAU1031

PAU150B1

PAP7015

PAP3011

PAQ101

PAU1014

PAP309

PAU1016

PAU150A2

PAU201

PAU1406

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAP5011

PAU1047

PAU202

PAU1601

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502PAP504

PAU1040

PAP409

PAQ301

PAU1028

PAC4702PAJ202

PAJ203PAR4801

PAP501

PAR4202

PAU1037

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043

PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501

PAU504PAC2101

PAR2201PAU503

PAC3001

PAR2501

PAR2602

PAC3002

PAU803

PAU901

PAU904PAC3101

PAR2601PAU903

PAC3701

PAJ101

PAL701

PAC3802


PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAC4902

PAR3801PAU1803

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002

PAR1802

PAD1102

PAR1902

PAL702

PAU13020

PAP604PAR3102

PAP606PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ303PAR3901

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1601PAU1801

PAR1801PAU13023

PAR1901PAU13022


PAR2301PAU401

PAU404

PAR2402

PAU304

PAU606

PAR2502PAR2801

PAU701PAR2701

PAU801

PAU804

PAR2802

PAU704

PAU1006

PAR2901

PAU1101

PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301PAR3402

PAU1305PAR3502

PAU13014

PAR3902

PAS405

PAR4002

PAU1903PAR4101

PAU1901PAU1904

PAR4201PAU2101

PAR4302

PAU2203PAR4401

PAU2201PAU2204

PAS103PAS303

PAP4010

PAU1027

PAU1403

PAP408

PAR3701

PAS401PAS402

PAU1029

PAF101 PAJ201


110

PAC102PAC101

COC1PAC202PAC201

COC2

PAC302PAC301

COC3PAC402

PAC401COC4

PAC502PAC501

COC5

PAC602PAC601

COC6PAC702PAC701

COC7PAC802PAC801

COC8

PAC902PAC901

COC9

PAC1001PAC1002COC10

PAC1102

PAC1101

COC11

PAC1202PAC1201

COC12

PAC1302PAC1301

COC13

PAC1401

PAC1402COC14

PAC1502


PAC1601COC16

PAC1702PAC1701

COC17

PAC1801

PAC1802

COC18PAC1902

PAC1901

COC19

PAC2002

PAC2001

COC20 PAC2102

PAC2101

COC21

PAC2202

PAC2201

COC22

PAC2302

PAC2301

COC23

PAC2402PAC2401

COC24

PAC2502PAC2501

COC25

PAC2602PAC2601

COC26

PAC2702PAC2701

COC27

PAC2801

PAC2802COC28

PAC2902 PAC2901

COC29

PAC3002

PAC3001

COC30 PAC3102

PAC3101

COC31

PAC3202

PAC3201


PAC3401

PAC3402


COC35

PAC3602PAC3601

COC36

PAC3702PAC3701

COC37PAC3802

PAC3801

COC38PAC3902PAC3901

COC39PAC4002

PAC4001

COC40

PAC4101

PAC4102

COC41

PAC4201

PAC4202

COC42

PAC4302PAC4301COC43

PAC4402PAC4401

COC44

PAC4502PAC4501

COC45

PAC4602PAC4601

COC46

PAC4702PAC4701

COC47

PAC4801PAC4802

COC48

PAC4901PAC4902

COC49

PAC5001PAC5002

COC50

PAC5101PAC5102

COC51

PAC5201PAC5202

COC52

PAC5302

PAC5301COC53

PAC5401

PAC5402

COC54

PAC5502

PAC5501 COC55

PAC5602

PAC5601

COC56

PACC10DH

COCC1

PACC20DH

COCC2

PACC30DH

COCC3

PACL10DHCOCL1

PACL20DHCOCL2

PAD102

PAD101

COD1

PAD202

PAD201

COD2

PAD301

PAD302

COD3

PAD402

PAD401

COD4

PAD1002PAD1001

COD10PAD1102PAD1101

COD11PAF102

PAF101COF1

PAJ105PAJ104

PAJ103PAJ102

PAJ101COJ1

PAJ202PAJ201 PAJ203

COJ2

PAL101PAL102

COL1

PAL201PAL202

COL2

PAL402

PAL401

COL4

PAL501PAL502

COL5 PAL701

PAL702COL7

PAP101

PAP102

PAP103

PAP104

PAP105

PAP106

PAP107

PAP108

PAP109

PAP1010

PAP1011

PAP1012

PAP1013

PAP1014COP1

PAP201

PAP202

PAP203

PAP204PAP205

PAP206

PAP207

PAP208PAP209

PAP2010

PAP2011

PAP2012

COP2

PAP301PAP302

PAP303PAP304

PAP305PAP306

PAP307PAP308

PAP309PAP3010

PAP3011PAP3012

COP3

PAP401

PAP402

PAP403

PAP404PAP405

PAP406

PAP407

PAP408PAP409

PAP4010

PAP4011

PAP4012

COP4PAP501

PAP502PAP503

PAP504PAP505

PAP506PAP507

PAP508PAP509

PAP5010PAP5011

PAP5012

COP5

PAP60U1PAP601

PAP603

PAP605

PAP607

PAP609

PAP6011

PAP6013

PAP6015

PAP6017

PAP6019

PAP602

PAP604

PAP606

PAP608

PAP6010

PAP6012

PAP6014

PAP6016

PAP6018

PAP6020

COP6

PAP70U1PAP701

PAP703

PAP705

PAP707

PAP709

PAP7011

PAP7013

PAP7015

PAP7017

PAP7019

PAP702

PAP704

PAP706

PAP708

PAP7010

PAP7012

PAP7014

PAP7016

PAP7018

PAP7020

COP7

PAQ103

PAQ102PAQ101

COQ1

PAQ201

PAQ202PAQ203

COQ2

PAQ303PAQ302

PAQ301COQ3

PAR101PAR102COR1

PAR201PAR202

COR2 PAR301PAR302

COR3

PAR401PAR402

COR4

PAR501PAR502

COR5

PAR601PAR602

COR6

PAR701PAR702COR7

PAR801

PAR802

COR8

PAR901

PAR902

COR9

PAR1002PAR1001

COR10

PAR1101PAR1102COR11

PAR1201PAR1202COR12

PAR1301PAR1302

COR13

PAR1401PAR1402

COR14

PAR1501PAR1502

COR15

PAR1602PAR1601COR16

PAR1701PAR1702

COR17

PAR1801PAR1802

COR18PAR1901

PAR1902

COR19

PAR2001PAR2002

COR20

PAR2101PAR2102

COR21

PAR2201

PAR2202

COR22PAR2301

PAR2302

COR23PAR2402

PAR2401COR24

PAR2501PAR2502

COR25

PAR2601

PAR2602COR26PAR2701

PAR2702

COR27PAR2801

PAR2802COR28

PAR2901PAR2902

COR29

PAR3001PAR3002COR30

PAR3102PAR3101COR31

PAR3201

PAR3202

COR32

PAR3302PAR3301COR33

PAR3401

PAR3402

COR34PAR3501

PAR3502

COR35

PAR3601

PAR3602

COR36

PAR3701

PAR3702

COR37

PAR3801PAR3802COR38

PAR3901PAR3902

COR39


PAR400MH PAR4001

COR40

PAR4101PAR4102

COR41PAR4201

PAR4202COR42


PAR430MH PAR4301

COR43

PAR4401PAR4402

COR44

PAR4701

PAR4702

COR47

PAR4801PAR4802COR48

PART101

PART102

CORT1

PAS102

PAS104PAS103

PAS101

COS1

PAS201PAS204

PAS202PAS203

COS2

PAS302

PAS304PAS303

PAS301

COS3

PAS403

PAS406PAS402

PAS404PAS405

PAS401

COS4


PAU1035

PAU1034

PAU1033

PAU1032PAU1031

PAU1030

PAU1029

PAU1028

PAU1027PAU1026

PAU1025




PAU1013PAU1012

PAU1011

PAU1010PAU109

PAU108

PAU107

PAU106

PAU105PAU104

PAU103

PAU102

PAU101

COU1

PAU201

PAU202

PAU203PAU204

PAU208

PAU207

PAU206PAU205

COU2

PAU301

PAU302PAU303

PAU304

PAU305

COU3

PAU401

PAU402PAU403

PAU404

PAU405

COU4

PAU501

PAU502PAU503

PAU504

PAU505

COU5

PAU60HSPAU604PAU605

PAU603PAU602

PAU606PAU601 COU6

PAU701

PAU702PAU703

PAU704

PAU705

COU7

PAU801

PAU802PAU803

PAU804

PAU805

COU8

PAU901

PAU902PAU903

PAU904

PAU905

COU9


PAU1003PAU1002

PAU1006PAU1001

COU10


COU11


PAU1203PAU1202

PAU1206PAU1201

COU12

PAU13015PAU13016PAU13017

PAU13018PAU13019

PAU13020PAU13021

PAU13022PAU13023PAU13024

PAU13025PAU13026

PAU13027PAU13028

PAU13014PAU13013PAU13012

PAU13011PAU13010

PAU1309PAU1308


PAU1304PAU1303

PAU1302PAU1301

COU13



COU14



COU15

PAU1601PAU1602


COU16


COU17

PAU1801

PAU1803

PAU1802

COU18


COU19

PAU2006


PAU2101PAU2102

PAU2103

COU21


PAU2202PAU2201

COU22

PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAF102 PART101

PAC5102

PAP204

PAR4102

PAU106

PAP2010

PAU1010

PAP205

PAU105

PAP3012

PAU1013

PAP3010


PAC5402

PAP206

PAR4402

PAU104

PAC602PAC702

PAC802

PAC1402

PAC1802PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501PAP2011

PAR4001PAR4301

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102

PAP402

PAR1402

PAU1035

PAC102PAC202

PAC2401

PAC4802

PAC5302

PAC5502

PAC5601

PAD101PAD201

PAD302

PAL101

PAL201

PAP602

PAR202PAR302

PAR402PAR502

PAR602

PAR1102PAR1202

PAR3702

PAR3802

PAR4702

PAS406

PAU208

PAU1405

PAU150B2

PAU1605

PAU2102

PAC1301

PAC1502

PAC1601PAC1701

PAC2301

PAC2601PAC2701

PAC3301

PAC5001

PAC5201

PAL502

PAU303

PAU305PAU405

PAU505

PAU703

PAU705PAU805

PAU905


PAC3202

PAP209

PAR2702

PAU109

PAC2202

PAP208

PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602PAC1702

PAC1801PAC1901

PAC2102

PAC2201

PAC2302

PAC2602PAC2702

PAC2801PAC2901

PAC3102

PAC3201 PAC3302

PAC3401PAC3501

PAC3601

PAC5002

PAC5101

PAC5202

PAC5401

PAL402

PAP2012

PAR3002

PAR4003PAR4303

PAU1012

PAU302PAU402

PAU502

PAU602PAU604

PAU702PAU802

PAU902

PAU1002PAU1004

PAU1102

PAU1202

PAU1204

PAU1902PAU2202

PAP7011

PAR3601

PAP701

PAS304

PAP7017

PAP703

PAP705

PAP707

PAP7019

PAP6014

PAP608

PAP6018

PAP1013

PAP1014

PAR102PAU1603

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAC4801 PAC4901

PAC5301

PAC5501

PAC5602

PAD402

PAJ104

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6020

PAP7013

PAQ102

PAQ202

PAQ302

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701PAR2001

PAR4802PAS102

PAS302

PAS403PAS404

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1404

PAU150A3

PAU1604

PAU1702

PAU1802

PAU2003

PAU2006

PAU2103PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAU1602

PAR401PAU1038

PAU150B3

PAP503

PAR1602

PAU1039

PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAS101PAS301

PAP105

PAP6012

PAC1202

PAP203

PAP6010

PAR1401

PAS104

PAU103

PAP709

PAP5012

PAR3101

PAR3201

PAU1048

PAR201

PAU1401PAR301

PAU1402

PAP4011

PAU1026

PAU1407PAP301

PAU1024

PAU1408

PAP401PAR601

PAU1036

PAU150A1

PAP406PAR501

PAU1031

PAU150B1

PAP7015

PAP3011

PAQ101

PAU1014

PAP309

PAU1016

PAU150A2

PAU201

PAU1406

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAP5011

PAU1047

PAU202

PAU1601

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502PAP504

PAU1040

PAP409

PAQ301

PAU1028

PAC4702PAJ202

PAJ203PAR4801

PAP501

PAR4202

PAU1037

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043

PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501

PAU504PAC2101

PAR2201PAU503

PAC3001

PAR2501

PAR2602

PAC3002

PAU803

PAU901

PAU904PAC3101

PAR2601PAU903

PAC3701

PAJ101

PAL701

PAC3802


PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAC4902

PAR3801PAU1803

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002

PAR1802

PAD1102

PAR1902

PAL702

PAU13020

PAP604PAR3102

PAP606PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ303PAR3901

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1601PAU1801

PAR1801PAU13023

PAR1901PAU13022


PAR2301PAU401

PAU404

PAR2402

PAU304

PAU606

PAR2502PAR2801

PAU701PAR2701

PAU801

PAU804

PAR2802

PAU704

PAU1006

PAR2901

PAU1101

PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301PAR3402

PAU1305PAR3502

PAU13014

PAR3902

PAS405

PAR4002

PAU1903PAR4101

PAU1901PAU1904

PAR4201PAU2101

PAR4302

PAU2203PAR4401

PAU2201PAU2204

PAS103PAS303

PAP4010

PAU1027

PAU1403

PAP408

PAR3701

PAS401PAS402

PAU1029

PAF101 PAJ201


111

PAC102PAC101

COC1PAC202PAC201

COC2

PAC302PAC301

COC3PAC402

PAC401COC4

PAC502PAC501

COC5

PAC602PAC601

COC6PAC702PAC701

COC7PAC802PAC801

COC8

PAC902PAC901

COC9

PAC1001PAC1002COC10

PAC1102

PAC1101

COC11

PAC1202PAC1201

COC12

PAC1302PAC1301

COC13

PAC1401

PAC1402COC14

PAC1502


PAC1601COC16

PAC1702PAC1701

COC17

PAC1801

PAC1802

COC18PAC1902

PAC1901

COC19

PAC2002

PAC2001

COC20 PAC2102

PAC2101

COC21

PAC2202

PAC2201

COC22

PAC2302

PAC2301

COC23

PAC2402PAC2401

COC24

PAC2502PAC2501

COC25

PAC2602PAC2601

COC26

PAC2702PAC2701

COC27

PAC2801

PAC2802COC28

PAC2902 PAC2901

COC29

PAC3002

PAC3001

COC30 PAC3102

PAC3101

COC31

PAC3202

PAC3201


PAC3401

PAC3402


COC35

PAC3602PAC3601

COC36

PAC3702PAC3701

COC37PAC3802

PAC3801

COC38PAC3902PAC3901

COC39PAC4002

PAC4001

COC40

PAC4101

PAC4102

COC41

PAC4201

PAC4202

COC42

PAC4302PAC4301COC43

PAC4402PAC4401

COC44

PAC4502PAC4501

COC45

PAC4602PAC4601

COC46

PAC4702PAC4701

COC47

PAC4801PAC4802

COC48

PAC4901PAC4902

COC49

PAC5001PAC5002

COC50

PAC5101PAC5102

COC51

PAC5201PAC5202

COC52

PAC5302

PAC5301COC53

PAC5401

PAC5402

COC54

PAC5502

PAC5501 COC55

PAC5602

PAC5601

COC56

PACC10DH

COCC1

PACC20DH

COCC2

PACC30DH

COCC3

PACL10DHCOCL1

PACL20DHCOCL2

PAD102

PAD101

COD1

PAD202

PAD201

COD2

PAD301

PAD302

COD3

PAD402

PAD401

COD4

PAD1002PAD1001

COD10PAD1102PAD1101

COD11PAF102

PAF101COF1

PAJ105PAJ104

PAJ103PAJ102

PAJ101COJ1

PAJ202PAJ201 PAJ203

COJ2

PAL101PAL102

COL1

PAL201PAL202

COL2

PAL402

PAL401

COL4

PAL501PAL502

COL5 PAL701

PAL702COL7

PAP101

PAP102

PAP103

PAP104

PAP105

PAP106

PAP107

PAP108

PAP109

PAP1010

PAP1011

PAP1012

PAP1013

PAP1014COP1

PAP201

PAP202

PAP203

PAP204PAP205

PAP206

PAP207

PAP208PAP209

PAP2010

PAP2011

PAP2012

COP2

PAP301PAP302

PAP303PAP304

PAP305PAP306

PAP307PAP308

PAP309PAP3010

PAP3011PAP3012

COP3

PAP401

PAP402

PAP403

PAP404PAP405

PAP406

PAP407

PAP408PAP409

PAP4010

PAP4011

PAP4012

COP4PAP501

PAP502PAP503

PAP504PAP505

PAP506PAP507

PAP508PAP509

PAP5010PAP5011

PAP5012

COP5

PAP60U1PAP601

PAP603

PAP605

PAP607

PAP609

PAP6011

PAP6013

PAP6015

PAP6017

PAP6019

PAP602

PAP604

PAP606

PAP608

PAP6010

PAP6012

PAP6014

PAP6016

PAP6018

PAP6020

COP6

PAP70U1PAP701

PAP703

PAP705

PAP707

PAP709

PAP7011

PAP7013

PAP7015

PAP7017

PAP7019

PAP702

PAP704

PAP706

PAP708

PAP7010

PAP7012

PAP7014

PAP7016

PAP7018

PAP7020

COP7

PAQ103

PAQ102PAQ101

COQ1

PAQ201

PAQ202PAQ203

COQ2

PAQ303PAQ302

PAQ301COQ3

PAR101PAR102COR1

PAR201PAR202

COR2 PAR301PAR302

COR3

PAR401PAR402

COR4

PAR501PAR502

COR5

PAR601PAR602

COR6

PAR701PAR702COR7

PAR801

PAR802

COR8

PAR901

PAR902

COR9

PAR1002PAR1001

COR10

PAR1101PAR1102COR11

PAR1201PAR1202COR12

PAR1301PAR1302

COR13

PAR1401PAR1402

COR14

PAR1501PAR1502

COR15

PAR1602PAR1601COR16

PAR1701PAR1702

COR17

PAR1801PAR1802

COR18PAR1901

PAR1902

COR19

PAR2001PAR2002

COR20

PAR2101PAR2102

COR21

PAR2201

PAR2202

COR22PAR2301

PAR2302

COR23PAR2402

PAR2401COR24

PAR2501PAR2502

COR25

PAR2601

PAR2602COR26PAR2701

PAR2702

COR27PAR2801

PAR2802COR28

PAR2901PAR2902

COR29

PAR3001PAR3002COR30

PAR3102PAR3101COR31

PAR3201

PAR3202

COR32

PAR3302PAR3301COR33

PAR3401

PAR3402

COR34PAR3501

PAR3502

COR35

PAR3601

PAR3602

COR36

PAR3701

PAR3702

COR37

PAR3801PAR3802COR38

PAR3901PAR3902

COR39


PAR400MH PAR4001

COR40

PAR4101PAR4102

COR41PAR4201

PAR4202COR42


PAR430MH PAR4301

COR43

PAR4401PAR4402

COR44

PAR4701

PAR4702

COR47

PAR4801PAR4802COR48

PART101

PART102

CORT1

PAS102

PAS104PAS103

PAS101

COS1

PAS201PAS204

PAS202PAS203

COS2

PAS302

PAS304PAS303

PAS301

COS3

PAS403

PAS406PAS402

PAS404PAS405

PAS401

COS4


PAU1035

PAU1034

PAU1033

PAU1032PAU1031

PAU1030

PAU1029

PAU1028

PAU1027PAU1026

PAU1025




PAU1013PAU1012

PAU1011

PAU1010PAU109

PAU108

PAU107

PAU106

PAU105PAU104

PAU103

PAU102

PAU101

COU1

PAU201

PAU202

PAU203PAU204

PAU208

PAU207

PAU206PAU205

COU2

PAU301

PAU302PAU303

PAU304

PAU305

COU3

PAU401

PAU402PAU403

PAU404

PAU405

COU4

PAU501

PAU502PAU503

PAU504

PAU505

COU5

PAU60HSPAU604PAU605

PAU603PAU602

PAU606PAU601 COU6

PAU701

PAU702PAU703

PAU704

PAU705

COU7

PAU801

PAU802PAU803

PAU804

PAU805

COU8

PAU901

PAU902PAU903

PAU904

PAU905

COU9


PAU1003PAU1002

PAU1006PAU1001

COU10


COU11


PAU1203PAU1202

PAU1206PAU1201

COU12

PAU13015PAU13016PAU13017

PAU13018PAU13019

PAU13020PAU13021

PAU13022PAU13023PAU13024

PAU13025PAU13026

PAU13027PAU13028

PAU13014PAU13013PAU13012

PAU13011PAU13010

PAU1309PAU1308


PAU1304PAU1303

PAU1302PAU1301

COU13



COU14



COU15

PAU1601PAU1602


COU16


COU17

PAU1801

PAU1803

PAU1802

COU18


COU19

PAU2006


PAU2101PAU2102

PAU2103

COU21


PAU2202PAU2201

COU22

PAC4102PAC4202

PAC4301PAU1703

PAU2001PAU2002

PAC4701

PAF102 PART101

PAC5102

PAP204

PAR4102

PAU106

PAP2010

PAU1010

PAP205

PAU105

PAP3012

PAU1013

PAP3010


PAC5402

PAP206

PAR4402

PAU104

PAC602PAC702

PAC802

PAC1402

PAC1802PAC1902

PAC2802PAC2902

PAC3402PAC3502

PAL202

PAL501PAP2011

PAR4001PAR4301

PAU1011

PAU601

PAU1001

PAU1201

PAC302

PAC402

PAC502

PAL102

PAP402

PAR1402

PAU1035

PAC102PAC202

PAC2401

PAC4802

PAC5302

PAC5502

PAC5601

PAD101PAD201

PAD302

PAL101

PAL201

PAP602

PAR202PAR302

PAR402PAR502

PAR602

PAR1102PAR1202

PAR3702

PAR3802

PAR4702

PAS406

PAU208

PAU1405

PAU150B2

PAU1605

PAU2102

PAC1301

PAC1502

PAC1601PAC1701

PAC2301

PAC2601PAC2701

PAC3301

PAC5001

PAC5201

PAL502

PAU303

PAU305PAU405

PAU505

PAU703

PAU705PAU805

PAU905


PAC3202

PAP209

PAR2702

PAU109

PAC2202

PAP208

PAR2302

PAU108

PAC3602

PAP207

PAR2902

PAU107

PAC1102

PAC1302

PAC1401PAC1501

PAC1602PAC1702

PAC1801PAC1901

PAC2102

PAC2201

PAC2302

PAC2602PAC2702

PAC2801PAC2901

PAC3102

PAC3201 PAC3302

PAC3401PAC3501

PAC3601

PAC5002

PAC5101

PAC5202

PAC5401

PAL402

PAP2012

PAR3002

PAR4003PAR4303

PAU1012

PAU302PAU402

PAU502

PAU602PAU604

PAU702PAU802

PAU902

PAU1002PAU1004

PAU1102

PAU1202

PAU1204

PAU1902PAU2202

PAP7011

PAR3601

PAP701

PAS304

PAP7017

PAP703

PAP705

PAP707

PAP7019

PAP6014

PAP608

PAP6018

PAP1013

PAP1014

PAR102PAU1603

PAC101PAC201

PAC301

PAC401

PAC501

PAC601PAC701

PAC801

PAC901

PAC1001

PAC1101

PAC1201

PAC2402

PAC2501

PAC3702PAC3801

PAC3901

PAC4001PAC4101

PAC4201

PAC4302

PAC4401PAC4501

PAC4601

PAC4801 PAC4901

PAC5301

PAC5501

PAC5602

PAD402

PAJ104

PAL401

PAP104PAP108

PAP1010PAP1012

PAP404

PAP508

PAP6020

PAP7013

PAQ102

PAQ202

PAQ302

PAR101

PAR701

PAR902

PAR1301PAR1501

PAR1701PAR2001

PAR4802PAS102

PAS302

PAS403PAS404

PAU1033

PAU1044

PAU204

PAU207

PAU1307

PAU13018PAU13021

PAU13025PAU13026

PAU1404

PAU150A3

PAU1604

PAU1702

PAU1802

PAU2003

PAU2006

PAU2103PAP506

PAU1042PAP306

PAR1702PAS201

PAU1019

PAU1602

PAR401PAU1038

PAU150B3

PAP503

PAR1602

PAU1039

PAP505

PAQ201

PAU1041

PAP308

PAU1017

PAU206

PAS101PAS301

PAP105

PAP6012

PAC1202

PAP203

PAP6010

PAR1401

PAS104

PAU103

PAP709

PAP5012

PAR3101

PAR3201

PAU1048

PAR201

PAU1401PAR301

PAU1402

PAP4011

PAU1026

PAU1407PAP301

PAU1024

PAU1408

PAP401PAR601

PAU1036

PAU150A1

PAP406PAR501

PAU1031

PAU150B1

PAP7015

PAP3011

PAQ101

PAU1014

PAP309

PAU1016

PAU150A2

PAU201

PAU1406

PAP1011

PAP302

PAU1023

PAP109PAP103

PAP305

PAU1020

PAP107

PAP303

PAR1502

PAR2002PAS202

PAU1022

PAP5011

PAU1047

PAU202

PAU1601

PAP101

PAP304

PAU1021

PAP102

PAP202

PAR1302

PAU102

PAP201

PAR3301

PAR3401

PAU101

PAC2502

PAD1001PAD1101

PAR3501

PAU1304

PAU13017

PAP502PAP504

PAU1040

PAP409

PAQ301

PAU1028

PAC4702PAJ202

PAJ203PAR4801

PAP501

PAR4202

PAU1037

PAP4012

PAU1025

PAU205

PAC902PAU1032

PAC1002

PAU1043

PAC2001

PAR2101

PAR2202

PAC2002

PAU403

PAU501

PAU504PAC2101

PAR2201PAU503

PAC3001

PAR2501

PAR2602

PAC3002

PAU803

PAU901

PAU904PAC3101

PAR2601PAU903

PAC3701

PAJ101

PAL701

PAC3802


PAJ103

PAU13015

PAC4002PART102

PAU1701

PAC4402PAC4502

PAR4701

PAU2004PAC4602

PAU2005

PAC4902

PAR3801PAU1803

PAD102PAR801

PAD202PAR901

PAD301

PAR1002

PAD401PAR3602

PAD1002

PAR1802

PAD1102

PAR1902

PAL702

PAU13020

PAP604PAR3102

PAP606PAR3302

PAQ103PAR802

PAQ203PAR1001

PAQ303PAR3901

PAR702PAU1034

PAR1101PAS204

PAR1201PAS203

PAR1601PAU1801

PAR1801PAU13023

PAR1901PAU13022


PAR2301PAU401

PAU404

PAR2402

PAU304

PAU606

PAR2502PAR2801

PAU701PAR2701

PAU801

PAU804

PAR2802

PAU704

PAU1006

PAR2901

PAU1101

PAU1104

PAR3001

PAU1103

PAU1206

PAR3202

PAU1301PAR3402

PAU1305PAR3502

PAU13014

PAR3902

PAS405

PAR4002

PAU1903PAR4101

PAU1901PAU1904

PAR4201PAU2101

PAR4302

PAU2203PAR4401

PAU2201PAU2204

PAS103PAS303

PAP4010

PAU1027

PAU1403

PAP408

PAR3701

PAS401PAS402

PAU1029

PAF101 PAJ201


112

B.6. Sensor Node Bill of Materials

B.6 Sensor Node Bill of Materials


113

Bill o

f Ma

teria

lsS

ou

rce D

ata

Fro

m:

Ro

om

Sen

so

r, Rev 1

Pro

ject:

Ro

om

Sen

so

r, Rev 1

Desig

nM

. Ald

rich

Report D

ate:1/12/2010

1:01:13 PMPrint D

ate:29-Apr-10

6:40:39 PM#

Desig

nato

rD

escrip

tion

Man

ufa

ctu

rer 1

Man

ufa

ctu

rer P

art N

um

ber 1

Fo

otp

rint

Su

pp

lier 1

Su

pp

lier P

art N

um

ber 1

Qu

an

tity

1C

C1, C

C2, C

C3

OP

AQ

UE

CO

SIN

E C

OR

RE

CTO

RN

/AN

/AN

/AN

/AN

/A3

2C

L1, CL2

F4.63 F2.8 LEN

S &

HO

LDE

RN

/AN

/AN

/AN

/AN

/A2

3N

/AR

oom S

ensor, Controls P

CB

Advanced C

ircuitsN

/AN

/AN

/AN

/A1

4C

1, C2

CA

P C

ER

AM

IC 2.2U

F 10V X5R

0805K

emet

C0805C

225K8P

AC

TUC

AP

C2012L

Digi-K

ey399-3125-1-N

D2

5C

3, C6, C

7, C9, C

10, C11, C

14C

AP

CE

RM

1UF 6.3V

X5R 0402

Kem

etC

0402C105K

9PA

CTU

CA

PC

1005LD

igi-Key

399-4873-1-ND

36

C4, C

8, C37, C

46C

AP

CE

R 10000P

F 25V 10%

X7R 0402

Murata E

lectronics North A

merica

GR

M155R

71E103K

A01DC

AP

C1005L

Digi-K

ey490-1312-1-N

D4

7C

5C

AP

CE

R 2.2P

F 50V C

0G 0402

Murata E

lectronics North A

merica

GR

M1555C

1H2R

2CZ01D

CA

PC

1005LD

igi-Key

490-1267-1-ND

18

C12

CA

P C

ER

1000PF 50V

C0G

0402M

urata Electronics N

orth Am

ericaG

RM

1555C1H

102JA01D

CA

PC

1005LD

igi-Key

490-3244-1-ND

19

C20, C

30C

AP

CE

R .47U

F 6.3V X5R

0402M

urata Electronics N

orth Am

ericaG

RM

155R60J474K

E19D

CA

PC

1005LD

igi-Key

490-3266-1-ND

210

C21, C

31 C

AP

CE

R .22U

F 6.3V X5R

0402M

urata Electronics N

orth Am

ericaG

RM

155R60J224K

E01D

CA

PC

1005LD

igi-Key

490-5407-1-ND

211

C13, C

15, C17, C

19, C23, C

24, C25, C

27, C29, C

33, C35, C

50, C52

CA

P C

ER

.1UF 10V

10% X5R

0402M

urata Electronics N

orth Am

ericaG

RM

155R61A

104KA01D

CA

PC

1005LD

igi-Key

490-1318-1-ND

1312

C16, C

26, C48, C

55, C56

CA

P C

ER

.1UF 50V

10% X7R

0603M

urata Electronics N

orth Am

ericaG

RM

188R71H

104KA

93DC

AP

C1608N

Digi-K

ey490-1519-1-N

D

513

C18, C

28, C34, C

49, C53

CA

P C

ER

1.0UF 10V

10% X5R

0603M

urata Electronics N

orth Am

ericaG

RM

188R61A

105KA61D

CA

PC

1608ND

igi-Key

490-1543-1-ND

314

C38, C

39C

AP

CE

R 30P

F 50V 5%

C0G

0402M

urata Electronics N

orth Am

ericaG

RM

1555C1H

300JZ01DC

AP

C1005L

Digi-K

ey490-1285-1-N

D2

15C

40, C41, C

42C

AP

CE

R 22U

F 25V X5R

1210Taiyo Y

udenTM

K325B

J226MM

-TC

AP

C3225N

Digi-K

ey587-2086-1-N

D3

16C

43, C44, C

45C

AP

CE

R 2.2U

F 6.3V 10%

X7R 0805

Murata E

lectronics North A

merica

GR

M21B

R70J225K

A01L

CA

PC

2012LD

igi-Key

490-1698-1-ND

317

C47

CA

P 47U

F 35V E

LEC

T NH

G R

AD

IAL

Panasonic - E

CG

EC

A-1V

M470

CA

PP

R2-5x11

Digi-K

eyP

5164-ND

118

C22, C

32, C36, C

51, C54

CA

P C

ER

3300PF 50V

X7R 0402

Murata E

lectronics North A

merica

GR

M155R

71H332K

A01D

CA

PC

1005LD

igi-Key

490-3248-1-ND

519

D2, D

10, D11

LED

570NM

GR

EE

N D

IFF 0603 SM

DA

vago Technologies US

Inc.H

SM

G-C

1901608

Digi-K

ey516-1425-1-N

D3

20D

1, D4

LED

630NM

HE

RE

D D

IFF 0603 SM

DA


Inc.H

SM

S-C

1901608

Digi-K

ey516-1422-1-N

D2

21D

3E

MITTE

R IR

5MM

HI E

FF 940NM

Vishay/S

emiconductors

TSA

L6200H

LMP

-5029D

igi-Key

751-1204-ND

122

(D3)

HO

US

ING

RA

FOR

5MM

HIG

H D

OM

E LE

DA


Inc.H

LMP

-5029N

/AD

igi-Key

516-1395-ND

123

F1P

TC R

ES

ET 33V

1.85A S

MD

2920Littelfuse Inc

2920L185DR

THD

igi-Key

F2871CT-N

D1

24J1

US

B R

CP

T B-TY

PE

R/A

FULL B

AC

KFC

I61729-1010B

LFTH

Digi-K

ey609-3656-N

D1

25J2

MA

LE B

AR

RE

L CO

NN

EC

TOR

2.1x5.5S

witchcraft Inc.

RA

PC

722XR

AP

C722X

Mouser

502-RA

PC

722X1

26L1

FER

RITE

CH

IP 60 O

HM

1700MA

0402M

urata Electronics N

orth Am

ericaB

LM15P

D600S

N1D

1005D

igi-Key

490-5201-1-ND

127

L2, L4FILTE

R C

HIP

220 OH

M 700M

A 0402

Murata E

lectronics North A

merica

BLM

15EG

221SN1D

1005D

igi-Key

490-4005-1-ND

228

L5IN

DU

CTO

R M

ULTILA

YE

R 330N

H 0402

TDK

Corporation

MLG

1005SR

33J1005

Digi-K

ey445-3077-1-N

D1

29(L5)

IND

UC

TOR

MU

LTILAY

ER

2.2UH

0402TD

K C

orporationM

LF1005A2R

2KT

1005D

igi-Key

445-3518-1-ND

130

(L5)FE

RR

ITE C

HIP

1000 OH

M 200M

A 0402

Murata E

lectronics North A

merica

BLM

15HG

102SN

1D1005

Digi-K

ey490-3999-1-N

D1

31L7

FER

RITE

CH

IP 40 O

HM

550MA

0402TD

K C

orporationM

MZ1005Y

400C1005

Digi-K

ey445-2150-1-N

D1

32P

1C

ON

N H

DR

DU

AL 14P

OS

.100 SR

T AU

Molex C

onnector Corporation

10-89-7142TH

Digi-K

eyW

M26814-N

D1

33P

2, P3, P

4, P5

CO

NN

HE

AD

ER

12PO

S .100 V

ER

T TINM

olex Connector C

orporation22-28-4120

THD

igi-Key

WM

6412-ND

434

P6, P

7C

ON

N R

EC

EP

T 20PO

S 2M

M S

TR D

L SM

DFC

I91596-120LF

SM

TD

igi-Key

609-2730-ND

235

Q1, Q

2, Q3

MO

SFE

T N-C

H 30V

1.7A S

SO

T3Fairchild S

emiconductor

ND

S355A

NS

OT-23

Digi-K

eyN

DS

355AN

CT-N

D3

36R

24, R28, R

30 (0-500 lx)R

ES

820K O

HM

1/16W 1%

0402 SM

D (0-500 lx)

Vishay/D

aleC

RC

W0402820K

FKE

DR

ES

C1005L

Digi-K

ey541-820K

LCT-N

D3

37(R

24, R28, R

30) (0-2,000 lx)R

ES

402K O

HM

1/16W 1%

0402 SM

D (0-2,000 lx)

Vishay/D

aleC

RC

W0402402K

FKE

DR

ES

C1005L

Digi-K

ey541-402K

LCT-N

D3

38(R

24, R28, R

30) (0-10,000 lx)R

ES

182K O

HM

1/16W 1%

0402 SM

D (0-10,000 lx)

Vishay/D

aleC

RC

W0402182K

FKE

DR

ES

C1005L

Digi-K

ey541-182K

LCT-N

D3

39(R

24, R28, R

30) (0-35,000 lx)R

ES

100K O

HM

1/16W 1%

0402 SM

D (0-35,000 lx)

Rohm

Sem

iconductor M

CR

01MZP

F1003R

ES

C1005L

Digi-K

eyR

HM

100KLC

T-ND

340

R13, R

15R

ES

20.0K O

HM

1/16W 1%

0402 SM

DR

ohm S

emiconductor

MC

R01M

ZPF2002

RE

SC

1005LD

igi-Key

RH

M20.0K

LCT-N

D2

41R

11, R12

RE

S 820 O

HM

1/16W 1%

0402 SM

DV

ishay/Dale

CR

CW

0402820RFK

ED

RE

SC

1005LD

igi-Key

541-820LCT-N

D2

42R

17, R20

RE

S 3.30K

OH

M 1/16W

1% 0402 S

MD

Rohm

Sem

iconductor M

CR

01MZP

F3301R

ES

C1005L

Digi-K

eyR

HM

3.30KLC

T-ND

243

R1, R

7, R31, R

32, R33, R

34, R35

RE

S 0.0 O

HM

1/16W 5%

0402 SM

DR

ohm S

emiconductor

MC

R01M

ZPJ000

RE

SC

1005LD

igi-Key

RH

M0.0JC

T-ND

744

R2, R

3, R4, R

5, R6, R

21, R22, R

25, R26

RE

S 10.0K

OH

M 1/16W

1% 0402 S

MD

Rohm

Sem

iconductor M

CR

01MZP

F1002R

ES

C1005L

Digi-K

eyR

HM

10.0KLC

T-ND

945

R23, R

27, R29, R

41, R44

RE

S 56.0 O

HM

1/16W 1%

0402 SM

DR

ohm S

emiconductor

MC

R01M

ZPF56R

0R

ES

C1005L

Digi-K

eyR

HM

56.0LCT-N

D5


114

46R

14R

ES

2.70K O

HM

1/16W 1%

0402 SM

DR

ohm S

emiconductor

MC

R01M

ZPF2701

RE

SC

1005LD

igi-Key

RH

M2.70K

LCT-N

D1

47R

8, R9, R

18, R19, R

36, R39

RE

S 560 O

HM

1/10W 1%

0603 SM

DR

ohm S

emiconductor

MC

R03E

ZPFX5600

RE

SC

1608ND

igi-Key

RH

M560H

CT-N

D

648

R16, R

42R

ES

0.0 OH

M 1/10W

5% 0603 S

MD

Rohm

Sem

iconductorM

CR

03EZP

J000R

ES

C1608N

Digi-K

eyR

HM

0.0GC

T-ND

249

R38

RE

S 100 O

HM

1/10W 1%

0603 SM

DR

ohm S

emiconductor

MC

R03E

ZPFX1000

RE

SC

1608ND

igi-Key

RH

M100H

CT-N

D1

50R

37R

ES

10.0K O

HM

1/10W 1%

0603 SM

DR

ohm S

emiconductor

MC

R03E

ZPFX1002

RE

SC

1608ND

igi-Key

RH

M10.0K

HC

T-ND

151

R10

RE

S 16.9 O

HM

1/8W 1%

0805 SM

DR

ohm S

emiconductor

MC

R10E

ZHF16R

9R

ES

C2012L

Digi-K

eyR

HM

16.9CC

T-ND

152

(R10)

RE

S 4.32 O

HM

1/8W 1%

0805 SM

DV

ishay/Dale

CR

CW

08054R32FK

EA

RE

SC

2012LD

igi-Key

541-4.32CC

CT-N

D1

53R

40, R43

Slide P

otentiometers 10K

OH

M B

TAP

ER

ALP

SR

S6011Y

A6009

THM

ouser688-R

S6011Y

A6009

254

R47, R

48R

ES

0.0 OH

M 1W

5% 2512 S

MD

Vishay/D

aleC

RC

W25120000Z0E

G2512

Digi-K

ey541-0.0XC

T-ND

255

RT1

CU

RR

EN

T LIMITE

R IN

RU

SH

5 0HM

20%C

anthermM

F72-005D7

THD

igi-Key

317-1148-ND

156

S1, S

3S

WITC

H TA

CT 6M

M 160G

F H=4.3M

MO

mron E

lectronics B

3S-1000P

THD

igi-Key

SW

836CT-N

D2

57S

2S

WITC

H D

IP H

ALF P

ITCH

2PO

SC

TS E

lectrocomponents

218-2LPS

TTH

Digi-K

eyC

T2182LPS

T-ND

158

S4

TAC

TILE S

WITC

HE

S 260gf R

ED

LED

Mountain S

witch

101-TS5511T26002R

-EV

THM

ouser101-TS

5511T26002R-E

V1

59(S

4)K

EY

CA

P R

OU

ND

BLA

CK

Mountain S

witch

101-TS55K

CR

-BLK

-EV

N/A

Mouser

101-TS55K

CR

-BLK

-EV

160

U1

IC M

CU

32-BIT W

/FLAS

H 48-LQ

FPTexas Instrum

entsTM

X320F28027PTA

48LQFP

Digi-K

ey296-24123-N

D1

61U

2IC

2-1LINE

DA

TA S

ELE

CT/M

UX S

M8

Texas Instruments

SN

74LVC

2G157D

CTR

TSO

P8

Digi-K

ey296-13266-1-N

D1

62U

3, U4, U

5, U7, U

8, U9, U

11, U19, U

22IC

OP

AM

P S

NG

L R-R

I/O LN

SO

T23-5A

nalog Devices

AD

8605AR

TZS

OT23-5

Digi-K

eyA

D8605A

RTZR

EE

L7CT-N

D9

63U

6, U10, U

12IC

PH

OTO

DE

TEC

TOR

AM

BIE

NT 6-O

DFN

IntersilIS

L29009IRO

Z-T76D

FND

igi-Key

ISL29009IR

OZ-T7C

T-ND

364

U13

IC U

SB

TO S

ER

IAL U

AR

T 28-SS

OP

FTDI

FT232RL R

28SS

OP

Digi-K

ey768-1007-1-N

D1

65U

14P

hotodiodes TriColor S

ensor LTF Low P

ower

TAO

STC

S3200D

-TR8S

OIC

Mouser

856-TCS

3200D-TR

166

U15

Photodiodes TriC

olor Sensor R

GB

, Clear C

hTA

OS

TCS

3414CS

6BG

AM

ouser856-TC

S3414C

S1

67U

16IC

PW

M TE

MP

SN

SR

CM

OS

/TTL SC

70-5A

nalog Devices Inc

TMP

05AK

SZ-500R

L7S

C-70

Digi-K

eyTM

P05A

KS

Z-500RL7C

T-ND

168

U17

SW

ITCH

ING

RE

GU

LATO

R, 5V

, 1.0 AR

ecom P

ower Incorporated

R-78B

5.0-1.0TH

Allied

394-00521

69U

18IC

IR R

CV

R M

OD

38KH

Z DO

ME

AXIA

LV

ishay/Sem

iconductorsTS

OP

38338TH

Digi-K

ey751-1388-N

D1

70U

20IC

LDO

RE

G 3.3V

1A S

OT223-6

Texas Instruments

TPS

79633DC

QR

SO

T223-6D

igi-Key

296-13766-1-ND

171

U21

SE

NS

OR

MO

TION

DIG

STD

WH

T PC

BP

anasonic Electric W

orksA

MN

11112TH

Digi-K

ey255-1812-N

D1

Ap

pro

ved

No

tes

167

Lin

e Ite

m (3

5) E

q. P

/N: 3

M, 9

53220-2

000-A

R-P

R; 3

M, 9

56220-2

000-A

R-P

R; S

ullin

s N

PP

N102G

HN

P-R

C (D

IGIK

EY

)

Lin

e Ite

m (6

6) E

q. P

/N: T

AO

S, T

CS

3210D

-TR

(MO

US

ER

)

Lin

e Ite

m (7

1) A

ltern

ate

Su

pp

lier: M

ou

ser, P

N: 5

95-T

PS

79633D

CQ


115

B.7. LED Ring Schematic

B.7 LED Ring Schematic


116

1 1

2 2

3 3

4 4

5 5

6 6

DD

CC

BB

AA

S01

S12

OE3

GND

4VD

D5

OUT

6

S27

S38

U1TCS32X0

SCLA1

SYNC

A2

GND

A3

SDAB1

VDD

B2

INTB3

U2TCS34X4CS

C1TBD 0603

C2TBD 0603

3V3

GND

GND

3V3

VDD

5GN

D4

IN2

FUNC

3TEM

P1

U3TMP05AK

SZ

C3TBD 0603

3V3

GND

Color to Frequency

I2C Color Sensor

PWM

Temp Sensor

3

D1REBEL

REDRTN

AMBRTN

GRNRTN

BLURTN

RED

AMB

GRN

BLU

HSHS

HSHS

HSHS

HSHS

11

22

33

44

55

66

77

88

99

1010

1111

1212

1313

1414

1515

1616

1717

1818

1919

2020

2121

2222

2323

2424

2525

2626

2727

2828

2929

3030

3131

3232

3333

3434

3535

3636

3737

3838

3939

4040

4141

4242

4343

4444

4545

4646

4747

4848

4949

5050

H1N2550-6002-RB

REDRTN

RED

AMB

REDRED

REDRTN

REDRTN

AMB

AMB

AMBRTN

AMBRTN

AMBRTN

GRNGRN

GRN

GRNRTNGRNRTNGRNRTN

BLUBLU

BLU

BLURTN

BLURTN

BLURTN

OES0S1S2S3

OUT

SYNC

SCLSDA

INT

INFUNC

TEMP

GND

OES1

S0S3

S23V3OU

T

SCL

SDA

SYNC

GND

INT

FUNC

INTEM

PGN

D

3

D6REBEL

3

D4REBEL

3

D9REBEL

3

D2REBEL

3

D7REBEL

3

D3REBEL

3

D8REBEL

CYNRTN

CYN

HSHS

3

D5REBEL

3

D10REBEL

CYN

CYN

CYN

CYNRTN

CYNRTN

CYNRTN

PIC101PIC102 COC1

PIC201PIC202

COC2PIC301PIC302 COC3

PID101

PID102

PID103 COD1P

ID201

PID202

PID203 COD2P

ID301

PID302

PID303 COD3 P

ID401

PID402

PID403 COD4P

ID501

PID502

PID503 COD5

PID601

PID602

PID603 COD6P

ID701

PID702

PID703 COD7P

ID801

PID802

PID803 COD8 P

ID901

PID902

PID903 COD9P

ID1001

PID1002

PID1003 COD10

PIH101

PIH102

PIH103PIH104

PIH105

PIH106

PIH107

PIH108

PIH109

PIH1010

PIH1011PIH1012

PIH1013PIH1014

PIH1015PIH1016

PIH1017PIH1018

PIH1019PIH1020

PIH1021PIH1022

PIH1023PIH1024

PIH1025PIH1026

PIH1027PIH1028

PIH1029PIH1030

PIH1031PIH1032

PIH1033PIH1034

PIH1035PIH1036

PIH1037PIH1038

PIH1039PIH1040

PIH1041PIH1042

PIH1043PIH1044

PIH1045PIH1046

PIH1047PIH1048

PIH1049PIH1050

COH1

PIU101

PIU102

PIU103

PIU104

PIU105

PIU106

PIU107

PIU108 COU1

PIU20A1

PIU20A2

PIU20A3

PIU20B1

PIU20B2

PIU20B3

COU2

PIU301PIU302

PIU303

PIU304PIU305 COU3 PIC102

PIC201

PIC302

PIH1038

PIU105

PIU20B2

PIU305

NL3V3

PID402

PIH107

PIH108

PIH109

NLAMB

PID901

PIH1010

PIH1011PIH1012

NLAMBRTN

PID302

PIH1013PIH1014

PIH1015

NLBLU

PID801

PIH1016

PIH1017PIH1018

NLBLURTN

PID102

PIH1025PIH1026

PIH1027

NLCYN

PID601

PIH1028

PIH1029PIH1030

NLCYNRTN

PIH1046

PIU303NLFUNC

PIC101

PIC202

PIC301

PIH1031

PIH1041

PIH1045

PIU104PIU20A3

PIU304

NLGND

PID202

PIH1019PIH1020

PIH1021

NLGRN

PID701

PIH1022

PIH1023PIH1024

NLGRNRTN

PID103

PID203

PID303

PID403

PID503

PID603

PID703

PID803

PID903

PID1003

NLHS

PIH1043

PIU302NLIN

PIH1040

PIU20B3NLINT

PID101

PID602

PID201

PID702

PID301

PID802

PID401

PID902

PID501

PID1002

PIH1042

PIH1049PIH1050

PIH1032

PIU103

NLOE

PIH1037

PIU106

NLOUT

PID502

PIH101

PIH102

PIH103

NLRED

PID1001

PIH104

PIH105

PIH106

NLREDRTN

PIH1034

PIU101NLS0

PIH1033

PIU102NLS1

PIH1036

PIU107

NLS2

PIH1035

PIU108

NLS3

PIH1047

PIU20A1NLSCL

PIH1039

PIU20B1NLSDA

PIH1048

PIU20A2NLSYNC

PIH1044

PIU301NLTEMP

B.7. LED Ring Schematic

117

B.8. LED Ring PCB

B.8 LED Ring PCB


118

PAC102PAC101

COC1

PAC202PAC201

COC2PAC302PAC301

COC3PAD103

PAD102PAD101

COD1

PAD203PAD202

PAD201

COD2

PAD303 PAD302

PAD301

COD3

PAD403

PAD402PAD401

COD4

PAD503

PAD502 PAD501

COD5

PAD603

PAD602 PAD601

COD6

PAD703

PAD702

PAD701COD7

PAD803PAD802

PAD801

COD8

PAD903

PAD902

PAD901

COD9

PAD1003

PAD1002PAD1001

COD10



PAH1017

PAH1019

PAH1021

PAH1023PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049



PAH1018

PAH1020

PAH1022

PAH1024PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAU105

PAU106

PAU107PAU108

PAU104

PAU103

PAU102PAU101

COU1



COU2PAU301PAU302

PAU305PAU304

PAU303

COU3

PAC102

PAC201 PAC302

PAH1038

PAU105

PAU20B2

PAU305

PAD402

PAH107 PAH108

PAH109

PAD901

PAH1010PAH1011 PAH1012

PAD302

PAH1013 PAH1014PAH1015

PAD801

PAH1016

PAH1017 PAH1018

PAD102

PAH1025 PAH1026

PAH1027

PAD601

PAH1028

PAH1029 PAH1030

PAH1046

PAU303

PAC101

PAC202 PAC301

PAH1031

PAH1041

PAH1045

PAU104

PAU20A3 PAU304

PAD202

PAH1019 PAH1020

PAH1021

PAD701

PAH1022

PAH1023 PAH1024

PAD103

PAD203

PAD303

PAD403

PAD503PAD603

PAD703

PAD803

PAD903

PAD1003

PAH1043

PAU302

PAH1040

PAU20B3

PAD101

PAD602

PAD201

PAD702

PAD301 PAD802

PAD401

PAD902

PAD501

PAD1002

PAH1032

PAU103

PAH1037

PAU106

PAD502

PAH101 PAH102PAH103

PAD1001

PAH104PAH105 PAH106

PAH1034

PAU101

PAH1033

PAU102

PAH1036

PAU107

PAH1035

PAU108

PAH1047

PAU20A1

PAH1039

PAU20B1

PAH1048

PAU20A2

PAH1044

PAU301

B.8. LED Ring PCB

119

PAC102PAC101

COC1

PAC202PAC201

COC2PAC302PAC301

COC3PAD103

PAD102PAD101

COD1

PAD203PAD202

PAD201

COD2

PAD303 PAD302

PAD301

COD3

PAD403

PAD402PAD401

COD4

PAD503

PAD502 PAD501

COD5

PAD603

PAD602 PAD601

COD6

PAD703

PAD702

PAD701COD7

PAD803PAD802

PAD801

COD8

PAD903

PAD902

PAD901

COD9

PAD1003

PAD1002PAD1001

COD10



PAH1017

PAH1019

PAH1021

PAH1023PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049



PAH1018

PAH1020

PAH1022

PAH1024PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAU105

PAU106

PAU107PAU108

PAU104

PAU103

PAU102PAU101

COU1



COU2PAU301PAU302

PAU305PAU304

PAU303

COU3

PAC102

PAC201 PAC302

PAH1038

PAU105

PAU20B2

PAU305

PAD402

PAH107 PAH108

PAH109

PAD901


PAD302


PAD801

PAH1016

PAH1017 PAH1018

PAD102

PAH1025 PAH1026

PAH1027

PAD601

PAH1028

PAH1029 PAH1030

PAH1046

PAU303

PAC101

PAC202 PAC301

PAH1031

PAH1041

PAH1045

PAU104

PAU20A3 PAU304

PAD202

PAH1019 PAH1020

PAH1021

PAD701

PAH1022

PAH1023 PAH1024

PAD103

PAD203

PAD303

PAD403

PAD503PAD603

PAD703

PAD803

PAD903

PAD1003

PAH1043

PAU302

PAH1040

PAU20B3

PAD101

PAD602

PAD201

PAD702

PAD301 PAD802

PAD401

PAD902

PAD501

PAD1002

PAH1032

PAU103

PAH1037

PAU106

PAD502

PAH101 PAH102PAH103

PAD1001

PAH104PAH105 PAH106

PAH1034

PAU101

PAH1033

PAU102

PAH1036

PAU107

PAH1035

PAU108

PAH1047

PAU20A1

PAH1039

PAU20B1

PAH1048

PAU20A2

PAH1044

PAU301

B.8. LED Ring PCB

120

PAC102PAC101

COC1

PAC202PAC201

COC2PAC302PAC301

COC3PAD103

PAD102PAD101

COD1

PAD203PAD202

PAD201

COD2

PAD303 PAD302

PAD301

COD3

PAD403

PAD402PAD401

COD4

PAD503

PAD502 PAD501

COD5

PAD603

PAD602 PAD601

COD6

PAD703

PAD702

PAD701COD7

PAD803PAD802

PAD801

COD8

PAD903

PAD902

PAD901

COD9

PAD1003

PAD1002PAD1001

COD10



PAH1017

PAH1019

PAH1021

PAH1023PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049



PAH1018

PAH1020

PAH1022

PAH1024PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAU105

PAU106

PAU107PAU108

PAU104

PAU103

PAU102PAU101

COU1



COU2PAU301PAU302

PAU305PAU304

PAU303

COU3

PAC102

PAC201 PAC302

PAH1038

PAU105

PAU20B2

PAU305

PAD402

PAH107 PAH108

PAH109

PAD901


PAD302


PAD801

PAH1016

PAH1017 PAH1018

PAD102

PAH1025 PAH1026

PAH1027

PAD601

PAH1028

PAH1029 PAH1030

PAH1046

PAU303

PAC101

PAC202 PAC301

PAH1031

PAH1041

PAH1045

PAU104

PAU20A3 PAU304

PAD202

PAH1019 PAH1020

PAH1021

PAD701

PAH1022

PAH1023 PAH1024

PAD103

PAD203

PAD303

PAD403

PAD503PAD603

PAD703

PAD803

PAD903

PAD1003

PAH1043

PAU302

PAH1040

PAU20B3

PAD101

PAD602

PAD201

PAD702

PAD301 PAD802

PAD401

PAD902

PAD501

PAD1002

PAH1032

PAU103

PAH1037

PAU106

PAD502

PAH101 PAH102PAH103

PAD1001

PAH104PAH105 PAH106

PAH1034

PAU101

PAH1033

PAU102

PAH1036

PAU107

PAH1035

PAU108

PAH1047

PAU20A1

PAH1039

PAU20B1

PAH1048

PAU20A2

PAH1044

PAU301

B.8. LED Ring PCB

121

PAC102PAC101

COC1

PAC202PAC201

COC2PAC302PAC301

COC3PAD103

PAD102PAD101

COD1

PAD203PAD202

PAD201

COD2

PAD303 PAD302

PAD301

COD3

PAD403

PAD402PAD401

COD4

PAD503

PAD502 PAD501

COD5

PAD603

PAD602 PAD601

COD6

PAD703

PAD702

PAD701COD7

PAD803PAD802

PAD801

COD8

PAD903

PAD902

PAD901

COD9

PAD1003

PAD1002PAD1001

COD10



PAH1017

PAH1019

PAH1021

PAH1023PAH1025

PAH1027

PAH1029

PAH1031

PAH1033

PAH1035

PAH1037

PAH1039

PAH1041

PAH1043

PAH1045

PAH1047

PAH1049



PAH1018

PAH1020

PAH1022

PAH1024PAH1026

PAH1028

PAH1030

PAH1032

PAH1034

PAH1036

PAH1038

PAH1040

PAH1042

PAH1044

PAH1046

PAH1048

PAH1050

COH1

PAU105

PAU106

PAU107PAU108

PAU104

PAU103

PAU102PAU101

COU1



COU2PAU301PAU302

PAU305PAU304

PAU303

COU3

PAC102

PAC201 PAC302

PAH1038

PAU105

PAU20B2

PAU305

PAD402

PAH107 PAH108

PAH109

PAD901


PAD302


PAD801

PAH1016

PAH1017 PAH1018

PAD102

PAH1025 PAH1026

PAH1027

PAD601

PAH1028

PAH1029 PAH1030

PAH1046

PAU303

PAC101

PAC202 PAC301

PAH1031

PAH1041

PAH1045

PAU104

PAU20A3 PAU304

PAD202

PAH1019 PAH1020

PAH1021

PAD701

PAH1022

PAH1023 PAH1024

PAD103

PAD203

PAD303

PAD403

PAD503PAD603

PAD703

PAD803

PAD903

PAD1003

PAH1043

PAU302

PAH1040

PAU20B3

PAD101

PAD602

PAD201

PAD702

PAD301 PAD802

PAD401

PAD902

PAD501

PAD1002

PAH1032

PAU103

PAH1037

PAU106

PAD502

PAH101 PAH102PAH103

PAD1001

PAH104PAH105 PAH106

PAH1034

PAU101

PAH1033

PAU102

PAH1036

PAU107

PAH1035

PAU108

PAH1047

PAU20A1

PAH1039

PAU20B1

PAH1048

PAU20A2

PAH1044

PAU301

B.8. LED Ring PCB

122

Appendix C

Firmware

Due to length, all source code and binaries for the LED controller and sensor node can befound at:

http://media.mit.edu/~maldrich/lighting/

123

http://media.mit.edu/~maldrich/lighting/

Appendix D

Radiometric Traceability

Attached are the traceability documents from Ocean Optics outlining NIST calibrationstandards and uncertainty that is carried over in the radiometric calibration source used inthese experiments.

124

Bibliography

[1] Abramson, M. A., 2003. Pattern search algorithms for mixed variable general con-strained optimization problems. Ph.D. thesis, Rice University.

[2] Abramson, M. A., Audet, C., Dennis, J. E., Le Digabel, S., 2009. Orthomads: Adeterministic MADS instance with orthogonal directions. SIAM J.Optim. 20 (2), 948–966.

[3] Ackermann, B., Schulz, V., Martiny, C., Hilgers, A., Zhu, X., 2006. Control of LEDs. In:Industry Applications Conference, 2006. 41st IAS Annual Meeting. Conference Recordof the 2006 IEEE. Vol. 5. pp. 2608–2615.

[4] Akashi, Y., Leslie, R., Novello, M., Nakamura, Y., 2001. Comparing lighting energyconservation measures in the United States and Japan 285, 127.

[5] Aldrich, M., Garbus, M., 2008. Calibration method and apparatus for lighting fix-tures using multiple spectrum light sources and light mixing. U.S. Patent Application,20080103714.

[6] Ashdown, I. E., 2004. Neural networks for LED color control. Vol. 5187. SPIE, pp.215–226.

[7] Audet, C., Béchard, V., Chaouki, J., Jun. 2008. Spent potliner treatment process op-timization using a MADS algorithm. Optimization and Engineering 9 (2), 143–160.

[8] Audet, C., Béchard, V., Digabel, S., 2008. Nonsmooth optimization through meshadaptive direct search and variable neighborhood search. J.Global Optimiz. 41 (2),299–318.

[9] Audet, C., Custódio, A. L., Dennis, Jr, J., 2007. Erratum: Mesh adaptive direct searchalgorithms for constrained optimization. SIAM J.on Optimization 18 (4), 1501–1503.

[10] Audet, C., Dennis, Jr, J., 2006. Mesh adaptive direct search algorithms for constrainedoptimization. SIAM J.on Optimization 17 (1), 188–217.

[11] Azevedo, I. L., Morgan, M. G., Morgan, F., 2009. The transition to solid-state lighting.Proceedings of the IEEE 97 (3), 481–510.

142

Bibliography

[12] Bergh, A., Craford, G., Duggal, A., Haitz, R., 2001. The promise and challenge ofsolid-state lighting. Phys Today 54 (12), 42–47.

[13] Bolt, R. A., 1980. "Put that there": Voice and gesture at the graphics interface. In:SIGGRAPH ’80: Proceedings of the 7th annual conference on Computer graphics andinteractive techniques. Vol. Seattle, Washington, United States. ACM, New York, NY,USA, pp. 262–270.

[14] Boynton, R. M., 1996. History and current status of a physiologically based system ofphotometry and colorimetry. J.Opt.Soc.Am.A 13 (8), 1609–1621.

[15] Burmen, M., Pernus, F., Likar, B., 2008. LED light sources: a survey of quality-affectingfactors and methods for their assessment. Measurement Science and Technology (12),122002–.

[16] Crisp, V. H. C., 1977. Preliminary study of automatic daylight control of artificiallighting. Lighting Research and Technology 9 (1), 31–41.

[17] Crisp, V. H. C., 1978. The light switch in buildings. Lighting Research and Technology10 (2), 69–82.

[18] Crisp, V. H. C., Henderson, G., 1982. The energy management of artificial lighting use.Lighting Research and Technology 14 (4), 193–206.

[19] Dantzig, G., 1974. Linear programming and extensions. Princeton University Press.

[20] Davis, W., Ohno, Y., 2005. Toward an improved color rendering metric. Vol. 5941.SPIE, pp. 59411G–.

[21] Esch, J., 2009. Prolog to the transition to solid-state lighting. Proceedings of the IEEE97; 97 (3), 478–480.

[22] Feldmeier, M., 2009. Personalized building comfort control. Ph.D. thesis, MassachusettsInstitute of Technology.

[23] Garcia, J., Dalla-Costa, M. A., Cardesin, J., Alonso, J. M., Rico-Secades, M., 2009.Dimming of high-brightness LEDs by means of luminous flux thermal estimation. PowerElectronics, IEEE Transactions on 24; 24 (4), 1107–1114.

[24] Graves, H., 2006. Developing the domestic luminaire market for LED technology - howcan policy help drive synchronous development? In: EEDAL 06 - Energy Efficiency inDomestic Appliances and Lighting, 4th Int’l Conference on.

[25] Guo, X., Houser, K., 2004. A review of colour rendering indices and their applicationto commercial light sources. Lighting Research and Technology 36 (3), 183–197.

143

Bibliography

[26] Harris, J. B., 1993. Electric lamps, past and present. Engineering Science and EducationJournal 2; 2 (4), 161–170.

[27] Hooke, R., Jeeves, T. A., 1961. "Direct search" solution of numerical and statisticalproblems. J.ACM 8 (2), 212–229.

[28] Hunt, D. R. G., 1977. Simple expressions for predicting energy savings from photo-electric control of lighting. Lighting Research and Technology 9 (2), 93–102.

[29] Hunt, D. R. G., 1979. Improved daylight data for predicting energy savings from pho-toelectric controls. Lighting Research and Technology 11 (1), 9–23.

[30] Hwang, J., Lee, D.-H., Park, S., Kim, Y.-W., Park, S.-N., 2009. Measurement un-certainty evaluation for emission color and luminance of displays. Appl. Opt. 48 (1),99–105.

[31] Jones, D. R., Perttunen, C. D., Stuckman, B. E., 1993. Lipschitzian optimization with-out the lipschitz constant. J.Optimiz.Theory Appl. 79 (1), 157–181.

[32] Kolda, T. G., Lewis, R. M., Torczon, V., 2004. Optimization by direct search: Newperspectives on some classical and modern methods. SIAM Rev 45 (3), 385–482.

[33] Krames, M. R., Shchekin, O. B., Mueller-Mach, R., Mueller, G. O., Zhou, L., Harbers,G., Craford, M. G., 2007. Status and future of high-power light-emitting diodes forsolid-state lighting. Display Technology, Journal of 3 (2), 160–175.

[34] Lee, E., 2008. Cyber physical systems: Design challenges. In: Object Oriented Real-Time Distributed Computing (ISORC), 2008 11th IEEE International Symposium on.pp. 363–369.

[35] Lei, Z., Xia, G., Ting, L., Xiaoling, G., Qiao Ming, L., Guangdi, S., 2007. Colorrendering and luminous efficacy of trichromatic and tetrachromatic LED-based whiteLEDs. Microelectron.J. 38 (1), 1–6.

[36] Lewis, R. M., Torczon, V., 2000. Pattern search methods for linearly constrained min-imization. SIAM Journal on Optimization 10 (3), 917–941.

[37] Lewis, R. M., Torczon, V., Trosset, M. W., 2000. Direct search methods: then andnow. J.Comput.Appl.Math. 124 (1-2), 191–207.

[38] Machado, C., Mendes, J. A., August, 2008. Automatic light control in domotics usingartificial neural networks. In: World Academy of Science, Engineering and Technology.Vol. 44. pp. 813–.

[39] Mistry, P., Maes, P., 2009. Sixthsense: a wearable gestural interface. In: SIGGRAPHASIA ’09: ACM SIGGRAPH ASIA 2009 Sketches. Vol. Yokohama, Japan. ACM, NewYork, NY, USA, pp. 1–1.

144

Bibliography

[40] Mozer, M., 1998. The neural network house: An environment that adapts to its inhab-itants. In: Proc. AAAI Spring Symp. Intelligent Environments.

[41] Muthu, S., Gaines, J., 2003. Red, green and blue LED-based white light source: imple-mentation challenges and control design. In: Industry Applications Conference, 2003.38th IAS Annual Meeting. Conference Record of the. Vol. 1; 1. pp. 515–522 vol.1.

[42] Muthu, S., Schuurmans, F. J., Pashley, M. D., 2002. Red, green, and blue LED basedwhite light generation: issues and control. In: Industry Applications Conference, 2002.37th IAS Annual Meeting. Conference Record of the. Vol. 1; 1. pp. 327–333 vol.1.

[43] Narendran, N., Deng, L., Pysar, R. M., Gu, Y., Yu, H., 2004. Performance characteris-tics of high-power light-emitting diodes. Vol. San Diego, CA, United States. SPIE, pp.267–275.

[44] Narendran, N., Gu, Y., 2005. Life of LED-based white light sources. Display Technology,Journal of 1 (1), 167–171.

[45] Navigent Consulting, Inc., 2009. Solid-state lighting research and development portfolio:Multi-year program plan fy’09-fy’15. Tech. rep., Lighting Research and DevelopmentBuilding Technologies Program, Office of Energy Efficiency and Renewable Energy,U.S. Dept. of Energy, Chicago.

[46] Nelder, J. A., Mead, R., 1965. A simplex method for function minimization. The Com-puter Journal 7 (4), 308–313.

[47] Ohno, Y., 2005. Spectral design considerations for white LED color rendering. Opt.Eng.44 (11), 111302–.

[48] Pacific Northwest National Laboratory, 2009. Summary of results: Round 9 of producttesting. Tech. Rep. 9, Department of Energy.

[49] Park, H., Burke, J., Srivastava, M. B., 2007. Design and implementation of a wirelesssensor network for intelligent light control. In: IPSN ’07: Proceedings of the 6th in-ternational conference on Information processing in sensor networks. Vol. Cambridge,Massachusetts, USA. ACM, New York, NY, USA, pp. 370–379.

[50] Park, S., Lee, D.-H., Kim, Y.-W., Park, S.-N., 2007. Uncertainty evaluation for thespectroradiometric measurement of the averaged light-emitting diode intensity. Appl.Opt. 46 (15), 2851–2858.

[51] Philips Lighting Research, 2010. Dynamic lighting. www.dynamiclighting.philips.com/,Accessed March 2010.

145

Bibliography

[52] Phillips, J., Coltrin, M., Crawford, M., Fischer, A., Krames, M., Mueller-Mach, R.,Mueller, G., Ohno, Y., Rohwer, L., Simmons, J., J.Y.Tsao, 2007. Research challengesto ultra-efficient inorganic solid-state lighting. Laser & Photonics Review 1 (4), 307–333.

[53] Powell, M. J. D., 1998. Direct search algorithms for optimization calculations. ActaNumerica 7 (-1), 287–.

[54] Rains, J., May, D., Ramer, D., February 2006. Optical integrating chamber lightingusing multiple color sources. U.S. Patent Number, 6,995,355.

[55] Ramer, D., Rains, J., Bagwell, R. S., January 2002. Direct view lighting system withconstructive occlusion. U.S. Patent Number, 6,334,700.

[56] Rosenbrock, H. H., 1960. An automatic method for finding the greatest or least valueof a function. The Computer Journal 3 (3), 175–184.

[57] Ryer, A., 1997. The Light Measurement Handbook, 2nd Edition. International Light,Inc., Newburyport, MA 01950-4092.

[58] Sandahl, L., Gilbride, T., Ledbetter, M., Steward, H., Calwell, C., 2006. Compactfluorescent lighting in America: lessons learned on the way to market. Richland WA:Pacific Northwest National Laboratory.

[59] Schubert, E. F., 2006. Innovations in light-emitting diodes. In: Lasers and Electro-Optics and 2006 Quantum Electronics and Laser Science Conference. CLEO/QELS2006. Conference on. pp. 1–2.

[60] Schubert, E. F., Kim, J. K., 2005. Solid-state light sources getting smart. Science308 (5726), 1274–1278.

[61] Schubert, E. F., Kim, J. K., Luo, H., Xi, J.-Q., 2006. Solid-state lighting - a benevolenttechnology. Reports on Progress in Physics 69 (12), 3069–3099.

[62] Sciacchitano, M., September 2009. Test fixture dome assembly. Solidworks 2008 SP0.0.

[63] Sha, L., Gopalakrishnan, S., Liu, X., Wang, Q., 2009. Cyber-physical systems: A newfrontier. pp. 3–13.

[64] Singhvi, V., Krause, A., Guestrin, C., Garrett, Jr, J. H., Matthews, H. S., 2005. In-telligent light control using sensor networks. In: SenSys ’05: Proceedings of the 3rdinternational conference on Embedded networked sensor systems. Vol. San Diego, Cal-ifornia, USA. ACM, New York, NY, USA, pp. 218–229.

[65] Slocum, A., 2005. Policy options to enhance technology diffusion: modeling the green-house gas reduction potential of solid-state lighting. Ph.D. thesis, Rochester Instituteof Technology.

146

Bibliography

[66] Sluis, V., Cortenraad, H. M. R., Diederiks, E. M. A., Hoonhout, H. C. M., Yalvac,M. A., 06 2007. Color matching for display system for shops. WO 2007/072376.

[67] Stanikunas, R., Vaitkevicius, H., Svegzda, A., Viliunas, V., Bliznikas, Z., Breive, K.,Vaicekauskas, R., Novickovas, A., Kurilcik, G., Zukauskas, A., Gaska, R., Shur, M. S.,2005. Polychromatic solid-state lamps versus tungsten radiator: hue changes of munsellsamples. J.Phys.D 38 (17), 3202–3207.

[68] Stevenson, R., 2009. The LED’s dark secret. IEEE Spectrum 46 (8), 26–31.

[69] Teixeira, T., Jung, D., Dublon, G., Savvides, A., 2009. Identifying people in cameranetworks using wearable accelerometers. In: PETRA ’09: Proceedings of the 2nd In-ternational Conference on PErvsive Technologies Related to Assistive Environments.Vol. Corfu, Greece. ACM, New York, NY, USA, pp. 1–8.

[70] Teixeira, T., Jung, D., Dublon, G., Savvides, A., 2009. Pem-id: Identifying peo-ple by gait-matching using cameras and wearable accelerometers. Vol. Como, Italy.ACM/IEEE, pp. 1–8.

[71] The MathWorks, Inc., March 2010. Global Optimization Toolbox 3. The Mathworks,Inc., 3 Apple Hill Drive Natick, MA 01760-2098, 14th Edition.

[72] Thompson, M. d. R., 2007. Psychophysical evaluations of modulated color rendering forenergy performance of LED-based architectural lighting. Ph.D. thesis, MassachusettsInstitute of Technology.

[73] Tor, J., December 2009. Test fixture heatsink. Solidworks 2008 SP0.0.

[74] Torczon, V., 1997. On the convergence of pattern search algorithms. SIAM J.on Opti-mization 7 (1), 1–25.

[75] Trevisanello, L. R., Meneghini, M., Mura, G., Sanna, C., Buso, S., Spiazzi, G., Vanzi,M., Meneghesso, G., Zanoni, E., 2007. Thermal stability analysis of high brightnessLED during high temperature and electrical aging. Vol. 6669. SPIE, pp. 666913–.

[76] van der Broeck, H., Sauerlander, G., Wendt, M., 2007. Power driver topologies andcontrol schemes for LEDs. In: Applied Power Electronics Conference, APEC 2007 -Twenty Second Annual IEEE. pp. 1319–1325.

[77] Webb, A. R., 2006. Considerations for lighting in the built environment: Non-visualeffects of light. Energy Build. 38 (7), 721–727.

[78] Wen, Y.-J., Granderson, J., Agogino, A. M., 2006. Towards embedded wireless-networked intelligent daylighting systems for commercial buildings. In: SUTC ’06:Proceedings of the IEEE International Conference on Sensor Networks, Ubiquitous,and Trustworthy Computing -Vol 1 (SUTC’06). IEEE Computer Society, Washington,DC, USA, pp. 326–331.

147

Bibliography

[79] Wyszecki, G., Stiles, W. S., 1982. Color Science : Concepts and Methods, QuantitativeData and Formulae. Vol. 2. Wiley, New York.

[80] Zachau, M., Becker, D., Berben, D., Fiedler, T., Jermann, F., Zwaschka, F., 2008.Phosphors for solid state lighting. Vol. 6910. SPIE, pp. 691010–.

[81] Zukauskas, A., Vaicekauskas, R., Ivanauskas, F., Gaska, R., Shur, M. S., 2002. Opti-mization of white polychromatic semiconductor lamps. Appl.Phys.Lett. 80 (2), 234–236.

[82] Zukauskas, A., Vaicekauskas, R., Ivanauskas, F., VaitkeviCius, H., Shur, M. S., 2008.Rendering a color palette by light-emitting diodes. Appl.Phys.Lett. 93 (2), 021109–.

148