Slide 1

Laser characterization and modelling for the development

of the Super LHC versatile transceiver

Sergio Rui SilvaMarie Curie Fellow PH-ESE-BE

CERN2 April 2009

Slide 2

• An upgrade of the current LHC (super LHC or SLHC), planned for 2013-18, is expected to increase the luminosity by an order of magnitude to 1035 particles/cm2/s.

• A tracking detector operating at the SLHC will require ten times more readout data bandwidth and radiation tolerance than at the current LHC detectors.

• Design a digital transceiver capable of operating at high speed (multiple GBits/s) and endure:– High radiation levels – High magnetic fields– Low temperatures

Project background

Slide 3

• With a device characterization one can: 1. Evaluate the LASER device2. Predict interaction between LASER, driver and

connection network:• Design matching network• Peaking circuit / Pre-Emphasis• DC electrical interface (Bias-T)

3. Predict performance degradation with environment conditions (temperature, radiation)

4. Design better systems

Why characterize LASER devices

Slide 4

Why characterize the LASER devices

• Interaction between LASER, driver and connection network:• Electrical impedance mismatch should be kept < 10%

Laser Driver Bias T Laser PD

+ -

VI

VR VR

VI P

Slide 5

What to characterize in the LASER device

• Laser characterization– Power / Current / Voltage Curves– S-Parameters

• Direct measurements• Static values obtained for a finite number of

device working conditions• Then, using input from characterization, make

model– Predict device behaviour

• Current threshold, BW & wavelength changes– Temp– Radiation

• Device package modifications

Slide 6

Types of LASER devices

• Semiconductor LASER devices:– FP (Fabry-Perot LASER)– DFB (Distributed Feedback LASER)– DBR (Distributed Bragg Reflector LASER)– VCSEL (Vertical Cavity Surface Emitting LASER)– HCSEL (Horizontal Cavity Surface Emitting

LASER)

Slide 7

Measurement setup

• TDR Setup Calibrate hereCooling system

Slide 8

Measurement setup

• VNA Setup Calibrate hereCooling system

VNA

DC Block Microstrip Fixture

LASER PD

Amplifier

Transmission Port

Reflection Port

LASER DC Bias Current

Peltier Element Temperature Sensor

Slide 9

Measurement setup

• VNA Setup

Slide 10

Measurement setup

• Laser assembly

Wired Laser Flex Strip Laser

Slide 11

Measurement setup

• Test fixture calibration set

Open Short Load

Wired Lasers Flex Strip Lasers

Open Short Load

Slide 12

• Testing & de-embedding of the test fixture:

Measurement setup

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1-1.5

-1

-0.5

0

0.5

1

1.5

Time (ns)

Test Fixture

OpenShortLoad

0 5000 10000 15000-30

-25

-20

-15

-10

-5

0

Frequency (MHz)

dB

S11

Open (W ire)Short (W ire)Load (W ire)Open (Flex)Short (Flex)Load (Flex)

0 5000 10000 15000-70

-60

-50

-40

-30

-20

-10

0

10

20

Frequency (MHz)

dB

One-Port, 3- Term Error (Diclad)

e00e11ee10e01

3-Term Error Matrix

Slide 13

Measurement setup

• Temperature control system:

Peltier based thermo-electric laser cooling system

Slide 14

Measurement setup

• Temperature control system:

Thermal image of the laser & cooling system set to 5°C

Slide 15

LASER model

• Measurements lead to simplified laser model– Chip & Package model– Intrinsic Laser

• Large signal model– DC model– Large modulation currents

• Small signal– Signal model– Small modulation currents

Slide 17

LASER model

• Model used

Source Chip & Package Intrinsic LaserTest Fixture

ILD

LP2 RP2 RS

Z0

L

CP2

CSUB

RSUB

ROutIS

LP1 RP1

CP1

Slide 18

LP2 RP2 RS

CP2

CSUB

RSUB

LP1 RP1

CP1

ZIn

ILD

LASER model

• Parasitic circuit model

Bias current should have a negligible

effect in S11: extract the parasitic model

parameters using this data

For IBias > Ith , |ZILD| << RS

Slide 19

LASER model

• Parasitic circuit simplified modelRS

CS

LP

ZIn

Parameters can be calculated directly from measurement data

Slide 20

LASER model

• ILD model: rate equations (semiconductor LASERs):

Carrier densityPhoton density

Optical power

Optical phase

• Model ILD using:• Explicitly through differential equations• Linear approximation at bias point

• For small signal model• Simpler simulations

Slide 21

• Intrinsic Laser Diode Modelling

The ILD parameters can be extracted

without the influence of the package

parasitic effects

Ref

( , ) ( , ) ( ) ( ) ( , )( , ) ( , ) ( ) ( ) ( , )

Global Bias ILD Bias PC TF ILD Bias

Global ILD Ref PC TF ILD Ref

H f I H f I H f H f H f IH f I H f I H f H f H f I

LASER model

Slide 22

• Laser Model Parameter ExtractionIntrinsic Laser

Diode Parameter Extraction

Test Fixture De-embedding

Parasitic Circuit Parameter Extraction

Global Response Parameter Fit

S-Parameters

ILD Parameters

PC Parameters

Complete Model Parameters

S21(f, IBias)

S11(f, IBias)

S(f, IBias)

Open, Short & Load Test Fixture S-Parameters

LASER model

Slide 23

• Laser model : Input Impedance

(VL-1310-5G-P2-P4\108725-07)

LASER model

Good fit in both magnitude and phase

over the frequency range of interest

0 1 2 3 4 5 6 7 8 9 100

50

100

150

200

250

300

350

400

450

500|Zin|: Measured & Modelled

|Zin

| (

)

Frequency (GHz)

MeasuredDe-embeddedModelled (Simple)Modelled

0 1 2 3 4 5 6 7 8 9 10

-150

-100

-50

0

50

100

150

Angle(Zin): Measured & Modelled

Ang

le(Z

in) (D

eg)

Frequency (GHz)

MeasuredDe-embeddedModelled (Simple)Modelled

Slide 24

• Laser model : Input Impedance

(VL-1310-5G-P2-P4\108725-07)

LASER model

0 5 10 150

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6I-V Curve

Current (mA)

Vol

tage

(V

)

RS=24 Ω (extracted from S11)RS=28 Ω (extracted from I-V)

Agreement of DC

internal resistance

values between

measures

Slide 25

• Laser model : Intrinsic Laser Model

(VL-1310-5G-P2-P4\108725-07)

LASER model

Good agreement

between the curves

obtained with the model and the

measured data

0 1 2 3 4 5 6 7 8 9 10-5

0

5

10

Intrinsic Laser Model Fit (H(f, IB)/H(f,I1))

Am

plitu

de (d

B)

Frequency (GHz)

Data I2/I1Data I3/I1Data I4/I1Model I2/I1Model I3/I1Model I4/I1

Slide 26

0 1 2 3 4 5 6 7 8 9 10-35

-30

-25

-20

-15

-10

-5

0

5Measurement data & Complete Laser Model

Am

plitu

de (d

B)

Frequency (GHz)

Data I1Data I2Data I3Data I4Model I1Model I2Model I3Model I4

• Laser model : Global model (Transfer function)

(VL-1310-5G-P2-P4\108725-07)

LASER model

Good model fit:

• Clear dependence on

bias• Bandwidth increases as

bias increases

IBias

Slide 27

• Laser model : Global model (I-L Curve)

(VL-1310-5G-P2-P4\108725-07)

LASER model

Model is able to predict

back a quantity that

was not included in

the model fit

0 5 10 15-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9LIV Curve

Current (mA)

Pow

er (m

W)

ModelledMeasured

Slide 28

On going work

• Model tuning (evaluate model with transceiver designer):

Simulation using the developed Verilog-A implementation of the laser model in

ADS

Slide 29

On going work

• Laser parameter extraction using noise (RIN) measurements:

0 500 1000 1500 2000 2500 3000-82

-80

-78

-76

-74

-72

-70

-68

-66

Am

plitu

de (d

B)

Frequency (MHz)

Laser RIN (Vertilas VCSEL 108725-06)

IB=0.0mA

IB=7.5mA

IB=8.0mA

IB=8.5mA

IB=9.0mA

IB=9.5mA

(VL-1310-5G-P2-P4\108725-07)

Slide 30

On going work

• Microstrip Bias-T:

Slide 31

Conclusion

• A laser model suitable for large variety of laser was developed.

• Accurate modelling of the input impedance was achieved:– Essential for designing matching network and

bias-T.• Good agreement obtained between the ILD model

and the measured data:– Transfer-function;– L-I curve;– Eye-Diagram.

• Development of good-practices in HF PCB layout.Thank you for your attention!