New Method for Spectral Irradiance and Radiance
Responsivity Calibration using Pulsed Tunable Lasers
The 11th International Conference on New
Developments and Applications in Optical Radiometry
September 19-23, 2011, Maui, Hawaii, USA
National Institute of Standards and Technology
Gaithersburg, Maryland
USA
Yuqin Zong
Steven Brown, George Eppeldauer, Keith Lykke, and Yoshi Ohno
Continuous wave (CW) tunable lasers
Shortcomings:
bulky
expensive
interference fringes
hard to operate and maintain
high power,
narrow bandwidth,
Being used for calibrations of primary detectors and
remote sensing instruments
Pulsed tunable lasers
Narrow pulse width, extremely low duty cycle (eg, 10-6)
Pulse to pulse variation, and hard to be stabilized.
Transimpedence amplifiers don’t work well.
Fully automated
Large tunable range
Finite bandwidth, no or less interference fringes
portable,
affordable.
Have not been used as calibration source yet!
Key questions to be answered
Will detectors be saturated?
Is a pulse laser equivalent a CW laser for
detector calibrations?
How to overcome fluctuation of a pulsed laser and
get repeatable results?
Can pulse lasers be used for calibration of
detectors with small uncertainties?
1000 Hz OPO
laser
Standard
Trap Detector
Test
Detector
Monitor
Detector
Ultrasound
bath
Computer
Schematic of the new measurement method
shutter
integrating
sphere
baffle
Electrometers (2)
Shutter controller
M
standard
M
teststandard /)()( QQRRtest
Optical
fiber
The OPO laser
210 nm to 2400 nm
tunable range,
1000 Hz repetitive
rate
5 ns pulse width
5 – 8 cm-1 bandwidth
Pulse waveform
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
-20 -10 0 10 20
Re
l. la
ser
po
we
r
Time, t (ns)
After 50 mmsphere
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
-20 -10 0 10 20
Re
l. la
ser
po
we
r
Time, t (ns)
After 5 mMM fiber
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
-20 -10 0 10 20
Re
l. la
ser
po
we
r
Time, t (ns)
OPO laser
Laser spectrum
0
0.2
0.4
0.6
0.8
1
1.2
348 349 350 351 352
Re
lati
ve S
PD
Wavelength, λ (nm)
350 nm
0
0.2
0.4
0.6
0.8
1
1.2
598 599 600 601 602
Re
lati
ve S
PD
Wavelength, λ (nm)
600 nm
0
0.2
0.4
0.6
0.8
1
1.2
1098 1099 1100 1101 1102
Re
lati
ve S
PD
Wavelength, λ (nm)
1100 nm
The electrometer
Charge measurement
function from 2 nC to 2 μC
High performance multichannel
switching card
< 3 fA bias current
< 20 μV burden voltage
No accurate timing and switching
Detector
+ -
DMM
C
V
Pulse train
i(t)
VCdttiQT
0 )(
With a switched integrator
transimpedance amplifier
Measurement timing
start
Electrometer’s synchronized
charge measurement
end
Laser shutter
open close
Measurement repeatability
Two S2281 Si photodiodes (PD)
standard deviation = 7 ppm!
0.999950
0.999970
0.999990
1.000010
1.000030
1.000050
0 5 10 15 20 25 30
Re
lati
ve c
har
ge r
atio
Measurement No., i
0.999950
0.999970
0.999990
1.000010
1.000030
1.000050
0 5 10 15 20 25 30
Re
lati
ve c
har
ge r
atio
Measurement No., i
One 3 Si PD trap and one
S2281 Si PD
standard deviation = 12 ppm!
1000 Hz OPO
laser
Test Detector
(S2281 PD)
Test
Detector
Ref. detect
(S2281 PD)
Optical
fiber
Computer
Linearity measurement
shutter
integrating
sphere
baffle
Electrometers (2)
Shutter controller
Ref.test /)( QQPr i
OD 2
attenuator
Result of detector linearity test
The relative responsivity is obtained by normalizing the charge
ratio r(Pi) of the test detector to reference detector .
0.9988
0.9990
0.9992
0.9994
0.9996
0.9998
1.0000
1.0002
1.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05
Re
lati
ve r
esp
on
sivi
ty
Averaged photocurrent, I (A)
OPO laser at 450 nm.
For a SS2281 PD,
Saturation starts at
Peak=100 mA,
Averaged=10-6 µA.
Points with regard to the linearity
1) Nonlinearity depends on the detector and the
laser wavelength.
2) The instantaneous photocurrent without causing
nonlinearity is several orders of magnitude
higher than the threshold nonlinear DC
photocurrent (0.1 – 1 mA typically).
3) The level of allowed averaged photocurrent is
several orders of magnitude lower than the
threshold nonlinear DC photocurrent.
Validation results using CW lasers
Difference in measured responsivity is only ≈ 0.02 %, well
within the instruments’ uncertainty (0.05 %).
0.000
0.010
0.020
0.030
0.040
0.050
400 450 500 550 600 650
R(λ
) D
iffe
ren
ce (
%)
Wavelength, λ (nm)
OPO/CW
gated CW/ CW
Uncertainties
Relative standard unc. (%)
Uncertainty component Type A Type B
Reference trap detector 0.028
Laser wavelength (0.01 nm) 0.005
Sphere source irradiance
uniformity 0.005
Detector reference plane 0.010
Detector linearity 0.005
Transfer to test detector 0.005
Electrometer (relative only) 0.01
Combined uncertainty (%) 0.033
Expanded uncertainty (k=2)
(%) 0.066
Conclusions
A new method using pulsed laser sources has been
developed for calibration of detectors and instruments.
The method has been validated and found to be
equivalent to CW laser method.
The averaged photocurrent should be kept several
orders of magnitude lower than the threshold
nonlinear DC photocurrent to avoid nonlinearity.
Pulsed laser sources have advantage over CW
lasers in reducing interference fringes.
Conclusions
This method can be used in other applications such
as measurement of material property of
transmittance and reflectance.
Compared to a monochromotor-based system,
calibration uncertainties are significantly lower
(eg, one order of magnitude).
THANK YOU