Download Miteq DR-125G-A, 12-GHZ Fiber

DOE/NV/25946--034
DOE/NV/11718--1276
MITEQ DR-125G-A, 12-GHZ FIBER-OPTIC DETECTOR EVALUATIONS FOR
THE PHOTONIC DOPPLER VELOCIMETRY DIAGNOSTIC
Araceli Rutkowski, Michael Rutkowski1
National Security Technologies, LLC, Los Alamos Operations
The 12.5-GHz bandwidth MITEQ DR-125G-A detector is used often in the photonic Doppler velocimetry
(PDV) diagnostic of Los Alamos National Laboratory (LANL). This paper presents detector characteristics
as applied to optical heterodyning. We propose a test setup to simulate the beat frequency generated when
incident and reflected light from a moving surface are mixed by optically combining a small-linewidth,
tunable laser with a fixed, 1550-nm, thin-linewidth, high-power laser. The detector’s transfer function,
harmonic content, and signal-to-noise ratio (SNR) were to be measured and plotted for different optical
power levels. Based on these results, appropriate light levels can be set to produce the highest dynamic range
and signal level for the beat frequency. The goal is to provide insight into setting up the diagnostic for
optimal detector performance using a specific optical input power.
I. Introduction
The PDV diagnostic has been deployed by LANL quite often in the last six months, with very good results. Setting
up the PDV detector for optimal performance has been largely a trial-and-error process, in which each configuration
has its share of unique issues. The MITEQ DR-125G-A detector consists of a PIN diode and a high-gain
transimpedance amplifier. Its total responsivity is approximately 1900 V/W. The detector’s 3-dB bandwidth ranges
from 10 kHz to 12.5 GHz, and maximum output can reach up to 3.5 V. Each channel in the detector module system
is often configured similarly to the block diagram shown in Figure 1. A high-power laser of 5 to 10 W is split to
produce and input optical power from 0.5 to 2 W at 1550 nm. An inline optical power meter monitors source power
with only a 0.3-dB loss. The source light (also called “unshifted light”) enters the optical circulator with a 0.77-dB
loss to Port 2, then onto the 90/10 fused coupler. The through leg of the coupler has a 0.72-dB loss and is used as the
output to the probe. The tap leg of the fused coupler has 11.5 dB of loss and generates unshifted light back into the
detector. The unshifted light travels into the optical attenuator, with 1 dB of loss at a 0-dB attenuation setting, to the
0.6-dB-loss, gold-capped optical reflector. From the reflector, the signal travels back to the attenuator to the tapped
leg of the fused coupler (another 11.5 dB of loss) to Port 2 of the optical circulator. The unshifted light goes through
to Port 3 (0.77 dB of loss), to a second power meter (0.41 dB of loss), to the MITEQ detector. The user can adjust
the amount of unshifted light at the detector using the optical attenuator.
Figure 1. Typical PDV block diagram
1
[email protected], (505) 663-2074
1
The incident light travels down some length of single-mode fiber (this may not always be the case but is desired).
As the surface moves, a Doppler-shifted light is reflected back up the single-mode fiber to the through leg of the
fused coupler (0.72 dB of loss). The Doppler-shifted light (also referred to as “shifted light”) enters optical circulator
Port 2 and exits Port 3 with a 0.77-dB lower signal. It travels through the second inline power meter (0.41 dB of
loss) to the MITEQ detector.
The interference between the shifted and unshifted light produces a frequency at the detector output. Unshifted light
at the detector can range from an estimated 0.5 to 20 mW, while shifted light can vary from 10 nW to 10 µW.
II. Test Setup
A test configuration was needed to generate both shifted and unshifted optical signals in the lab without having to
use the Doppler-shifted light reflected from a moving surface. The basic requirements for the optical sources were:
adjustable optical output power, good optical power and wavelength stability and a maximum wavelength difference
between the optical sources of a few picometers. The test configuration shown in Figure 2 was set up and evaluated
for use to characterize the MITEQ DR-125G-A detector. An IPG Photonics ELR-5-1550-SF laser generated the
unshifted light. The randomly polarized laser has a 50-kHz linewidth (FWHM). Its output power is
5 W, scalable down to a few mW. This laser is very similar to the laser source utilized in the PDV diagnostic. To
generate the shifted light, we used the Agilent 81940A Compact Tunable Laser Source. This laser has a linewidth
of 100 kHz (FWHM) and a wavelength stability of ±2.5 pm over 24 hours. Its wavelength resolution is 1 pm at
1550 nm. Optical attenuators were connected to the output of each laser to adjust their output power levels. The IPG
Photonics laser was connected to the JDS Uniphase VA4 optical attenuator, while the Agilent tunable laser was
connected to the HP 8158B optical attenuator. The attenuators’ outputs were fed into a JDS Uniphase 50/50 fused
optical coupler. An EigenLight M320 series power meter was placed in line with the detector under test (DUT) to
minimize fiber disruption during data recording. The EigenLight meter has an absolute accuracy of ±0.2 dB. The
output of the DUT was connected to the Agilent E4440A high-performance spectrum analyzer, with a 26.5-GHz
bandwidth and 140 dB of dynamic range. All fibers were single-mode SMF-28 9/125-µm Corning fiber. All fiber
connections were FC/APC.
Figure 2. Test-setup block diagram
The IPG laser had a measured wavelength of 1549.944 nm, while the tunable laser was set to 1549.974 nm. These
settings produced a beat frequency of 3.8 ±0.3 GHz. Spectrum analyzer settings were:
•
•
•
•
Video bandwidth: 50 MHz
Resolution bandwidth: 8 MHz
Frequency span: 1 GHz
Center frequency: 4 GHz.
2
Because the 3.8-GHz signal amplitude varied by 2 to 3 dB, we used the analyzer’s Max Hold feature to obtain the
reading. Due to polarization mismatch, the beat signal proved very sensitive in amplitude as the optical fiber cable
was moved. It was stabilized by taping all fiber jumpers to the lab bench.
III. Transfer Function
To determine the optimum value of unshifted light into the detector, we measured the root-mean-square output
voltage vs. unshifted optical power while holding the shifted light constant. Figure 3 shows the resulting transfer
curves of the MITEQ detector, S/N#1092539, when the shifted light was held constant at 100 nW, and the unshifted
light was adjusted from 0 to 16 mW. There was good correlation between the calculated and measured data, until the
optical power reached 10 mW. From 10 to 11.5 mW, the amplitude leveled off to 78 mV, then declined to 40 mV, at
an input power of 16 mW. This data supports the conclusion that there is an optimum optical level at which one can
achieve maximum output voltage.
MITEQ 1092539 OUTPUT VS. UNSHIFTED LIGHT POWER WITH A FIXED SHIFTED LIGHT AT
100 nW
120
MITEQ OUTPUT (mV-RMS)
100
Vout = Responsivity *2
(Power shifted * Power unshifted)
80
Measured
MITEQ Vout
60
40
20
0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
OPTICAL POWER (mW)
OUTPUT VOLTAGE (mV rms)@ 100nW SHIFTED LIGHT
12.0
13.0
14.0
15.0
16.0
CALCULATED OUTPUT VOLTAGE (mV rms)@ 100nW SHIFTED LIGHT
Figure 3. MITEQ transfer function calculated vs. measured
The graph in Figure 4 shows the same MITEQ detector with the output radio frequency (RF) level at four different
shifted-light levels. In this case, the data demonstrates that the optimum input optical power level occurs at the same
point independently of the shifted-light level. Figure 4 also demonstrates how quickly the output voltage declines
with decreasing shifted-light levels. The graph in Figure 5 shows a closer view of the MITEQ transfer function at
lower unshifted light power levels.
3
MITEQ 1092539 OUTPUT VS. UNSHIFTED POWER
800
700
output (mV rms)
600
500
400
300
200
100
0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
unshifted power (mW)
OUTPUT VOLTAGE (mV rms)@ 10nW SHIFTED LIGHT
OUTPUT VOLTAGE (mV rms)@ 100nW SHIFTED LIGHT
OUTPUT VOLTAGE (mV rms)@ 1uW SHIFTED LIGHT
OUTPUT VOLTAGE (mV rms)@ 10uW SHIFTED LIGHT
Figure 4. MITEQ measured transfer function for different shifted optical powers
MITEQ 1092539 OUTPUT VS. UNSHIFTED POWER (lower limit)
400
350
output (mV-rms)
300
250
200
150
100
50
0
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
2.20
2.40
unshifted power (mW)
OUTPUT VOLTAGE (mV rms)@ 10nW SHIFTED LIGHT
OUTPUT VOLTAGE (mV rms)@ 100nW SHIFTED LIGHT
OUTPUT VOLTAGE (mV rms)@ 1uW SHIFTED LIGHT
OUTPUT VOLTAGE (mV rms)@ 10uW SHIFTED LIGHT
Figure 5. MITEQ transfer function at lower unshifted optical power
4
2.60
The graph in Figure 6 plots a second MITEQ detector S/N#1104469. This data show that, under the same test
conditions, its output level at 10-µW shifted light (white trace) is ~30% higher (~250 mV) than the output signal of
the MITEQ S/N#1092539 (see Figure 5). Notice that the peaking of the output signal on S/N#1104469 occurs at
3.5 mW lower than that of the S/N#1092539. Based on this information, a total of four MITEQ detectors were
evaluated under the same test conditions. Results appear in Figure 7. This graph shows that each MITEQ detector
has its own “operational sweet spot” for optimum output voltage versus input optical power. This “spot” varies from
7 to 12 mW of input power and 80 to 120 mV of output voltage, or possibly even more.
MITEQ S/N# 1104469 OUTPUT VOLTAGE VS. UNSHIFTED OPTICAL POWER (AT
DIFFERENT SHIFTED OPTICAL POWERS)
1.0E+0
900.0E-3
output voltage (volts-rms)
800.0E-3
700.0E-3
600.0E-3
500.0E-3
400.0E-3
300.0E-3
200.0E-3
100.0E-3
000.0E+0
1.0E-3
3.0E+0
6.0E+0
9.0E+0
12.0E+0
15.0E+0
18.0E+0
unshifted optical power (mW)
10nW shifted opt pwr
100nW shifted opt pwr
1uW shifted opt pwr
10uW shifted opt pwr
Figure 6. Second MITEQ detector transfer function at different shifted optical powers
5
MITEQ OUTPUT VOLTAGE VS. UNSHIFTED OPTICAL POWER AT SHIFTED POWER OF 100 nW
(FOR 4 MITEQ RECEIVERS)
140.0E-3
VOLTAGE (Volts-rms)
120.0E-3
100.0E-3
80.0E-3
60.0E-3
40.0E-3
20.0E-3
000.0E+0
000.0E+0
2.0E-3
4.0E-3
6.0E-3
8.0E-3
10.0E-3
12.0E-3
14.0E-3
16.0E-3
18.0E-3
20.0E-3
22.0E-3
24.0E-3
UNSHIFTED OPTICAL POWER (Watts)
S/N# 1092520
S/N# 1092543
S/N# 1092539
S/N# 1104469
Figure 7. Transfer function for 4 different MITEQ detectors
IV. Harmonic Levels
The second harmonic of the beat frequency was plotted for two different shifted-light levels. Figure 8 shows the
fundamental frequency and the second harmonic levels. It also presents the difference (delta) in decibels between the
fundamental and the second harmonic at a fixed shifted-light level of 10 µW. Figure 9 illustrates the second
harmonic level at a 1-µW fixed shifted-light level. As the fundamental signal approaches the compression region,
the relative harmonic amplitude will start to increase faster than the output. In Figure 8, the signal starts entering the
compression region at 5 mW of input power, with a shifted-light level of 10 µW. In Figure 9, the signal begins
entering the compression region at 9 mW. Depending on application, a maximum desirable second harmonic level is
between 35 and 40 dBc. Figures 9 and 10 demonstrate that, for the same unshifted optical power, the harmonic
distortion level is proportional to the shifted optical power. In addition, the second harmonic appears highest at the
compression point.
6
MITEQ 1092539 OUTPUT (4-GHz), 2ND HARMONIC LEVEL (8-GHz) AND 2ND HARM.
DELTA VS. UNSHIFTED OPTICAL PWR @ 10-uW SHIFTED
800
45
40
700
OUTPUT & 2ND HARM ONIC
LEVEL (m V-rm s)
HARMONIC DELTA (dBc)
35
600
30
500
25
400
20
300
15
200
10
100
5
0
0.000
2.000
4.000
6.000
8.000
10.000
12.000
14.000
0
16.000
OPTICAL PWR (mW)
output(mV)
2nd harmonic level (mV)
2nd Harmonic DELTA (dBc)@ 10uW SHIFTED LIGHT
Figure 8. MITEQ harmonic distortion at 10-µW shifted power
MITEQ 1092539 OUTPUT (4GHz),2ND HARMONIC LEVEL(8GHz) AND 2ND
HARM. DELTA VS. UNSHIFTED OPTICAL PWR @ 1uW SHIFTED
300
55
50
45
40
200
35
30
150
25
20
100
15
10
50
5
0
0.000
2.000
4.000
6.000
8.000
10.000
12.000
14.000
0
16.000
OPTICAL PWR (mW)
output (mV)
2nd harmonic level (mV)
2nd Harmonic DELTA (dBc)@ 1uW SHIFTED LIGHT
Figure 9. MITEQ harmonic distortion at 1-µW shifted power
7
HARMONIC DELTA (dBc)
OUTPUT,2ND HARMONIC LEVEL
(mV-rms)
250
V. Signal-to-Noise Evaluation
The SNR was first normalized over a 1-Hz bandwidth. The same four fixed, shifted-light levels were used (see
results in Figure 11). The data show the SNR peaking at the unshifted-light level of 0 dBm. Figures 4 and 10
indicate that the optical input power for highest output voltage is not the same as the optical power level for
optimum SNR. Therefore, a trade-off exists between output level and SNR. Figures 11–14 show the same SNR plots
as Figure 10, except that the bandwidth was changed from 1 Hz to 1 GHz, 4 GHz, 8 GHz, and 12 GHz, respectively.
Bandwidth plays an important role in optimizing the SNR for low shifted-light signals. Note that the peak SNR of
the 100-nW curve is 12 dB in a 4-GHz bandwidth (Figure 12). When the bandwidth doubles to 8 GHz, the SNR for
the same 100-nW curve reduces to 9 dB (Figure 13). Generally, the SNR will increase 3 dB for a 50% reduction in
operating bandwidth. Figure 11 shows that in order to optimize the SNR for this particular MITEQ unit, the
unshifted optical power level should fall within the range of 0 to –2.5 dBm.
MITEQ 1092539 SNR (1Hz BW) VS. UNSHIFTED POWER
130
SNR (dB)
120
110
100
90
80
-27.5
-25
-22.5
-20
-17.5
-15
-12.5 -10 -7.5
-5
-2.5
UNSHIFTED POWER (dBm)
0
2.5
5
7.5
10
SNR (dB/Hz)@ 10nW SHIFTED LIGHT
SNR (dB/Hz)@ 100nW SHIFTED LIGHT
SNR (dB/Hz)@ 1uW SHIFTED LIGHT
SNR (dB/Hz)@ 10uW SHIFTED LIGHT
Figure 10. MITEQ signal-to-noise ratio in 1-Hz bandwidth
8
12.5
MITEQ 1092539 SNR 1-GHz VS. UNSHIFTED POWER
40
36
32
SNR (dB)
28
24
20
16
12
8
4
0
-27.5
-25
-22.5
-20
-17.5
-15
-12.5 -10
-7.5
-5
-2.5
UNSHIFTED POWER (dBm)
0
2.5
5
7.5
10
SNR (dB/1GHz)@ 10nW SHIFTED LIGHT
SNR (dB/1GHz)@ 100nW SHIFTED LIGHT
SNR (dB/1GHz)@ 1uW SHIFTED LIGHT
SNR (dB/1GHz)@ 10uW SHIFTED LIGHT
Figure 11. MITEQ signal-to-noise ratio in 1-GHz bandwidth
MITEQ 1092539 SNR 4GHz VS. UNSHIFTED POWER
32
28
24
SNR (dB)
20
16
12
8
4
0
-27.5
-25
-22.5
-20
-17.5
-15
-12.5
-10
-7.5
-5
-2.5
UNSHIFTED POWER (dBm)
0
2.5
5
7.5
10
SNR (dB/4GHz)@ 10nW SHIFTED LIGHT
SNR (dB/4GHz)@ 100nW SHIFTED LIGHT
SNR (dB/4GHz)@ 1uW SHIFTED LIGHT
SNR (dB/4GHz)@ 10uW SHIFTED LIGHT
Figure 12. MITEQ signal-to-noise ratio in 4-GHz bandwidth
9
12.5
12.5
MITEQ 1092539 SNR 8GHz VS. UNSHIFTED POWER
32
28
24
SNR (dB)
20
16
12
8
4
0
-27.5
-25
-22.5
-20
-17.5
-15
-12.5
-10
-7.5
-5
-2.5
UNSHIFTED POWER (dBm)
0
2.5
5
7.5
10
SNR (dB/8GHz)@ 10nW SHIFTED LIGHT
SNR (dB/8GHz)@ 100nW SHIFTED LIGHT
SNR (dB/8GHz)@ 1uW SHIFTED LIGHT
SNR (dB/8GHz)@ 10uW SHIFTED LIGHT
12.5
Figure 13. MITEQ signal-to-noise ratio in 8-GHz bandwidth
MITEQ 1092539 SNR 12-GHz VS. UNSHIFTED POWER
28
24
S N R (dB )
20
16
12
8
4
0
-27.5
-25
-22.5
-20
-17.5
-15
-12.5 -10 -7.5
-5
-2.5
UNSHIFTED POWER (dBm)
0
2.5
5
7.5
10
SNR (dB/12GHz)@ 10nW SHIFTED LIGHT
SNR (dB/12GHz)@ 100nW SHIFTED LIGHT
SNR (dB/12GHz)@ 1uW SHIFTED LIGHT
SNR (dB/12GHz)@ 10uW SHIFTED LIGHT
Figure 14. MITEQ signal-to-noise ratio in 12-GHz bandwidth
10
12.5
VI. Conclusion
Four MITEQ DR-125G-A detectors were lab-tested and evaluated for transfer function, second harmonic level, and
SNR. The transfer function curves show that an optimum input optical power level will yield a maximum output
voltage. If this optimum optical input power is exceeded, the detector will go into compression (output signal will
decrease for increasing optical input power). This will also produce high second harmonic distortions (10–20 dB
down from the beat frequency) and reduce the SNR, consequently decreasing the dynamic range. The optimum
optical power level for output voltage differs from that of an optimum SNR, thus necessitating a trade-off between
output and SNR. The optimum optical input level for output voltage varies for each MITEQ detector; it may also be
temperature-dependent. Thus, in certain applications, a detector screening may need to be performed. If the
detector’s operating point remains within the linear portion of the curve (input optical power at 1 mW), the SNR and
second harmonic distortion will be optimal. An external postamplifier can be used to achieve an output signal high
enough for the digitizers.
Operating the detectors at the full 12.5-GHz bandwidth may work acceptably for a shifted-light level >1 µW.
Sensitivity and dynamic range are usually a concern, particularly for shifted-light levels below 1µW; in such cases
the operational bandwidth may need to be reduced with a filter. By halving the bandwidth, the SNR will increase by
3 dB. This, along with proper biasing, will greatly improve detector sensitivity and dynamic range.
Copyright. This manuscript has been authored by Bechtel Nevada and National Security Technologies under Contract Nos.
DE-AC08-96NV11718 and DE-AC52-06NA25946 with the U.S. Department of Energy. The United States Government retains
and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a nonexclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others
to do so, for United States Government purposes.
Disclaimer. This paper was prepared as an account of work sponsored by an agency of the U.S. Government. Neither the U.S.
Government nor any agency thereof, nor any of their employees, nor any of their contractors, subcontractors or their employees,
makes any warranty or representation, express or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not
infringe privately own rights. Reference herein to any specific commercial product, process, or service by trade name, trademark,
manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the U.S.
Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those
of the U.S. Government or any agency thereof.
11