Академический Документы
Профессиональный Документы
Культура Документы
DU T
2
Introduction
Contents Polarization-dependent loss
(PDL) is a major concern for
Introduction 2
scientists and engineers because
analog and digital signals are
Description of PDL 3
adversely affected by too much
or too little PDL. The PDL of
Typical PDL Values 3
optical components and systems
can be accurately measured
HP 11896A Polarization
using the measurement methods
Controller Description 4
described in this product note.
The HP 11896A Polarization
Selecting the Proper
Controller is an important part
PDL Measurement Method 5
of these measurement methods.
Polarization-Dependent This product note discusses
Loss Measurements the concept of PDL and gives
• Single-Wavelength, High Accuracy Data: examples of typical values that
HP 8153A/HP 11896A are expected for common light-
Power Meter PDL Test System 6 wave components.
Component PDL
• 0.8 dB –0.6 dBm
Figure 1. Polarization-
dependent loss causes
power variations.
4
Select the proper Among the three methods offered, The Swept-Wavelength Method
PDL measurement the Power Meter Method offers delivers high-speed measure-
the highest single-wavelength ments of PDL across a specified
method PDL measurement accuracy. wavelength span. This method
Table 2 shows the typical PDL This method is well suited for is ideal for wavelength-division-
measurement capabilities of each general-purpose applications. multiplexing applications where
method compared to possible single-wavelength PDL infor-
The Power Max./Min. Method
performance requirements. The mation is not sufficient.
provides single-wavelength PDL
power and PDL ranges shown
data plus polarization state Note that the PDL Measurement
in this table will be referred to
information about which relative Ranges shown in Table 2 relate
throughout this document. Use
states of polarization create to the Measurement Accuracy.
this table to decide which mea-
maximum and minimum power PDL values greater than the
surement method is most appro-
points. These power points can be ranges shown can be measured
priate for a specific application.
recreated and device performance however measurement accuracy
adjusted using the assistance of may decrease and measurement
Table 2. a Poincare sphere display and time (also known as scan time)
PDL measurement Stokes parameter data. will increase.
method performance
comparison
Measurement Method
Swept-Wavelength Method
Power Meter Method Power Max/Min Method Swept-Wavelength Method for WDM Components
Performance Requirement HP 8153A/11896A HP 8509/11896A HP 71451B Option 002/003 HP 71452B/11896A
Measurement Accuracy
Worst-Case Value <±0.011 dB <±0.1 dB <±0.1 dB <±0.1 dB
RSS Value <±0.007 dB <±0.06 dB <±0.05 dB <±0.06 dB
(Typical repeatability) (±0.001 dB) (<±0.01 dB) (±0.01 dB) (±0.01 dB)
PDL Measurement Range
(based on accuracy) 0 to 3.0 dB 0 to 30.0 dB 0 to 5.0 dB 0 to 2 dB
Wavelength Range 1250 to 1600 nm 1250 to 1600 nm 1250 to 1600 nm 1525 to 1565 nm
DUT Output Power Range +3 to –50 dBm +10 to –55 dBm 0 to –65 dBm/nm –10 to –70 dBm/nm
Approximate Test Time 10 sec. 10 sec. to 1.5 min. 5 min. 1 to 5 min.
Automatically-Swept Wavelength No No Yes Yes
Polarization Information No Yes No No
6
% PDL Error
uncertainty contribution. The 40
PDL measurement system is
percent-PDL uncertainty equa- 30
the sum of the measurement
tions shown in Table 4 are pro- 20
system’s inherent uncertainty
vided for estimating scan time 10
(values shown in Table 3) and
(also known as measurement 0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
the uncertainty created by the
time) uncertainty contributions Seconds of Polarization Scan Time
measurement’s finite scan time
for a given scan time. Figure 7
relative to the PDL of the test
shows how percent-PDL uncer- Figure 7. Typical
device (see Table 4). Below is an
tainty decreases with increased HP 8153A/11896A
explanation of these terms and PDL polarization
scan time.
an uncertainty calculation. scan time uncertainty
contribution (for power
It is assumed that the DUT Table 3. HP 8153A/11896A measurement averaging
output power range is between PDL Measurement System time T=20 msec only)
uncertainty contribution
+3.0 to –50 dBm and the PDL calculation
value is <3.0 dB, as shown in
Element Description Value Example of
Table 2. Actual system perfor-
uncertainty calculation:
mance will vary depending upon Source drift (10 sec. test time) <0.001 dB Below is an example for calcu-
the individual performance of
HP 11896A loss variation <0.004 dB lating the worst-case uncertainty
the specific equipment in the
Effective power sensor PDL <0.005 dB for an HP 8153A/HP 11896A
system. Refer to the Appendix
PDL measurement of 0.3 dB.
for additional ways to ensure Power sensor linearity <0.001 dB
Values shown are derived from
maximum PDL measurement Connector pair PDL <0.02 dB information in Tables 3 and 4.
accuracy from a given PDL
measurement system. Measurement System
Uncertainty Contribution Table 5. Uncertainty
contribution for an
Inherent PDL Worst case HP 8153A/11896A PDL
Fusion spliced 0.011 dB Measurement System
measurement system
Connectors 0.031 dB
uncertainty contribution: Measurement Conditions
Individual elements in a PDL RSS case
Fusion spliced 0.007 dB DUT’s measured PDL value 0.3 dB
measurement system randomly Connectors 0.021 dB
combine to create a system Power measurement averaging time (T) 20 msec
uncertainty contribution value Scan time (ST) 10 sec
as shown in Table 3. Table 4. Relationships
between HP 8153A Uncertainty Contributions
PDL measurement time and Power Meter averaging
time (T), HP 11896A System uncertainty:
scan rate relationships for scan rate and polariza- Worst case 0.011 dB
uncertainty contribution: tion scan time RSS case 0.007 dB
Various PDL measurement con- Scan time uncertainty
ditions may require different % PDL (=PDL value x 5%) 0.015 dB
Optical Uncertainty
power measurement averaging Power Meter versus Measurement Uncertainty
times in order for the HP 8153A Averaging Scan Scan Scan Time
Lightwave Multimeter to Time (T) Rate Time (ST in sec.) Worst case 0.026 dB
achieve optimum performance. 20 msec 5 10 sec (50/ST)% RSS case 0.022 dB
Table 4 shows the relationships
50 msec 4 25 sec (125/ST)%
between the averaging time, the
HP 11896A Polarization 100 msec 3 50 sec (250/ST)%
Controller scan rate and the 200 msec 2 100 sec (500/ST)%
overall polarization scan time.
8
HP 8509/11896A
power max./min.
PDL test system
The HP 8509/HP 11896A
Power Max./Min. PDL test
system is a semi-automatic,
single-wavelength PDL measure-
ment system. It provides a PDL
value and absolute maximum
and minimum power values and
polarization information (see
Figure 8).
The Power Max./Min. PDL
measurement method is similar
to the HP 8153A/HP11896A
power meter method discussed
in this document. During this
measurement, the HP 8509A or
HP 8509B Polarization Analyzer
monitors the average output Figure 8. Typical display
power and state of polarization. for HP 8509/11896A
PDL Markers are automatically Power Max./Min. PDL
Measurement System
placed on the Poincare sphere
(see Figure 8) to indicate the
points at which optical power Figure 9. Setup
maximum and minimum occur. for PDL measure-
The respective Stokes parameters ments using the
HP 8509B Lightwave HP 8509/11896A
for the Max. and Min. points Polarization Analyzer Power Max./Min.
are also shown in Figure 8. The PDL measure-
polarization state of the test sig- ment system
nal is automatically adjusted
using the HP 11896A Polarization Optical Optical
HP 11896A Opt 025 Output Input
Controller shown in Figure 9. Polarization Controller
2. Select Reset. This is the angle between the Power sensor PDL 0.02 dB
Max. and Min. PDL markers on Power linearity 0.05 dB
HP 11896A settings (1/1): the HP 8509 Poincare sphere
3. Press SCANRATE. Connector pair PDL <0.02 dB
display.
4. Use the far right knob to Measurement Uncertainty
HP 8509 settings (1/2):
select SCAN RATE: 5 and press
7. The measurement is complete. Worst case 0.104 dB
ENTER. This scan rate is slow
PDL data is shown in the “Max- RSS case 0.058 dB
enough for the measurement
Min Power” data box (see
rate of the HP 8509 yet fast
Figure 8). It is assumed that the PDL of
enough to perform an accurate
measurement in a short amount 8. Begin another measurement the DUT is less than 3.0 dB and
of time. by selecting Reset and return- the HP 8509 optical input power
ing to step #5. is between +10 to –55 dBm at all
5. Press AUTOSCAN.
times during the measurement,
as shown in Table 2.
Refer to the Appendix for addi-
tional ways to ensure maximum
possible PDL measurement
accuracy from a given PDL
measurement system.
10
HP 71451B settings (2/2): allow the maximum and mini- Additional information about
13. PDL across the measured mum traces (shown in Figure 10) swept-wavelength HP 71451B
wavelength range is automati- to stabilize. Actual system per- settings and polarization
cally calculated and graphically formance will vary depending scan rate
displayed by pressing CALC+ upon the individual performance
Typical HP 71451B Option 002
DISPLAY. Press AUTO SCALE of the specific equipment in the
(white light source option)/003
and marker functions as needed system.
(swept PDL measurement kit
to optimize the view of the mea-
Refer to the Appendix for addi- with HP 11896A included) set-
surement trace.
tional ways to ensure maximum tings are shown below. The PDL
14. To test additional devices, possible PDL measurement down loadable program (included
connect a new test device as accuracy from a given PDL with Option 003) automatically
shown in Figure 12 and return measurement system. sets values to within these typical
to step #8. ranges.
Table 7. HP 71451B
Uncertainty analysis Option 002/003 and Wavelength range 1250 to 1600 nm
HP 11896A Swept-Wavelength
Measurement wavelength span 100 to 350 nm
PDL measurement uncertainty PDL Measurement System
for the HP 71451B/HP 11896A uncertainty calculation Reference level ≤50 dBm
Swept-Wavelength PDL Test Resolution bandwidth 10 nm
Element Description Value Video bandwidth 30 to 300 Hz
System is based upon the inher-
ent uncertainty of the measure- White light source drift <0.01 dB (typical)
ment system (values shown in Under these conditions and
HP 11896A loss variation <0.004 dB
Table 7). an HP 11896A Polarization
OSA Detector PDL <0.04 dB (typical) Controller scan rate of 1, the
It is assumed that the HP 71451B Scale fidelity <0.02 dB (typical) HP 71451B is able to make an
Option 002 white light source entire sweep at approximately
Signal digitization <0.01 dB
is used and the PDL value is the same polarization state. Over
between 0 to 5 dB, as shown Connector pair PDL <0.02 dB (typical)
a five minute period, a sufficiently
in Table 2. The HP 11896A Measurement Uncertainty large sampling of random polar-
Polarization Controller scan ization states can be measured
time should be long enough to Worst case 0.104 dB
in order to accurately measure
RSS case 0.051 dB PDL over a specified wave-
length range.
12
HP 83438A/HP 71452B/ However, if the HP 11896A runs actual time may be longer or
HP 11896A swept- slowly and the HP 71452B Opti- shorter. The SOP rarely moves
wavelength measurement cal Spectrum Analyzer sweeps exactly along a great circle.
of DWDM components quickly, then trace A of the OSA Rather it rather moves in a fig-
can capture the maxima and ure “8” as well as other figures
The HP 83438A #009/HP 71452B/ trace B the minima of each sweep. along the whole surface of the
HP 11896A setup as shown in The difference between trace A Poincaré sphere.
Figure 13 characterizes compo- and trace B shows the polariza-
nents for dense wavelength- HP recommends that Tc be about
tion dependent loss versus wave-
division multiplexing (DWDM) ten to forty times larger than
length. For low PDL values
applications fast and accurately the sweep time of the OSA. In
(<1 dB), the average of the two
versus wavelength. Under remote this case the state of polarization
traces is approximately the
control, it can measure insertion changes little during a single
insertion loss for unpolarized
loss, crosstalk and PDL auto- OSA sweep. The total measure-
light.
matically. Other characteristics ment time should exceed twenty
can be calculated too, such as to fifty times Tc, so that the
Measurement procedure HP 11896A exposes the DUT to
the polarization dependence of
the center wavelength or band- This procedure enables PDL all SOPs during hundreds of
width of a filter. measurements to be performed OSA sweeps.
over a specified wavelength span In the example below, the OSA
HP 83438A which is equal to or less than the sweeps in 110 ms. From Table 8
#009 Erbium width of the source used in the
ASE Source we selected the scan rate 3 (Tc =
HP 71452B test system. The best setting for 3 sec is greater than 20 * 110 ms)
Optical
Spectrum Analyzer the HP 11896A depends on the and let the measurement run
OSA. The following explains the for two minutes (40 * Tc).
procedure and the relationship
HP 11896A between the HP 11896A and the
Polarization Table 8. Typical
Controller OSA for the best measurement HP 11896A times to
statistically complete
results. one great circle on
Device the Poincaré sphere
Under Test HP 11896A settings:
Statistical Time
In this setup, the HP 11896A Displayed For a Great Circle
Figure 13. AUTO SCAN mode is used. How- Scan Rate Tc (seconds)
ever, what scan rate is the best?
The Erbium ASE Source 1 12
To minimize the overall mea- 2 6
HP 83438A with option 009 pro- surement time, the scan rate 3 3
vides polarized light in the range should not be too slow. To mea- 4 1.2
1525 to 1565 nm. Its polarization 5 0.6
sure the PDL at one state of 6 0.3
state is automatically random- polarization, that state should 7 0.25
ized using the HP 11896A stay long enough for the OSA to 8 0.2
Polarization Controller. If the complete a sweep; therefore the
HP 11896A runs fast and the OSA sweep speed should be HP 71452B settings:
HP 71452B Optical Spectrum appropriately fast. Set wavelength and amplitude
Analyzer averages 100 or more settings to match the device under
sweeps, then the resulting trace Table 8 lists the relationship test. In the example shown here,
represents the loss versus wave- between displayed scan rate and START is 1542 nm, STOP is
length of unpolarized light. the time Tc required for one great 1562 nm, and resolution band-
circle on the Poincaré sphere. width (RES BW) is 0.2 nm. Our
The HP 11896A randomizes the DUT has approximately 7 dB
SOP. Therefore, the time Tc is insertion loss.
only a statistical measure. The
13