Transcript Optical Wave Guide Polarization Measurement System Students

Optical Polarization Dependent Loss Measurement System
Students: Jaspar Siu and Ronald B. Williams
Research Mentor: Dr. Christi K. Madsen
Department of Electrical Engineering, Texas A&M University
Research Goal: To create an automated polarization measurement system that characterizes the polarization dependent loss over the
wavelength range of 1520-1570 nm and finds the principal polarization states for different integrated optic waveguides and fiber devices.
Figure 5 –
Calculation Flow Chart
Introduction
Fiber optics is an important field in Electrical Engineering due to the many advantages it
has over traditional wires. Data is sent at the speed of light and fiber optics enables a
much bigger bandwidth. However, when we use light to transfer data, there are many
factors that affect the signal that is being transferred and polarization is one of the main
factors. Light is an electromagnetic wave. Polarization has to do with how the electric and
magnetic fields align themselves. There are 3 types of polarization for light. They are
linear, circular and elliptical polarization. If a device is polarization dependent and light is
not aligned to its most efficient polarization state, you may loose some of the data. The
difference between the best and worst case transmissions is called the polarization
dependent loss (PDL) and is typically expressed in dB.
For this research project, the objective is to develop a fully automated polarization
measurement system that measures the PDL for a range of wavelengths. The
measurement system calculates the first row of the Mueller matrix and the input Stokes
vectors. The Stokes vector describes the state of polarization of the light, for example, at
the input to the device. The Mueller matrix describes the action of a device on the state of
polarization. In this experiment we use the Mueller matrix to find the maximum and
minimum power transfer at each wavelength. We then calculate the most efficient
polarization state, ie. input Stokes vector, for different devices.
The measurement system makes use of 2 pieces of equipment, the Finisar and the
Polarization Controller to acquire the measurements we need.
The programming language we used is LabView because it can communicate with
different measuring devices through a General Purpose Interface Bus (GPIB). Therefore,
we can communicate with the devices directly through the computer without having to
manually change the parameters.
Methodology
Equipment Description
1.) Reference measurements are taken without the device under test. The reference
measurements are needed to account for variations in the input power. The user can
control the range of wavelengths. Once the program runs, the Finisar will output a laser
that is swept within the range of wavelengths. The laser will go through the polarization
controller and back to the detector on the Finisar. This process is repeated 4 times to go
through the 4 states (Vertical, Horizontal, 45° linear, and Circular) of polarization. This is
achieved by changing the angles of the quarter and half wave plates on the polarization
controller to various degrees. The whole process is repeated 4 times such that 4
measurements of the 4 states of polarization are taken.
Once all the measurements are made, averages are taken such that we get the average
measurements for the 4 states of polarization under the variable names Sv, Sh, S45, Sc.
The 4 sets of measurements are then compared with each other such that an array of noise
can be calculated and averaged. Standard deviation of the noise is also calculated to see
how consistent the measurements are. The 4 variables are saved in a binary format, such
that memory space can be saved. The data can be retrieved from an add-on program to
LabView called DIAdem. Graphs, reports and data manipulations can be easily made in
DIAdem.
Results
The following results are for a specific DUT device called Single Ring Micro Resonator:
Figure 6 – PDL against Wavelength Plot
At certain wavelengths PDL is very high
This device is polarization dependent
Figure 7 – Tmax and Tmin against Wave Length Plot
When you zoom into the data, T min is slightly delayed
from Tmax. Tmin and Tmax follow the same path, but Tmax
yield higher power
Figure 8 – S’s against Wavelength Plot (Reference Measurements)
The plots are almost identical, so there isn’t much noise
in the equipment.
Figure 9 – T’s against Wavelength Plot (Actual Measurements)
Different states of polarization produce different power
outputs.
2.) Actual measurements are taken with the device under test. The same process is used as
the reference measurements. The average actual measurements for the 4 states of
polarization are saved under the variable names Tv, Th, T45, Tc.
3.) Calculations are made immediately after the actual measurements. The variables Sv, Sh,
S45, Sc, Tv, Th, T45 and Tc are all arrays of data in the form of decibels (dB). They are all
converted into power units (mW) before they can be manipulated. The following
equations are then used to get the first row of Mueller Matrix:
M00 = (1/2) * ( (Th/Sh) + (Tv/Sv) )
M02 = (T45/S45) – M00
M01 = (1/2) * ( (Th/Sh) - (Tv/Sv) )
M03 = (Tc/Sc) – M00
The first row of Mueller Matrix determines the insertion loss for the DUT.
Figure 1 - Finisar CT-RTS
The Finisar CT-RTS is a real time continuous device that outputs a multimode sweeping
laser that sweeps through a limited range of wavelengths. The maximum range of
wavelength it can output is 1520nm to 1570nm. The Finisar has 2 detectors and it can
display the incoming signals on its LCD screen.
Figure 2 – HP 8169A Polarization Controller
The following equation is used to determine the intermediate variable A:
A is a variable that is needed to make further calculations.
The following equations are used to determine Tmax and Tmin:
Tmax = M00 + a
Tmin = M00 – a
Tmax and Tmin represent the maximum and minimum power transfer at a specific
wavelength.
Figure 10 – Smax Matrixes against Wavelength Plot
Figure 11 – Smin matrixes against Wavelength Plot
At each wavelength there are a set of 4 Smax elements.
At each wavelength there are a set of 4 Smin elements.
These can be used to calculate the angles for the quarter and half These can be used to calculate the angles for the quarter and
wave plates to get the most efficient polarization state for
half wave plates to get the least efficient polarization state for
the device.
the device.
Tmax and Tmin can be used to calculate the polarization dependent loss (PDL) of the device.
The HP 8169A Polarization Controller is a device that changes the polarization of an input PDL = 10*Log (Tmax/Tmin)
signal, and outputs a new polarized signal. The 3 parameters you can change are the
PDL shows the loss that is due to polarization. If a waveguide is polarization independent • The measurement system is fully automated, which yields to the least human
polarization reference, the quarter wave plate and the half wave plate.
then PDL will be low throughout the range of wavelength.
intervention and less error.
• The measurement system is fast paced since measurements and calculations are done
Program Flow Chart
Finally, Smin and Smax can be calculated using the following equations:
in less than a minute.
• The measurement system calculates PDL, Tmax, Tmin, Smax and Smin for the whole range
Figure 3 – Reference
of wavelengths, which allows the user to determine which wavelength yields the least
Measurement Flow Chart
PDL, and the most efficient state of polarization for the waveguide.
Smax and Smin are vectors that represent the maximum and minimum Stokes vector. From • The measurement system saves the data in a binary format which saves memory
Figure 4 – Actual
space, and it allows for easy retrieval of data for graphing, and making reports.
the maximum and minimum stokes vector you can calculate the angles for the
Measurement Flow Chart
•The measurement system determines the noise in the equipment, and it takes
quarter and half wave plates such that you get the most efficient state of polarization.
averages to minimize the amount of error due to noise.
Conclusion
This project is funded by National Science Foundation (NSF) and Department of Defense (DOD). Special thanks for the valuable guidance of Dr. Christi K. Madsen and the generous help of Mehmet Solmaz, Mike Thompson, and Donald Benjamin Adams.