Overview
When testing optical components, understanding polarization alignment, potential crosstalk, and fiber drift between the Optical Vector Analyzer (OVA) and the Device Under Test (DUT) is a common requirement. This article outlines how the LUNA OVA manages internal polarization states, how to leverage its software for device analysis, and how to properly calibrate the system to minimize fiber-induced polarization errors.
LUNA OVA Polarization Architecture
Dual Orthogonal Launch
Polarization management within the OVA relies on an automated matrix architecture rather than manual physical alignment. To measure the complete Jones Matrix transfer function, the OVA automatically transmits two copies of light that are polarized perfectly orthogonal to one another. The initial physical launch angle of this light is arbitrary and is not referenced to the connector keyway. Despite this arbitrary launch angle, the two distinct states of light are guaranteed to remain strictly orthogonal to each other.
Jones Matrix Calculations
The OVA receives the transmitted light from the device using polarization-diverse detectors. By processing the two orthogonal inputs across these diverse detectors, the system mathematically calculates the full Jones Matrix of the DUT. This complete matrix allows users to accurately study and isolate the polarization states inside the device, even if the physical light entering from the OVA is not aligned with the internal TE and TM axes of the waveguide.
Simulating Polarization States via Software
The optional LUNA Polarization Analysis Software provides tools to characterize how a device responds to specific polarization orientations without needing physical adjustments. Users can simulate the expected optical response for any user-specified input vector by varying the Phi and Theta angle values within the software. This simulation feature allows operators to identify the precise angles that correspond to the minimum or maximum transmission states of a device.
More information about Luna's Polarization Analysis Software can be found in chapter 6 of the OVA User's Manual.
Best Practices: Mitigating Fiber Drift and Crosstalk
Because the input of many devices is coupled via Single-Mode fiber, polarization states will inevitably wander or rotate within the lead fibers connecting the OVA to the DUT. To minimize these effects and ensure the OVA measurement coordinate system aligns with the DUT, follow this recommended calibration workflow:
Secure the Setup: Tape down all coupling and lead fibers firmly to the work surface so they cannot physically move during testing.
Connect the Leads: Connect the lead fibers directly to each other using an optical coupler, omitting the DUT for the baseline scan.
Perform Calibration: Complete a system calibration in this connected state. This process effectively "zeros out" the polarization rotation and phase effects introduced by the lead fibers themselves.
Insert the DUT: Carefully remove the optical coupler, insert the DUT into the optical path, and proceed with taking your measurements.
Note: While disconnecting the coupler and inserting the DUT requires minor physical handling, this method remains the best available practice to minimize path alterations and maintain measurement integrity.
Limitations on Pure TE/TM Mode Isolation & Birefringence
The Challenge of Pure Mode Excitation
A common testing goal is to isolate a specific polarization (such as pure TE or pure TM) to measure localized insertion loss to deduce birefringence from end-facet reflection delays. Users often worry that a misalignment between the LUNA source and the waveguide axes will excite an unwanted mixture or superposition of TE and TM modes. Because the OVA inherently relies on a dual-polarization orthogonal structure to calculate its datasets, it is not possible to physically illuminate a DUT with only a single, isolated mode (such as pure TE or pure TM) to back out birefringence.
For specialized applications requiring isolated, single-mode excitation, an Optical Backscatter Reflectometer (OBR) is recommended. The OBR inherently outputs only a single polarization state at a time, though like the OVA, its physical launch angle is not locked to the fiber keyway. By using an external polarization controller, the output of the OBR can be manually polarization-aligned with the TE or TM modes of the DUT.