Hi, we use keysight N7700A IL/PDL software to measure the WDM IL and PLD spectrum.
The software will output 2 spectrum curve : TE and TM after sweep.
Is there any naming rule for TE/TM curve?
When we measure repeatedly, the same curve sometimes named different name.
for example,
it outputs two curves, A named TE, B named TM
but sometimes A named TM, B named TE
this makes us a little confuse.
Hi, that's a good point to clarify. The two curves produced by the TE/TM analysis correspond to the spectra for polarization aligned with the two axes of the device under test. But the software analysis alone cannot determine the assignment of the axes to TE and TM, or even if those are useful labels for the axes.
The algorithm chooses a wavelength where the measurement results give a good distinction for determining the two axes and the curve with higher transmission (lower IL) at this wavelength is shown with the solid line in the graph. I think the exported data for Excel puts this curve under the label TE.
Anyway, in some measurements, small changes due to noise or temperature change of the device can change the wavelength used or the relative ordering of the curves at that wavelength. That is apparently what happens in your measurements. In that case, you should not use the "TE" and "TM" labels provided by the software but may need to use other knowledge of the DUT to assign the axes based on the expected behavior. Your expectations should be trusted more than the somewhat arbitrary assignment of the TE/TM analysis algorithm.
In some cases you may also find the new function of this software useful for setting the polarization controller to one of the axes based on either the minimum or maximum signal at a self-chosen wavelength or based on the TE/TM analysis.
That is described in this application note.
Solving Polarization Alignment with the IL/PDL Test System
Best regards,
Mike