AnsweredAssumed Answered

PER "jumps" when using N7788B with N7783B thermal cycling unit

Question asked by maryrae on Jan 16, 2013
How can I fix my PER measurement?

Problems using PER (Heater) button on Polarization Navigator software:

I'm characterizing the PER of a PMF port of a 3-port DUT across the C-band.  Port PER is typically around 45dB, but I'm seeing occasional sudden drops to 30dB or even 10~20dB.  When I go back and re-measure, PER is around 45dB.

There's no physical reason the DUT should behave this way.  It should be roughly 45dB across the C-band.  

One concern is that the yellow arc on the PER tab often only traces out about 200 degrees rather than a complete circle.

Also, our "good" PER values (that is, not counting the jumps) are a few dB better than those of our DUT vendor.  Vendor uses same setup--N7788B + N7783B+ pol controller + low-SSE TLS.

I've tried both a low-SSE laser and a higher power (not-low-SSE) Agilent 81682A.  Either way, DoP is >98%.
I'm using the Mueller Matrix method and a polarization analyzer/controller to maximize the light into the N7788B. Depending on the TLS, input to N7788B is -5 to -11dBm during PER test.
PM splice is between heater and N7788B (in other words, heater is between DUT and splice).