Hello board, I am actually working with the optical component analyzer N7788B and need some help for the measuring method. I mean the Application "PMD/PDL/Loss" from the Polarisation Navigator Software. What I want to do: We have a PMD emulator from fiberpro (PE4200) in our lab and we are not shure how exact/valid are the emulated values (DGD and SoPMD). Therefore I want to meassure that device with the N7788B and a tunable Laser Source (Agilent 81600B). However, the setup is well described in the manual (N7788B) and my measurings are working verry well. BUT...
My Problems: I am able to set specific values on my DUT (PE4200) such as DGD:100ps / SoPMD:1000s^2.To measure these values I'm using the Application "PMD/PDL/Loss" from the Polarisation Navigator Software. The measuremt results are extremely depended from the Sweep Rate and that does not mean that the highest/lowest rate produce the best result. I did measurings for more sweep rates and somehow the "best" (nearest to the set value) results at a sweep rate between min and max rate. This is also different for varied DUT set values. In the beginning I did all the measuring with the highest post processing resolution. Unfortunately I saw that the post processing resolution has also a big influence to the measured value. Here also the highest resolution does not end in the best result value. Resolution values between highest and lower are more effective.
NEED HELP!!! Does someone here know the reason of these impacts? Does someone know what the best practice is? Does someone here know a better description as this one from the Agilent N7788B Manual? Does someone here know some white papers for this kind of DGD, SecondOrderPMD measuring?
I’m sincerely thankful for every help on this side.
We'll probably need to see in more detail what kind of differences are produced. You could post attach a couple files (the best would be the .pbin measurement files). If you prefer to me with the PM mail system of this forum or let me know if you want to e-mail them directly.
The PMD measurements are made with a single-sweep implementation of the Jones Matrix Eigenanalysis method. With this method, the PMD is determined from the rate of change of the Jones Matrix vs. wavelength. This does mean that higher resolution measurements become noisier, essentially because the result is formed with a divisor that is approaching zero. Maybe that is part of what you are seeing.
The wavelength resolution of the N7788B PMD measurement scales directly with the sweep speed, and is only 12 pm at 40nm/s. Since many DGD/PMD measurements, like on fiber, do not require this resolution, the result can be smoothed with the resolution setting that you refer to. This is essentially a post-processing.
Anyway when the device you are measuring is set for a higher value of 2nd-order PMD, the PMD vector and DGD are varying with wavelength. Is the period of this variation close to the measurement resolution in your case? Or also to the smoothing resolution? That might be a cause of speed dependence. Do you have less dependence when the SOPMD is set lower?
Hi Mike, Thank you for your answer, I thought I will be informed by a mail at new forum replies… however. Yes I will send you some example measurement. Ok, seems that .pbin files are not allowed to be uploaded. Maybe the best way is to send them directly to you. Or I try it via PM mail. I understand your description, and yes… that is also my actual view… But how I should now what is the best sweep rate and what can be the best post processing resolution, when I have a black-box measuring and don’t know what DGD, SoPMD my DUT has. Thank you. Regards, Dietmar