AnsweredAssumed Answered

Averaged Measurement Starts Low in PLTS v5.5

Question asked by cram869 on Nov 15, 2010
Latest reply on Nov 15, 2010 by cram869
I apologize for the subject line in this post.  I am not sure how to label it otherwise.  This issue is probably another case of user error on my part, but it has been plaguing me for some time now.

I set up my measurement using the measurement wizard in PLTS.  I load a previous calibration. (I don't think I have the same issue if I perform the calibration and then immediately perform a measurement.)  Then, I collect the measurement.  I can visibly watch the return loss on my PNA N5242A drop from the previous level to a very low value.  As more sweeps are taken to collect the number of averages I requested  (usually 8 averages), the end result usually is close to but not quite at the level it should be.

I didn't notice what was happening at first until I started looking more closely at the insertion loss in my measurements.  I count on a certain amount of copper loss even at DC, but I was seeing perhaps a dB of loss at 10 MHz on a channel that I know wasn't that lossy.  Eventually, I realized that if I just click the Measure -> Start to get another measurement that I would get a more realistic value.

I assume it has something to do with resetting the averaging at the time PLTS takes the measurement.  It really looks like some old result is recalled at that point, and the first measurement I take is 1 unknown measurement + 7 DUT measurements averaged out. 

I've attached two measurements from the same DUT demonstrating the result.  The one ending in *2.dut is the first measurement, and the other is the subsequent measurement I get when I click Measure -> Start.