Hi, I'm trying to measure an amplifier's stability. I'm using the network analyzer's equation for calculating and showing mu/mu prime. I'm also saving the s2p file and plotting stability in ADS.

I'm seeing that the stability differs between the two methods although both should be based on the same s parameter data. The formula for stability is pretty straightforward so I'm wondering what is different in the network analyzer?

By the way, the analyzer is saying my amplifier is stable while ADS shows potential instability so I do have a need to clarify what is correct.

I'm seeing that the stability differs between the two methods although both should be based on the same s parameter data. The formula for stability is pretty straightforward so I'm wondering what is different in the network analyzer?

By the way, the analyzer is saying my amplifier is stable while ADS shows potential instability so I do have a need to clarify what is correct.

If the S11 and S22 are close to zero, then small errors (even rounding errors) can cause trouble. If you are re-simulating in ADS and don't use exactly the same frequencies, then interpolation can cause trouble.

I did notice that my cal only consisted of 201 points over a 10GHz span. I'll try increasing that and seeing how the data will differ. It's so close that cal/port extension may very well be affecting it (my marginal results that is). I tried simulating with the same span and number of points but still see the same result. Unfortunately I did not save the stability measurement on the analyzer and can't show you, I'll be repeating it.

Update: As Dr_joel pointed out, I believe this is due to rounding when saving the s2p file. The difference between ads and the network analyzer isn't that great but since I'm so close to gamma = 1 it was enough to warrant potential instability.

Edited by: ervman1 on Apr 2, 2014 10:36 AM

## Attachments