Quantcast
Channel: Cadence Custom IC Design Forum
Viewing all articles
Browse latest Browse all 4954

Transient assisted PSS convergence problem

$
0
0

Hi there,

I have experienced a PSS convergence issue and needed some advice.

My circuit is a PLL that requires a certain time to settle, and I wanted to simulate the close-loop performance of the PLL.

I do PSS simulation, with "Run Transient" set to "Yes". Then I specify the proper time for the PLL to settle and I also saved the transient results.

The problem comes with the stop time setting. I know that the circuit will settle within 3us, verified by transient simulation, and I specify the stop time to be longer than that, for instance 4us.

What I noticed is that as I change the stop time, althoughalways longer than setting time, sometimes PSS converges and sometimes it does not.

For instance, I set 4us stop time, simulator converges and simulation finishes with expected results. Now I increase the stop time a bit longer to 4.1 us, then pss fails to converge. If I increase the stop time to 4.3us, it converges, and with 4.5us, it fails again.

The circuit is already settled at 4us, pss converges, and the results are as expected. What I don't understand is why the simulator can't converge at 4.1us.

What I noticed is that when it converges, it converges pretty fast, with just a few tries. when it fails, it also fails very fast, within a few iterations, "Conv norm" goes to very large number.

I wonder what could I do to improve this situation. I used default PSS convergence and accuracy settings, and I am using ic213_isr22 

Thanks!


Viewing all articles
Browse latest Browse all 4954

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>