The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Measurements >> Phase Noise and Jitter Measurements >> different jitter results from transient to pss+pnoise analysis
https://designers-guide.org/forum/YaBB.pl?num=1217583830

Message started by trond on Aug 1^st, 2008, 2:43am

Title: different jitter results from transient to pss+pnoise analysis
Post by trond on Aug 1^st, 2008, 2:43am

I want to verify obtained jitter results from a PSS+pnoise analysis with a transient simulation to be sure the obtained values are correct.

1) I have a SCL divide-by-2 circuit driven with a 4GHz signal (o/p is at 2GHz). I do a PSS and pnoise analysis with frequency range from 1k to f_out with a beat frequency of f_out. I use the pnoise jitter plot form to calculate the jitter Jee over a range from 1k to f_out/2. The obtained Jee is approx. 350fs.
(3 harmonics, 30 sidebands, tstab 20ns)

2) I run a transient simulation and write all periods of the output clock to a file and import it to matlab. I use the matlab function J=std(periods); to obtain the absolute jitter. J=0.13fs.

Q1) Why are the values so different? I was expecting similar results?

Thanks

Title: Re: different jitter results from transient to pss+pnoise analysis
Post by Stefan on Aug 1^st, 2008, 3:24am

How many periods did you write to your output file?
Are you sure that your simulator correctly implements the noise?
When using spectre you can verify this by setting transient noise to another multiplicator or simply by changing the simulation temperature. Anyway, the difference (if calculated correctly) is too much ...

Title: Re: different jitter results from transient to pss+pnoise analysis
Post by trond on Aug 1^st, 2008, 6:03am

Thanks for the prompt reply Stefan.

I simulated over 5000 periods. I am running a longer simulation just now to see if it makes a difference. In the transient simulation I did not have the transient noise option enabled.

1) Should I do so? If so, the pss+pnoise analysis does not introduce random noise, does it?

At the moment I only have an RCL network on the power supply and all resistors have noise as well.

2) In theory the results should be pretty much identical, should they not?

Thanks,

Title: Re: different jitter results from transient to pss+pnoise analysis
Post by trond on Aug 2^nd, 2008, 11:15am

Using the noise option in the transient simulation does increase the jitter. For example. the jitter calculated in Matlab from the obtained periods is approx. 230fs.
As mentioned earlier the pss+pnoise gave 350fs as Jee [1k..f_out/2].

1) But I am still not certain why the two values are different, and more importantly, which value is the correct one.

Thanks,
Trond