Relaxation Oscillator Jitter Simulation in cadence - Eye diagram confusion...

[wandola]

Dear All,

I am designing a conventional relaxation oscillator. The moninal frequency is around 2.85 MHz.

When I try to run the transient noise simulation and use calculator to obtain the eye diagram, I got some weird diagram and i am very much confused.



Attached you can see some waveforms obtained using the function eye diagram. The setting is pretty simple. Just set the start and stop time. The period is the oscillator output frequency period. I just took an average value for this. The weird thing is that when I changed the period by a little bit, the plotted diagram is very different.

Can some expert help me with this? I don't know how i should proceed.

For the first graph, the period is 351.5 ns; for the second one, the period is 351.3 ns; the last one has period of 351.25 ns.

Please help me with this. thanks a lot guys.

 

[D.A.(Tony)Stewart]

What did you expect? a perfect sawtooth? I see a lot of instability modes with FM. I can guess but I"d rather see your design.

 

[crutschow]

For an eye diagram you must look at only one or two waveform periods, not many as your graphs show.

 

[D.A.(Tony)Stewart]

He does not have a stable pattern to even view an eye pattern with. It is over modulating or something. One can only guess.

 

[erikl]

... the last one has period of 351.25 ns.

I'd think isn't so bad at all - I just wouldn't call it an eye diagram.

In my eyes ;-) it depicts a time jitter of about 15ns, corresponding to a frequency jitter of about 60kHz at a mean frequency of 2.850MHz , or a phase jitter of about 15° .

But may be I'm totally wrong. Let's wait for the experts!