How to simulate pss in spectre for a periodically turned on/off VCO

[DoctorWho]

Dear guys,

I'm now designing an VCO. It's a bit different in that it's periodically turned on/off.

Thus, I have a voltage pulse at a frequency around 1MHz to implement the periodical on/off. Meanwhile, the VCO is oscillating at several GHz. I want to use pss and pnoise to see spectral purity of the VCO when it's on but cannot setup pss simulation correctly.

The simulator seems to consider the frequency of pulsed on/off signal as a fundamental and says that the fundamental tone should be integer multiples of the 'Beat frequency', which I usually set as VCO frequency.

Anyone can help me out?

Thanks in advance.

 

[dick_freebird]

You might want to play with a two step solution - a transient
long timescale run to get the thing to turned-on-steady-state,
save the final node voltages at a timepoint that is a clean
place for activity (far from phase compare edges and so on)
and then jam that initial condition onto a start-from-0 single
reference clock period that has all of the stimuli aligned the
same, relative to ref clock and VCO phase, as that .IC final
cycle.

Then the on/off periodicity will be out of play because it's
static within that second run.

I think your reference clock is the "fundamental" of interest.

 

[DoctorWho]

You might want to play with a two step solution - a transient
long timescale run to get the thing to turned-on-steady-state,
save the final node voltages at a timepoint that is a clean
place for activity (far from phase compare edges and so on)
and then jam that initial condition onto a start-from-0 single
reference clock period that has all of the stimuli aligned the
same, relative to ref clock and VCO phase, as that .IC final
cycle.

Then the on/off periodicity will be out of play because it's
static within that second run.

I think your reference clock is the "fundamental" of interest.

No, I'm not interested in spectral purity of reference clock.(If I understand it right, this is the clock that controls on/off of the VCO)

I'm interested in spectral purity of the VCO signal running at several GHz. Actually, I'm more focused on the effect of settling after switching on spectral purity. I need to see whether the settling effect has negligible effect on spectral purity.

I can see roughly what you're suggesting. But I'm not sure: Are you still suggesting assessing spectral purity with pss, followed by pnoise, analysis?

 

[dick_freebird]

Reference clock is the PLL comparison frequency. The on/off
enable, I'd think would be slower.

If you want to observe the spectral purity in relation to a
settling phenomenon, you are a priori violating one of the
precepts of pss/pnoise - it can't be steady state if it hasn't
settled.

In that case I think you probably want something more
like a windowed FFT or even a simplistic period-by-period
frequency number (I have a veriloga widget that does
that, buried in some design project somewhere) that you
can see at least fundamental drift on. Maybe a series of
windowed FFTs spaced in time?

 

[DoctorWho]

I have had a little progress. I removed the 'period' parameter in vpulse. Instead, I connect 4 vpulses in series. The first one has an offset of 0ns, pulse width 50ns. Second one: offset 100ns, width 50ns ... Fourth one: offset 300ns, pulse width not specified(Infinite, I guess). With this, I managed to make pss converge.

Indeed. You're right. pss analysis seems automatically detecting the steady state. The last switching activity in the circuit occurs at 300ns and 'pss' simulation widows shows several time windows: time = (0s-> 301.2ns), time = (301.2ns-> 301.4ns), time = (301.2ns-> 301.401ns), time = (301.2ns-> 301.401ns)... The pss seems to find the circuit after 301.2ns to be stable and converges.

I think you're right on windowed FFT. I think what I'm looking for is to do FFT from 300ns. I have also been suggested something like time-domain noise simulator from Berkeley. Anyway, I think this is the correct track to go.

Thanks for your help!