Designer’s Guide Community

sandman

Community Member

Offline

Posts: 73

VCO simulation
Oct 25^th, 2007, 3:41am

Hi,

I'm trying to simulate a current-controlled-oscillator (defined using verilog-A) and view it's oscillations from a transient response in Cadence. I sweep the current using the 'parametric simulations' in Spectre and obtain the PSD using the dft (of the transient response) from the Calculator tool.

Since the dft of the transient signal generates many peaks (including harmonics) for each of the current values in the sweep, I was wondering if it was possible to determine which of the peaks is the fundamental and also if there's any way in which I can automatically calculate and store the peak (value from the dft plot) of this fundamental using some sort of a script of any other method to a file for further processing, say, to generate some sort of a statistic (current versus peak frequency generated).

A very long and confusing question, but any help would be greatly appreciated !

Cheers!
Sandman.

Back to top

IP Logged

ACWWong

Community Fellow

Offline

Posts: 539
Oxford, UK

Re: VCO simulation
Reply #1 - Oct 25^th, 2007, 4:07am

The issue with your approach is that the period over which you do your dft will determine the frequency resolution. As such finding the frequency at which the dft is maximised will probably carry a fair error (which reduces with longer dft period/more simulation time). Also you need to ensure your dft expression is setup correctly to ensure you pickup your fundamental frequency correctly without aliasing. Anyway simply using the calculator xmax(mag(your dft expression)) will give you the frequency of the maximum amplitude dft tone.
If you just want to know the frequency you can simply use the clip and frequency calculator functions directly on the transient waveforms. Using this method you could just do a single swept transient where your current is ramped versus time and you plot using the frequency calculator function on the transient waveform.

A better approach would be to do parameteric pss analyses which will yield the fundamental oscillation frequency accurately in each parametric simulation... assuming your verilogA model is suitable for pss analyses.

Hope this helps

aw

Back to top

IP Logged

sandman Community Member Offline Posts: 73	Re: VCO simulation Reply #2 - Oct 27^th, 2007, 4:15am I haven't been able to try it, but, many thanks for the help, aw.
Back to top	IP Logged