Designer’s Guide Community

vakilia

Junior Member

Offline

Posts: 15
Atlanta, GA

Frequency versus control voltage in VCO
May 14^th, 2003, 7:55am

I have designed this ring VCO and the time domain analysis shows me how the frequency (Frq) changes with the control voltage (Vctrl), which is not a linear behavior actually. To find the VCO gain I would like to be able to find the transfer function (Frq/Vctrl) in frequency domain but I am not sure if Spectre will take care of the non-linear behavior of the VCO and even if it is possible to do so?!

Thanks

Babak

Back to top

IP Logged

Eugene

Senior Member

Offline

Posts: 262

Re: Frequency versus control voltage in VCO
Reply #1 - May 14^th, 2003, 12:12pm

You mention a frequency domain transfer function and nonlinearity in the same inquiry. Exactly what do you mean by frequency domain transfer function? Do you mean the static relationship between VCO tuning voltage and the VCO's steady state operating frequency? Or do you mean the small signal transfer function about an operating point (which would be proportional to 1/s)? Or do you mean something like a describing function, which is essentially the linear transfer function that best approximates the large signal response?

Back to top

IP Logged

vakilia Junior Member Offline Posts: 15 Atlanta, GA	Re: Frequency versus control voltage in VCO Reply #2 - May 14^th, 2003, 3:51pm I actully mean the static relationship between VCO tuning voltage and the VCO's steady state operating frequency.
Back to top	IP Logged

Eugene

Senior Member

Offline

Posts: 262

Re: Frequency versus control voltage in VCO
Reply #3 - May 14^th, 2003, 8:32pm

I don't believe you can easily generate this curve using the basic Spectre. However, I have generated such curves with relative ease using SpectreRF. If memory serves, I ran a swept autonomous PSS analysis and plotted "beat frequency" against the sweep variable. The sweep variable was of course the VCO tuning voltage.

Back to top

IP Logged

Andrew Beckett

Senior Fellow

Offline

Life, don't talk to
me about Life...

Posts: 1742
Bracknell, UK

Re: Frequency versus control voltage in VCO
Reply #4 - May 15^th, 2003, 7:00am

Another way you can do this is to do a transient and then
sweep the control voltage. At the output of the VCO, put a
freq_meter component (in ahdlLib under <instdir>/tools/dfII/samples/local/artist/ahdlLib at the output.
This is a verilog-a model which captures the instantaneous frequency.

However, in order to allow it to settle you probably need to sweep the
control voltage slowly, or perhaps use a pwl source to step the
control voltage up so you can leave it steady for a few cycles.

Best is to use the autonomous PSS method described previously in this
thread, though.

Regards,

Andrew.

Back to top

IP Logged

vakilia

Junior Member

Offline

Posts: 15
Atlanta, GA

Re: Frequency versus control voltage in VCO
Reply #5 - May 16^th, 2003, 4:25pm

Eugene and Andrew;
I actually used the freq_meter component in ahdlLib and changed the Vctl to pwl with approperiate slope and ran the transient simulation. In the results window I plotted output frequency versus control voltage by changing the x-axis to Vctr net name and it showed the static frequency responce perfectly.
I will try the other way as well.

Thanks for your helpful comments,
Babak

Back to top

IP Logged