The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> VCO start-up problem
https://designers-guide.org/forum/YaBB.pl?num=1190353499

Message started by mobil on Sep 20^th, 2007, 10:44pm

Title: VCO start-up problem
Post by mobil on Sep 20^th, 2007, 10:44pm

Hello all,

Just mentioned a old topic about how to trigger the VCO in simulation. I know it was discussed before in
http://www.designers-guide.org/Forum/YaBB.pl?num=1066188298/7
In above post, the suggestion is to add impulse (voltage or current) to 'kick' VCO or play around the initial conditions.

But, I remember there was another simple method (I can't find the post in the forum) mentioned that just setup the
'maxstep' in transient simulation (SpectreRF). e.g. if the frequency of vco is 5 GHz, let the maxstep = 5p. After doing so,
the 'impulse' component isn't necessary and the VCO will start by itself.

Unfortunately, I can't remind the detail of how to setup this 'maxstep' under the operating frequency. Does someone
know this or find the original post? Thank you.

Mobil

Title: Re: VCO start-up problem
Post by Ken Kundert on Sep 21^st, 2007, 8:50am

To start an oscillator you need to kick it somehow. You can use either a finite duration source or an initial condition. For ring oscillators, I prefer initial conditions. For resonant oscillators I prefer a damped sinusoid with a frequency that matches the oscillators. Set maxstep small to nurture the oscillation (oscillations when small can be quenched by large timesteps).

This is all discussed in some detail in my book (http://www.designers-guide.org/Books/dg-spice/index.html).

-Ken

Title: Re: VCO start-up problem
Post by mobil on Sep 21^st, 2007, 6:15pm

Thanks your reply, Ken.

I'll read that carefully.