Designer’s Guide Community

Nick

Junior Member

Offline

Posts: 21

Starting up an oscillator in sim
Sep 27^th, 2006, 8:25am

Hi All,

I am designing an LC oscillator. To start it up, I ramp up the supplies to their final value and then apply a dumped isin current source on one of the tank nodes. I have noticed that below an amplitude value of the isin the oscillator does not start up, or starts up but fails to keep the oscillation. That value is around 10nA for this circuit. Can anyone advise me as to how low/high the isin source amplitude should be? What does it mean in terms of simulator that for some values of amplitude the oscillator does not start? I am using the gear2only method for the transient analysis.

Thank you,
Nick

Back to top

Nick

IP Logged

bernd Senior Member Offline Posts: 229 Munich/Germany	Re: Starting up an oscillator in sim Reply #1 - Sep 27^th, 2006, 8:30am Quote: I am using the gear2only method for the transient analysis. Try with traponly. Bernd
Back to top	Just another lonesome cad guy IP Logged

Nick Junior Member Offline Posts: 21	Re: Starting up an oscillator in sim Reply #2 - Sep 27^th, 2006, 8:37am Hi, thanks for the reply. Traponly and gear2only have the same results. When using Euler, I need even larger amplitudes to get start-up, but I think that has to do with the method itself. Any other ideas? Could it be circuit dependent? Thanks, Nick
Back to top	Nick IP Logged

bernd Senior Member Offline Posts: 229 Munich/Germany	Re: Starting up an oscillator in sim Reply #3 - Sep 27^th, 2006, 8:48am Try to set an 'Initial Condition" of the value of your supply voltage to the tank node rather then using an isin source. Bernd
Back to top	Just another lonesome cad guy IP Logged

Nick

Junior Member

Offline

Posts: 21

Re: Starting up an oscillator in sim
Reply #4 - Sep 27^th, 2006, 9:42am

That doesn't work.
I have tried playing with initial conditions and also with vsin sources on the supply and on the tank.
Only the isin source seams to work. I apply it with a delay, after ramping up the supplies of the circuit to their final value. I cannot understand the dependence on the amplitude of the current though.

Thanks,
Nick

Back to top

Nick

IP Logged

ACWWong

Community Fellow

Offline

Posts: 539
Oxford, UK

Re: Starting up an oscillator in sim
Reply #5 - Sep 27^th, 2006, 12:13pm

Nick wrote on Sep 27^th, 2006, 8:25am:

I have noticed that below an amplitude value of the isin the oscillator does not start up, or starts up but fails to keep the oscillation.

if you're using traponly and the LC oscillator starts but then the amplitude dies away, then i guess you don;t have enough negative resistance in your active core to overcome losses in your tank (ie you don't have a net negative resistance). you will find if you increase your negative resistance the oscillator will start easier in simulation (when given the current kick), as in real life. using gear or euler could kill the oscillation due to numerical damping.
as to the size of the kick, as oscillators start on noise, you could give it a kick in the order of the size of the noise you expect to give a better estimate of start-up time. although i normally give it just a big kick!
(aside: i think there is a fairly new transient noise function which could do this for you? i haven;t used it myself but might be worth looking into ?)

Back to top

IP Logged