The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> LC VCO topology
https://designers-guide.org/forum/YaBB.pl?num=1376932265

Message started by JV on Aug 19^th, 2013, 10:11am

Title: LC VCO topology
Post by JV on Aug 19^th, 2013, 10:11am

Hi guys,

I see that a widely used topology for this kind of oscillator is the LC tank resonator with cross-coupled pair and current source, as in attachment.

I was wondering why most people use the nMOS cross-coupled pair with nMOS bias transistor? What's wrong with the nMOS cross-coupled pair and pMOS bias transistor for example? The pMOS transistor has much less flicker noise, so it will improve the phase noise performance of the VCO. Is there any disadvantage to that topology (except area)?

Is it important for the cross-coupled pair and the bias transistor to be of the same type?

Thanks a lot!
JV

Title: Re: LC VCO topology
Post by raja.cedt on Aug 19^th, 2013, 11:06am

Hello,
It depends many factors, however you could freely chose all 4 variable combination(n or p cross coupled, n or p current source). The main advantage of current architecture are

1. There is no DC voltage change change across Varactor as inductor has 0 dc potential, hence less 1/f bias noise up conversion (common mode to frequency or AM to FM conversion). You can refer The impact of device type and sizing on phase noise mechanisms for further understanding.
2. Supply noise spurs also will be very less, whenever there is noise on supply same noise noise will be appeared at node vx and vy and if loop filter also refereed to same supply then Vtune also will also be impacted by same noise, hence no noise across varactor and no spurs.

Thanks,
Raj.

Title: Re: LC VCO topology
Post by JV on Aug 19^th, 2013, 3:58pm

Thanks for the quick and clear reply, Raj, I appreciate that!

JV

Title: Re: LC VCO topology
Post by aaron_do on Aug 19^th, 2013, 5:51pm

Hi Raj.,

Quote:

and if loop filter also refereed to same supply then Vtune also will also be impacted by same noise

in this case, is it better to tie the loop filter capacitors from the supply to the charge-pump output instead of from charge-pump output to ground? My thinking is the supply may not be ideally referenced to ground. If you don't do that, I guess your supply cap should be significantly larger than your loop-filter cap right?

thanks,
Aaron

Title: Re: LC VCO topology
Post by raja.cedt on Aug 20^th, 2013, 12:48am

Dear Aaron,
unfortunately i didn't get what you are saying. What do you mean by supply may not be ideally referenced to ground. If you don't do that, I guess your supply cap should be significantly larger than your loop-filter cap right.

Thanks,
Raj.

Title: Re: LC VCO topology
Post by aaron_do on Aug 20^th, 2013, 1:10am

Hi Raj.,

you mentioned that the loop filter it referenced to the same supply as the VCO. So what you're saying is that if there is any noise on the supply then it will appear on both terminals of the varactors, so the voltage across the varactors will be zero.

I was just saying that your loop filter has some large capacitors connected between the CP output and ground. Ideally ground and supply have the same AC potential, but in practice, its not really true, and there will be some power supply noise wrt ground. So is it better to connect the loop filter capacitors between the CP output and the supply instead. That way, the CP output will track variations in the power supply better.

regards,
Aaron

Title: Re: LC VCO topology
Post by raja.cedt on Aug 20^th, 2013, 2:00am

Dear Aaron,
I am also saying the same(Please refer my first post), loop filter should be connected between VDD and control node. If you are talking about parasitic bottom plate cap or some transistor diffusion cap between control voltage node and gnd then noise appeared on node X will be slightly lesser than at Vtune.

Thanks,
Raj.

Title: Re: LC VCO topology
Post by aaron_do on Aug 20^th, 2013, 2:38am

Hi Raj.

yes that's what I meant.

thanks,
Aaron