The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> LC-VCO varactor
https://designers-guide.org/forum/YaBB.pl?num=1267075294

Message started by unevb on Feb 24^th, 2010, 9:21pm

Title: LC-VCO varactor
Post by unevb on Feb 24^th, 2010, 9:21pm

I have a question on how to use a varactor in the tank circuit for a LC-VCO

Structure of the VCO : It has an NMOS and PMOS cross-coupled negative gm-stage. A L-C tank connected across the differential arms of the VCO. The control line of the VCO is driven by the output of the charge pump and loop filter. The control line can vary from GND to SUPPLY.

My questions are

i) If we use a NMOS varactor with the gates connected to the VCO output and the control line connected to Source,Drain,Body (shorted together) , then the Vgs of the NMOS is continuosly varying as the VCO oscillates. The capacitance will keep changing in this case. Cap change means the resonant frequency is continuously changing. How does it work then?

ii) There will be some combinations of Control Voltage and VCO output for which the Vgs will not be sufficient to create the cap!

iii) With the above mentioned circuit configuration. When I change the control line from GND to SUPPLY, then there is a change in the amplitude of the oscillations at the output. Is this common? Only possible reason I can think of is the parasitic resistance being a function of the control voltage. and increases in one case. Any other possible reasons?

Thanks

Title: Re: LC-VCO varactor
Post by rfcooltools.com on Feb 26^th, 2010, 6:52pm

unevb,
i)The Capacitance will keep changing, but the frequency of oscillation will be close to the average capacitance presented through out a full period. Signal varying capacitance will be a form of distortion which may make the phase noise greater than optimal.
ii) basically the KVco will radically change effecting the loop stability and loop gain. It will have an effect on the loop bandwidth and gain dropping due to KV will probably make it more stable so stability may not be a concern. with that said my guess is that phase noise may suffer. Try to keep the VCO out of this region for the most part unless the system is reciprocal mixing limited and not integrated phase error limited as in a jammerless data system.
iii) This makes sense since the varactor Q is probably quite a bit lower than the other capacitors in the circuit. When this lower Q varactor's weighted Q becomes more dominant then the total tank Q decreases thus the amplitude will be lower.

Also worth noting is that most mos devices are not modeled correctly to be suitable as varactors for low noise operation primarily due to gate resistance not being modeled correctly or at all in many cases. In other words the results will be better in simulation then measured. Consider trying to unload the burden of tuning a wide frequency range by the varactor only instead choose a method of switching in capacitance when varactor voltage exceeds a range. VCO design is a lesser of all evils multi variable optimization. Where simulation is hard to get accurate and modeling is everything.

J

Title: Re: LC-VCO varactor
Post by tkhan on Feb 27^th, 2010, 7:50am

unevb wrote on Feb 24^th, 2010, 9:21pm:

If we use a NMOS varactor with the gates connected to the VCO output and the control line connected to Source,Drain,Body (shorted together) , then the Vgs of the NMOS is continuosly varying as the VCO oscillates.

Have you tried something like this:

http://www.experts.com/Images/articles/Gutierrez1/vco.gif