Designer’s Guide Community

mixed_signal

Senior Member

Offline

Posts: 183

varactor in lc vco
Dec 26^th, 2012, 5:30pm

Hi,
a) I am designing a LC vco and using the varactor shown in fig. 1 where C is mim cap and C2 is diode varactor in IBM130nm
Resistor sets the node V1 near 0V.
b) I want tuning range from 786Mhzz to 800Mhz which requires C to change from 1394f to 1342f i.e. delta_C=52f

My doubts:
1. How C1 and C2 and Resistor should be chosen/ratio? I know C1 should be very large.

2. I want to check the standalone equivalent capacitance through simulation using AC 1V, freq=1/2pi,sweeping DC volt across varactor (Method by Ken Kundert) .The varactor alone works fine.

3. But the cap of combination is not C1 series C2 . Instead in Fig.2 it is C1 and in Fig. 3 it is C2. The node V1 seems to be an AC ground thereby preventing series combination of C1 and C2.
My question is how this circuit works? Am I doing mistakr by simulating this way?

Back to top

varactor.jpg

IP Logged

raja.cedt

Senior Fellow

Offline

Posts: 1516
Germany

Re: varactor in lc vco
Reply #1 - Dec 26^th, 2012, 9:52pm

Hello,
Very interesting Question.
This ckt contains 4 ports, V+,V-,Vctl, biasing terminal (gnd terminal of R), you have to find cap between two ports instead of that you have shorted 2 ports always (in fig2 V+ and bias,fig3 Vctl and bias) and calculated cap between other terminal and shorted terminals.
You have to find capacitance between V+ and V-, so please add a 1V ac source between those and find cap.
or V+ and control node also fine.

Thanks,
Raj.

Thanks,
Raj.

Back to top

IP Logged

rfidea

Senior Member

Offline

Posts: 159
Europe

Re: varactor in lc vco
Reply #2 - Dec 27^th, 2012, 7:11am

Strange. I agree with raja that you should measure the differential capacitance. But you should get ok values even in the test you have done. You state that you are using "freq=1/2pi". What are you meaning by that? freq=0.159?

For "normal" design the resistor value should be higher than your equivalent parallel resistor from your tank. Otherwise if will degrade your Q. Usually the C1 is chosen much higher than C2. But if you have a very small tuning range you can relax this and then not be able to tune so much.

Back to top

IP Logged

raja.cedt Senior Fellow Offline Posts: 1516 Germany	Re: varactor in lc vco Reply #3 - Dec 27^th, 2012, 10:00am Hello Mixed_signal, As rfidea pointed, You are finding capacitance at very low frequency (0.159HZ), where c1 impedance is not very low compared to R1, that's why you are getting strange result but if you calculate the same cap at frequency>1/(2PIRC) you should get same cap. Thanks, Raj.
Back to top	IP Logged

mixed_signal Senior Member Offline Posts: 183	Re: varactor in lc vco Reply #4 - Dec 27^th, 2012, 2:11pm Hi raja.cedt & rfidea, All my circuits are correct. The 1/2pi term is to get rid of division by 2pi for freq in Hz I did a mistake by choosing extremely small frequency (1/2pi) at which capacitive impedance is very large compared to resistive. The problem is solved at high frequency. Thanks both of you!
Back to top	IP Logged