Some questions on VCO design

[Adam2008]

Hi All,

I am designing a cross coupled VCO using CMOS technology. Now I got some question of the tuning frequency.
The frequency is always much higher than I expected. It can not achieve the lowest frequency which the varactor is supposed to. What is the reason of this problem? Can I measure the capacitance of the varactor? I am using Cadence spectreRF.

Thanks.

 

[khouly]

mmm , strange
usually the oscillator will give u lower frequency due to alot of parasictics u didn't take into account in the hand calculation

khouly

 

[jecyhale]

Hi All,

I am designing a cross coupled VCO using CMOS technology. Now I got some question of the tuning frequency.
The frequency is always much higher than I expected. It can not achieve the lowest frequency which the varactor is supposed to. What is the reason of this problem? Can I measure the capacitance of the varactor? I am using Cadence spectreRF.

Thanks.

Maybe the inductor model is not precise.
Or do you use smoe special structure?
Usually, the frequency should be low than we design.

 

[khouly]

how much differnece between the calculated and simulated frequqncy

try to send a schematic of ur oscillator

thanks

khouly

 

[Adam2008]

I am using one symmetric inductor. WIll the value of inductance be different from claimed in the manual?

 

[khouly]

mmm , maybe there is a difference between them but how much difference u have in the frequency

khouly

 

[Adam2008]

The difference is quite big. If the calculated max frequency is 4.5GHz, the simulated frequency is between 5GHz and 6GHz.

 

[khouly]

man , there is some error
may be u put the capacitor or the vacator less , than the design

recheck ur tank circuit

is it a crosscoupled oscillator

khouly

 

[Adam2008]

yes, it is cross coupled vco.

 

[khouly]

mmm ,
the cross coupled oscillator add a parasitic capacitance due to transistors ,

u need to check the resonator , may be the varactors are not right

khouly

 

[yours]

The difference is quite big. If the calculated max frequency is 4.5GHz, the simulated frequency is between 5GHz and 6GHz.

How did you caculate the frequency?By programming or other quomodos.
Maybe your model is not exact enough.
I need knowing your means.
thank you.

 

[Adam2008]

I calculated the frequency by f = 1/ (2 *pi*sqrt(LC)).

 

[khouly]

this is the right equation ,
C is the varactor capacitance + parasitic capacitances
L is the inductance of the coil.
then u have an error in one term of these

another thing , u may have an error in SpectreRF setups. did u follow the manuals and appnotes given by cadence

Khouly