Simulator LC-VCO by Cadence

[Helena1999]

Hello everyone,
I recently try to work on LC-VCO. May I ask how is the difference between .tran and .pss >>SIMULATION.
when i simulate by pss, i can see the VCO oscillate. But I use .tran, it didn't work. I simulator the differential vop/von.
Thank you !
pic 1 is my circuit
pic2 is pss result
pic3 is tran result

 

[Dominik Przyborowski]

Pss - periodic steady state is showing results for oscillator in established state. Transient is calculating total time evolution, so require a time to achieve steady state (oscillation). Moreover, default integration method (gear) is dumping oscillations in transient and can counteract to circuit behavior.
If you want to see oscillation in transient, force proper initial conditions and/or change integration method to trapezoidal (if initial conditions are not enough).

 

[Helena1999]

Pss - periodic steady state is showing results for oscillator in established state. Transient is calculating total time evolution, so require a time to achieve steady state (oscillation). Moreover, default integration method (gear) is dumping oscillations in transient and can counteract to circuit behavior.
If you want to see oscillation in transient, force proper initial conditions and/or change integration method to trapezoidal (if initial conditions are not enough).

thank you. It's really helpful.
Can I ask why my .tran result is strange
>> do i need set some initial conditions?

 

[Dominik Przyborowski]

Circuit is perfectly symmetrical and the most probably dc solution found by simulator (as the starting point) is with both transistors fully open (gates on supply and thus both sources and drain too). If you set initial condition of one output grounded and other on supply (the best one would be similar as obtained by pss) then is a chance to start simulation in proper operating point.

 

[alex_fd]

tran simulation need more time to stablize.so maybe you can set the time of tran simulation to 1us.And there is another issue,the time step.you can set the max time step to be 4ps in the option box of tran .

 

[Helena1999]

tran simulation need more time to stablize.so maybe you can set the time of tran simulation to 1us.And there is another issue,the time step.you can set the max time step to be 4ps in the option box of tran .

I appreciate it. Then I try this solution to run . But the waveform is still kinda weird. Can I ask how to figure this?
Or should I did something wrong.

 

[BigBoss]

Inject a Start-Up Current Pulse to any Drain for a limited time. Yo can use PWL Current Source for this purpose.