LC VCO doesn't oscillate in transient analysis

[boshiouke]

Can anyone comment on how to get the LC VCO to oscillate in transient analysis please? Initial conditions are set, and PSS/Pnoise works fine. But in transient analysis VCO simply doesn't oscillate. Thank you.

 

[tfwee]

I think you may want to check the following:

1) have to inject impluse current at the output node of the VCO to initiate or speed up your oscillation.
2) You may have to run a longer period to check whether your circuit take a longer time to start oscillating.
3) To limit your maxstep in the time step parameter in the option of transient analysis if your are using spectre
4) Another method is to ramp up your supply voltage in transient analysis.

One thing to point out, if your transient analysis is not oscillating, are you sure that the result you get for PSS/Pnoise is working fine. In PSS, you should be able to plot the output waveform too, can you confirm that the output is correct.

 

[dingjingfeng]

maxstep is usually set to at least 1/10 of the aimed frequency period.

 

[pi331133]

The transient may need long time to start up oscillation.

 

[Willem]

What software are you using to simulate this oscillator? If its ADS or Pspice i can help.

 

[nike2000]

The transient analysis need long time at least >1ms tran time.
vdd signal from 0 to vdd level with rising time.
Also the fact using .ic (initial condition) it uses it is one method.

 

[flatulent]

Inside SPICE, Overcoming the Obstacles of Circuit Simulation by Ron Kielkowski deals with this and other problems. The book is from 1994 and may be found in a local library.

In general you have to tighten up all of the tolerance statements by factors of 10-1000 and shorten the time step. This increases the simulation time, but it will produce correct results.

Do not depend on the help department of the software vendors. They are non-hardware engineers who only know the answer to the 100 most frequently asked questions which are mostly about how to use the program.

 

[eejli]

try to power down and power up your vco. It will startup.

 

[neoflash]

It might not be a simulation issue, since you already tried add pulse injection.

Are you sure your circuit will oscillate?

What happens if you inject a pulse into your output node? Does it ring and damp?

What happens if you increase your tail current or your cross-coupled NMOS pair?

 

[billywu]

Hi,
What do you mean by "ramp up the supply voltage in tran."?
To add rise time for my Vsupply? Or to increase and decrease my supply periodically?

4) Another method is to ramp up your supply voltage in transient analysis.