four stages ring vco phase noise

[jugemu1234]

ring vco

Hi all,

I've been trying to downsize the volume of VCO noise at 3GHz use.
According to spectreRF (MMSIM7.0.1) pss+pnoise analysis for VCO only, higher Kvco gives same phase noise over entire offset.
But I believe higher Kvco should bring worse phase margin since noise sensitivity gets higher.

Could anybody explain what is going on? Am I misconfiguring Or is this happening because of the fact with ideal Vctrl?

Thank you very much in advance.

 

[saro_k_82]

vco vcnt

As much as I can see and understand, you are not varying the KVCO in your sims. You are just simulating the phase noise for different VCO inputs (thereby output frequencies) so KVCO can be same. But as the frequencies are different, their phase noise can be expected to be different should they maintain the same swings. The difference you must see is just 0.26dB (3.3/3.2) which might very well be present.

 

[jugemu1234]

how many stages are in vco?

Hi Saro_k_82,

Thank you for your comment and sorry for my word short, behind the simulation, I turned on Kvco boost switch in VCO and lowered Vcnt accordingly to realize same level of osc frequency. So, Kvco is really different with same swing.

I am thinking this unchanged phase noise is due to ideal forced control line which does not contain any noise from PFD and so on. What would you think?

Thanks

 

[saro_k_82]

But I believe higher Kvco should bring worse phase noise since noise sensitivity gets higher.

Very true. But here you are talking about VCO's response to a noise at it's input. Normally when we run pnoise analysis for VCO, we are concerned about the noise generated internally.
You cant observe this as you have ideal voltage source for the control voltage., Here KVCO plays no role in phase noise. The best you can do is to run a PXF analysis and find the noise contribution from the control line to the output for different KVCOs.