The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> Phase noise floor CMOS buffer
https://designers-guide.org/forum/YaBB.pl?num=1216229011

Message started by Visjnoe on Jul 16^th, 2008, 10:23am

Title: Phase noise floor CMOS buffer
Post by Visjnoe on Jul 16^th, 2008, 10:23am

Dear all,

when simulating the phase noise of an oscillator + CMOS buffer it turns out that the wideband phase noise is dominated by contributions of the CMOS buffer. The simulator returns a phase noise floor of -160dBc/Hz (broadband noise). Can the simulator be trusted to resolve such low levels accurately (assuming accurate device models)?

I would say yes, but I would like to hear if someone thinks otherwise. The reason for asking is a discrepancy between measurements and simulations. Assuming a phase noise floor of -150dBc/Hz would lead to agreement between measurements and simulations.
Personally, I think 10dB is a huge difference so I don't think simulator inaccuracy is a likely cause...

Regards

Peter

Title: Re: Phase noise floor CMOS buffer
Post by ACWWong on Jul 16^th, 2008, 12:42pm

I don't think simulator accuracy is to blame... maybe simulator setup or model accuracy, test bench setup, inclusion of all losses like layout parasitics etc.
Previously I have simulated around -160dBc/Hz and also measured within a dB or so of the simulation on the bench.
Are you getting the expected output power & frequency ?

Title: Re: Phase noise floor CMOS buffer
Post by Visjnoe on Jul 16^th, 2008, 11:30pm

Dear ACWong,

yes, power and frequency are correct.
I also ran simulations on an extracted netlist, including bondpad/package parasitics + noise from the biasing.

Regards

Peter

Title: Re: Phase noise floor CMOS buffer
Post by ACWWong on Jul 17^th, 2008, 2:12am

i guess if the output power is expected then this is evidence that the VCO drive to the buffer is as expected also. is there any VCO measurement without the buffer ?

how is the measurement setup, is it capable of measuring -160dBc/Hz ?
is the setup run from a battery ?

otherwise i guess the noise modelling of the devices isn't up to scratch then... 10dB is quite alot.