Designer’s Guide Community

sivankumarpsp New Member Offline Posts: 6 India	PSRR of VCO Mar 26^th, 2009, 12:40am How do i measure the PSRR for VCO? What is the exact definition of PSRR for VCO? I am trying to give a sine wave of some frequency at the supply and trying to find out the change in the frequency and then take the ratio of frequency change/supply chage. Is it correct? can any one explain?
Back to top	IP Logged

raja.cedt

Senior Fellow

Offline

Posts: 1516
Germany

Re: PSRR of VCO
Reply #1 - Mar 26^th, 2009, 1:16am

hi,
whats your supply noise profile,because if it is really contains only one harmonic then sine is ok..but most of the supply noise spectrum's are like band pass pass with cutoff frequency in mega hertz.so better always apply square wave with magnitude based on the peak of the spur in noise spectrum(some times people find PSRR by using AC analysis,but this is acceptable when noise magnitude is very low,but now a days noise having considerable magnitude and that to your doing PLL i guess..so you have to use square wave with proper rise time).I don't think variation in frequency is important..because any how PLL loop with take care..so rather than finding frequency variation find change in Kvco..because it will effect loop dynamics...
with my very less design experience i gave the above ans..please correct me if i am wrong

Thanks,
Rajasekhar.

Back to top

IP Logged

Visjnoe

Senior Member

Offline

Posts: 233

Re: PSRR of VCO
Reply #2 - Mar 26^th, 2009, 8:08am

Several things can be checked here:

1. Sensitivity of the VCO frequency with regard to supply variations.
As mentioned, typically not very relevant since the PLL will *eventually* counteract the change...but take care! If the changes happen faster than your PLL/loop can follow, you will see an effect.

2. The appearance of spurs in the VCO spectrum due to tones on the supply. See theory about FM modulation. If possible use a separate regulator for your VCO and don't run anything else from it.

What tones can you expect? Look at your system: harmonics or mixing products of clocks.

Regards

Peter

Back to top

IP Logged

raja.cedt

Senior Fellow

Offline

Posts: 1516
Germany

Re: PSRR of VCO
Reply #3 - Mar 26^th, 2009, 12:15pm

hi peter,
Thanks for your reply,if i use any seperate Regulator only for VCO that's fine...i think in that case for PSRR ac analisys is enough,but its bandwidth should be very high..In your reply i have one doubt..even though VCO frequency changes and finally after several cycles PLL will correct even thogh it's bandwidth is low

Thanks,
Rajasekhar.

Back to top

IP Logged

sivankumarpsp New Member Offline Posts: 6 India	Re: PSRR of VCO Reply #4 - Mar 30^th, 2009, 6:08am Hi, Thanks a lot for your answers. Can i do PSS and PNOISE analysis to find out the phase noise in the output of VCO due to the ripples in the supply? If that can be done, what kind of test bench i need to make? Regards, Sivan
Back to top	IP Logged

Ken Kundert Global Moderator Offline Posts: 2384 Silicon Valley	Re: PSRR of VCO Reply #5 - Mar 30^th, 2009, 4:04pm You would use PXF. -Ken
Back to top	IP Logged

sivankumarpsp New Member Offline Posts: 6 India	Re: PSRR of VCO Reply #6 - Mar 30^th, 2009, 9:18pm Hi, Can any one give me some idea how to use PXF simulation for finding out phase noise? Sivan
Back to top	IP Logged

Andrew Beckett

Senior Fellow

Offline

Life, don't talk to
me about Life...

Posts: 1742
Bracknell, UK

Re: PSRR of VCO
Reply #7 - Mar 31^st, 2009, 3:18am

I presume you don't really mean phase noise? For that you'd use pnoise analysis.

To find the transfer function from a perturbation on the supply to the output of the VCO, you'd use PXF, as Ken says. There is also the "modulated" mode of PXF, which allows you to separate the PM and AM components of the modulation - some of the supply variation may not affect the frequency, but just the amplitude - so being able to separate them may be useful.

To run PXF, you wouldn't apply any varying signal on the supply - you'd tell PXF where the output of the circuit is, and the output frequency (much as you'd specify the output for pnoise). Then when the simulation is finished, you can use the direct plot form to click on the source for the supply, and see the transfer function(s) from this source to the outputs.

Regards,

Andrew.

Back to top

IP Logged

Yutao Liu

Community Member

Offline

Posts: 76
Guangzhou, China

Re: PSRR of VCO
Reply #8 - Sep 5^th, 2009, 1:20am

Andrew Beckett wrote on Mar 31^st, 2009, 3:18am:

I presume you don't really mean phase noise? For that you'd use pnoise analysis.

To find the transfer function from a perturbation on the supply to the output of the VCO, you'd use PXF, as Ken says. There is also the "modulated" mode of PXF, which allows you to separate the PM and AM components of the modulation - some of the supply variation may not affect the frequency, but just the amplitude �- so being able to separate them may be useful.

To run PXF, you wouldn't apply any varying signal on the supply - you'd tell PXF where the output of the circuit is, and the output frequency (much as you'd specify the output for pnoise). Then when the simulation is finished, you can use the direct plot form to click on the source for the supply, and see the transfer function(s) from this source to the outputs.

Regards,

Andrew.

Hello Andrew, I have the same question as Sivan's.

Since PXF gives us the transfer function from the power supply to the output of VCO, we only get to know VCO's power supply sensitivity, which is not the phase noise due to the ripple on supply path.
And in the "modulated" mode in PXF, even when PM is chosen in "Main Form", I only get the phase component of modulation from power supply to output of vco, not the phase noise yet.

Do you have any idea to change this result to the phase noise contributed by supply noise?

Regards,
Yutao

Back to top

IP Logged