The Designer's Guide Community
Forum
Welcome, Guest. Please Login or Register. Please follow the Forum guidelines.
Jul 16th, 2024, 4:07pm
Pages: 1
Send Topic Print
VCO Pnoise simulation setup question (Read 7911 times)
James
Community Member
***
Offline

snowfiled

Posts: 32

VCO Pnoise simulation setup question
Jul 08th, 2005, 9:54pm
 
Dear all,

When I setup the Pnoise Analysis for my VCO, in the setup form, there is an item"Input Source", I choose "voltage", then it shows "Input Voltage Source" with a bank form on the right. I should select a source.

Which voltage source I should choose? the power supply source (vdd)  or the input signal source(i.e. the input control voltage of VCO)?

Thanks a lot.
Back to top
 
 

James
View Profile   IP Logged
Ken Kundert
Global Moderator
*****
Offline



Posts: 2386
Silicon Valley
Re: VCO Pnoise simulation setup question
Reply #1 - Jul 8th, 2005, 10:08pm
 
If you specify an input source the PNoise analysis will compute the noise figure or input referred noise. That is appropriate for amplifiers and mixers, but not oscillators. Normally, one does not specify an input source with oscillators.

-Ken
Back to top
 
 
View Profile WWW   IP Logged
James
Community Member
***
Offline

snowfiled

Posts: 32

Re: VCO Pnoise simulation setup question
Reply #2 - Jul 9th, 2005, 9:20am
 
Dear Ken,

Where I can find the detailed setup information for PSS and Pnoise analysis for VCO, PFD/CP and divider? Now you tell me how to setup "input source", but maybe I have other mistakes I dont know. So I cannot get correct result with that wrong setup.

Best
Back to top
 
 

James
View Profile   IP Logged
Ken Kundert
Global Moderator
*****
Offline



Posts: 2386
Silicon Valley
Re: VCO Pnoise simulation setup question
Reply #3 - Jul 9th, 2005, 10:24pm
 
Run cdsdoc, open the SpectreRF user manual, and read chapter 5 "Simulating Oscillators".

-Ken
Back to top
 
 
View Profile WWW   IP Logged
James
Community Member
***
Offline

snowfiled

Posts: 32

Re: VCO Pnoise simulation setup question
Reply #4 - Jul 17th, 2005, 10:14am
 
Dear Ken,

Thanks for your answer which helped me a lot in VCO simulation. Now I am simulating another block's pnoise, a 3 stage ripple counter.

The input frequency is 200M, the output is 25M. In the PSS setup form, I set beat frequency as 25M.

I have a few questions on Pnoise setup:
1. For the "Input Source", which option I should choose. I choose voltage and tried to select the wire of the input signale, but it give me the "object selected is not a valid selection object" message. Then I changed to current, port, but it gave me the same msg.

Should I choose input source as "none" ? But in your pnoise+jitter, it says, for divider, the input noise dominant. So I don't think it's should be set to none. But in my simulation, I set it to none since I couldn't set it to other values.

2. In pnoise setup form, I choose Noise Type as jitter. I think it will directly give me the ripple counter output jitter. Since the PFD is the next stage after the counter, I set the cross direction as "rise" and Threshold Value as 0.63V. I choose 0.63V because the nmos threshold voltage is 0.63V. I am not sure if I am correct on setting this value.

3. After simulation, I try to watch the result. In the pnoise jitter result option, there are "cycle to cycle", "long term", "edge to edge" options. Which one we usually use for jitter calculation?

4. I choose "cycle to cycle", and integrte range is from 1 to f0/2, f0 is the beat frequency. That is from 1 to 12.5MHz. But the result is rather small: 67fs. I changed to "long term" and "edge to edge", althouge the results are difterent, they are still in fs range. I wonder this result comes from my possibly wrong "threshould value", after I reset it to 2.5V (the supply if 5V), I got jitter 71fs. At this point, I suspect it's the reason I set "Input source" as none, but I am not sure. Your help will be highly appreciated.
Back to top
 
 

James
View Profile   IP Logged
Ken Kundert
Global Moderator
*****
Offline



Posts: 2386
Silicon Valley
Re: VCO Pnoise simulation setup question
Reply #5 - Jul 17th, 2005, 12:36pm
 
1. Specify that the input is a vsource, and then actually choose a vsource (not a wire).
2. Seems reasonable to me, but you are neglecting the pmos turn-on threshold. Perhaps it would be better to choose a value between the two?
3. I have no guidance to offer on this one. It really depends on your needs. Look at my PLL jitter paper to find the definitions for the three types (edge-to-edge is what I call period jitter).
4. I'm afraid you will have to try things to either debug the problem or convince yourself you are getting the right answer. Not having an input source will not affect the output noise results. It only determines whether SpectreRF calculates the the input-referred noise.

-Ken
Back to top
 
 
View Profile WWW   IP Logged
Pages: 1
Send Topic Print
Copyright 2002-2024 Designer’s Guide Consulting, Inc. Designer’s Guide® is a registered trademark of Designer’s Guide Consulting, Inc. All rights reserved. Send comments or questions to editor@designers-guide.org. Consider submitting a paper or model.