The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> How to model the power supply noise??
https://designers-guide.org/forum/YaBB.pl?num=1260884939

Message started by Yutao Liu on Dec 15^th, 2009, 5:48am

Title: How to model the power supply noise??
Post by Yutao Liu on Dec 15^th, 2009, 5:48am

Dear everyone,
I designed a LDO and want to testify it could block the power supply noise and improve the performance of VCO.

I tried to use "vsource" as the power supply of the VCO and defined sine signal to model the noise. But I can't see any degradation in the transient waveform of VCO output. How should I model the power supply noise?

There are thermal noise, fickle noise, and so on. Is there any corresponding models for each of them?

Thanks in advance.
Yutao Liu

Title: Re: How to model the power supply noise??
Post by Ken Kundert on Dec 15^th, 2009, 11:59am

The easiest way is to simulate the LDO and VCO using PXF analysis, and observe the transfer function from the raw supply voltage to the output of the VCO.

-Ken

Title: Re: How to model the power supply noise??
Post by Yutao Liu on Dec 15^th, 2009, 5:52pm

Thanks for Ken's reply.
I did PXF analysis and used "vsource" as the supply. But the result showed that VCO without LDO is less sensitive to power than the one with LDO. I think lacking noise source in the power supply caused this result. So I asked how to module different kinds of noise in the simulation.

And I also tried to set a sine signal in the "vsource" and then run PXF, but there was a warning that the power supply is inconstant voltage. How should I set the PXF when there is another periodic signal in a oscillator circuit?

In addition, PXF only presents the difference in the sensitivity from the power supply to the output of VCO. If I want to check out the difference in phase noise, how should I do? Is there any additional configuration in the pnoise setting?

-Yutao Liu

Title: Re: How to model the power supply noise??
Post by sheldon on Dec 16^th, 2009, 6:17am

Yutao,

Spectre RF supports semi-autonomous periodic steady-state
analysis, that allows you to simulate circuits that contain both
driven and autonomous blocks. Also, I thought that you could
define the noise of a source. If that works then you could just
define the VCO supply noise and perform pnoise analysis. BTW,
Semi-Autonomous requires a Spectre-GXL license.

Don't quite understand your comment on the relationship
between LDO/no LDO and power supply noise.

Best Regards,

Sheldon

Title: Re: How to model the power supply noise??
Post by Ken Kundert on Dec 17^th, 2009, 8:10am

I don't understand how it is that 1) you say that PXF gives results that don't seem correct and 2) you cannot get PXF to work. Those two statements seem inconsistent.

You can perform XF analysis of the LDO alone and PXF analysis of the VCO alone and multiply the transfer functions together, and that would give you an sanity check on the result from using PXF both of them together. The result will not be identical, but it should be very similar.

You would not apply a sine signal to the power supply when using PXF. PXF applies small-signal test signals automatically as needed.

I believe ADE provides a way of configuring PXF to allow the output to be phase. If not, you can apply the equations you find in http://www.designers-guide.org/Analysis/rf-sim.pdf.

-Ken