Measure the inputrefered voltage&current noise in CDN without destroying the DC bias?

[ustligen1]

Measure the inputrefered voltage&current noise in CDN without destroying the DC bias?

Hi,

I try to measure conventional noise in CDN, but something strange happened.

At first, I just put two no value voltage sources @input and output respectively and do the noise simulation. Then I find it totally destroyed the DC biasing.

Then I add some choke ideal ind and cap with large values and do the simulation once again. Then the noise report is very very very suspicious with a input referred current noise three magnitude smaller than hand calc. Meanwhile, the contributors are unreasonable.

Who can teach me how to measure input referred current and volt noise. Thanks a lot.

Attachment is my schematic.

Justin

 

[erikl]

Re: Measure the inputrefered voltage&current noise in CDN without destroying the DC b

I think the noise simulation doesn't need any input nor output sources in order to introduce/measure noise: you may remove I7, C0, V1, L1, IPRBB0 and L0 (which is a cap). Just refer the resultant output noise @ vo to the input noise @ the drain-to-source connection node, which is calculated automatically from model parameters.

Of course you are free to inject additional noise via V1 - L1 . But consider how much of it will arrive at your input node due to the (frequency dependent) voltage division ratio.