noise vs s-paramter simulation for noise figure

Dear asetji,

asetji said:

i have tried both noise analysis and parameter analysis. However, i get different results from the two analysis.

What is the source of your S-parameter file? Is it measured data or simulated data? Have you examined your S-parameters in ViVA? If I were Tawna Wilsey, whose background and knowledge I can only aspire to, she would immediately ask you to study your S-parameters carefully as it is very common to get incorrect simulation results if the S-parameters are not appropriate for the simulation. I might suggest you examine Tawna's note(s) at URL:

support.cadence.com/.../ArticleAttachmentPortal

If you are not familiar with the Rapid Adoption Kit (RAK) for Noise and S-Parameter Measurements of a FET at URL:

support.cadence.com/.../ArticleAttachmentPortal

it might be worth reviewing it to make sure your simulation set-up is not overlooking something.

Sorry I cannot provide more specific advice asetji!

Shawn

Thanks ShawnLogan,

unfortunately my reference key is expired so cannot access the support documents that u have referred.

Regarding the circuit, it is a mixture of extracted schematics (R,C,CC) and EM simulated layouts.

BR

Dear acetyl,

asetji said:

unfortunately my reference key is expired so cannot access the support documents that u have referred.

Rats - that is too bad as I think they might provide some insights into your quandary! 

asetji said:

Regarding the circuit, it is a mixture of extracted schematics (R,C,CC) and EM simulated layouts.

I assume the EM simulated layouts are modeled as S-parameters. If so, you can still examine the S-parameters for the relevant subcircuits in ViVA directly. From te=he ViVA window, under the menu item File, choose Open Results.... Select your S-parameter file. Its sufficient should be something like <s_paramter_filename>.snp where n is the number of ports.

Meanwhile, perhaps you could ask whomever is responsible for your Cadence tools for access to the On-line support portal.

Shawn

Hi,

finally got the renewed key and had a look at the support documents.

However, they didn't really help me in solving this mystery.

Any other ideas?

BR

Dear asetji,

asetji said:

finally got the renewed key and had a look at the support documents.

However, they didn't really help me in solving this mystery.

Any other ideas?

Great!  How did the S-parameters for each file you are using look? Is the frequency resolution sufficient? What are your simulation settings? Are your networks non-linear? Are your tool versions consistent with those recommended in the Cadence On-line support articles whose links I included?

Speaking for myself only, without more details about your simulation setups, S-parameter files, and circuit, all I can do is guess by asking questions about them. Andrew often recommends that one contact customer support if the details are proprietary.

Sorry I cannot be more help...

Shawn

Hi asetji,

In situations like yours I start by creating a simple test case circuit that I fully understand and know in advance what the simulation results will be. Perhaps a simple text book circuit or create your own. Simulate the test case circuit and experiment with the test bench until you get the expected results. If you can't get the simple circuit to work as expected there is no benefit in wasting time on a more complicated circuit. You must conquer the basic circuit simulation first. Then add more complexity to the circuit (perhaps replace the IMN with s-parameter file) and repeat.

Brad