cmdmprobe from analogLib

Dear illaoi,

illaoi said:

(3) the current the "iprove vinj" has times CMDM?

Yes.

illaoi said:

Also is CMDM=1 differential or common-mode stability?

From URL:

support.cadence.com/.../techpubDocViewerPage

"Common Model Differential Model Probe
This is a Spectre subcircuit component used in Spectre stability analysis for measuring
differential stability. It measures the common-mode stability when CMDM is set to 1 and
differential-mode stability when CMDM is set to -1."

I am not sure of your need for this probe. However, as I detailed in the Troubleshooting article in this recent post:

community.cadence.com/.../fully-differential-opamp-stb-simulation

Cadence recommends the use of diffstbprobe for both common-mode and differential mode stability analyses. If you are looking to just understand the theory behind this schematic, Frank Wiedmann has an extensive note detailing the background (including references) of single and double injection as a means of determining loop gain and phase. His note is at URL:

https://sites.google.com/site/frankwiedmann/loopgain

I will not speak for Frank, but I believe he also recommends the use of the Cadence diffstbprobe in lieu of the cmdmprobe from his following response some years ago

at URL:

https://community.cadence.com/cadence_technology_forums/f/custom-ic-design/23009/cannot-perform-stb-analysis-in-cadence

I hope this helps a bit illaoi!

Shawn 

Dear illaoi,

illaoi said:

(3) the current the "iprove vinj" has times CMDM?

Yes.

illaoi said:

Also is CMDM=1 differential or common-mode stability?

From URL:

support.cadence.com/.../techpubDocViewerPage

"Common Model Differential Model Probe
This is a Spectre subcircuit component used in Spectre stability analysis for measuring
differential stability. It measures the common-mode stability when CMDM is set to 1 and
differential-mode stability when CMDM is set to -1."

I am not sure of your need for this probe. However, as I detailed in the Troubleshooting article in this recent post:

community.cadence.com/.../fully-differential-opamp-stb-simulation

Cadence recommends the use of diffstbprobe for both common-mode and differential mode stability analyses. If you are looking to just understand the theory behind this schematic, Frank Wiedmann has an extensive note detailing the background (including references) of single and double injection as a means of determining loop gain and phase. His note is at URL:

https://sites.google.com/site/frankwiedmann/loopgain

I will not speak for Frank, but I believe he also recommends the use of the Cadence diffstbprobe in lieu of the cmdmprobe from his following response some years ago

at URL:

https://community.cadence.com/cadence_technology_forums/f/custom-ic-design/23009/cannot-perform-stb-analysis-in-cadence

I hope this helps a bit illaoi!

Shawn 

Following on from Shawn's reply, you should not use cmdmprobe. It is obsolete. The big clue is when you try to instantiate it:

It was replaced because it didn't handle unbalanced differential loops, and the diffstbprobe is a much better way of doing this - then the choice as to whether you are measuring common-mode or differential-mode loop gain is a choice on the stb analysis form rather than being a parameter of the probe device. 

Andrew

ShawnLogan said:

I will not speak for Frank, but I believe he also recommends the use of the Cadence diffstbprobe in lieu of the cmdmprobe

This is of course absolutely correct. Thank you also for referencing my website; I explain the diffstbprobe there in the section Loop Gain of Differential Circuits.

Thanks to all the replies, understood.

Since we are on the note of stability, I wanted to make sure on another thing, assuming a fully differential block, it can have cascade of differential amplifiers, or amplifier plus active filter, ..., all differentials.

I want to know, is having two ports at input and output, and looking at k-factor sufficient for the system to be stable differentially?

also, for common mode stability, is shorting the inputs and connecting to input port, and shorting the outputs and connecting to output port and looking at k-factor if greater than one is sufficient to make sure the system has common mode stability at all cases?

In other words, is choosing the short route of doing so rather than doing individual block differential and common mode stability more efficient?

K-factor stability is a different concept with limited applicability, see for example www.microwaves101.com/.../Stability-Analysis-Presentation.pdf

Dear illaoi,

illaoi said:

Since we are on the note of stability, I wanted to make sure on another thing, assuming a fully differential block, it can have cascade of differential amplifiers, or amplifier plus active filter, ..., all differentials.

I want to know, is having two ports at input and output, and looking at k-factor sufficient for the system to be stable differentially?

The use of the K factor to determine if an amplifier is unconditional stable is limited to a single-stage amplifiers. It may not predict oscillation between its stages.

illaoi said:

also, for common mode stability, is shorting the inputs and connecting to input port, and shorting the outputs and connecting to output port and looking at k-factor if greater than one is sufficient to make sure the system has common mode stability at all cases?

The use of K to predict unconditional stability is really limited to single-stage amplifiers where the input and output impedances can be well defined. Stability circles can be more helpful to assess multistage amplifier stability.

illaoi said:

In other words, is choosing the short route of doing so rather than doing individual block differential and common mode stability more efficient?

It is more efficient, but it may not lead to a correct assessment of stability!

As a side note, the use of solely K to predict single amplifier stability can also lead to incorrect results. There are a couple of examples in reference [1] where the specific conditions for K to predict stability are specifically defined.

Shawn

[1] A. Platzker, W. Struble and K. T. Hetzler, "Instabilities diagnosis and the role of K in microwave circuits," 1993 IEEE MTT-S International Microwave Symposium Digest, Atlanta, GA, USA, 1993, pp. 1185-1188 vol.3, doi: 10.1109/MWSYM.1993.277082.