Questions about the stabilty of a PA

linear PA which usually have two amplifying stages.
1. Should each of the stage have a stability factor K larger than unity? Or the K of the whole PA?
2. Should the K larger than unit in the low frequency range(<1MHz)?
3. How to simulation a large signal s parameter to find out the LSSP stability,
is there an example in ADS?
3ks

There are lot of examples available at the installation folder of Agilent ADS

For a multi-stage RF PA important is the total K stability factor.
The PA has to be stable for different output loads and phases, and also should be stable at low frequencies.

for multistage amplifier, i think each of stage should have stability factor K larger than unity. IMO, if each stage has K larger than unity, we can assure that total K larger than unity.

in my experience, total K larger than unity, doesnt guarantee that every single stage has K>1. so a single stage possibly experiences instability. instability of a single stage causes instability of PA.

regards

I think both single stage and multistages k factor should be larger than unity.

I have to explain a little bit more.
First of all, if K factor is lower than unity doesn?t mean that the stage it will oscillate. That means the stage is potential unstable for different input/output impedances.
The changes of the impedances internally are not so significant compared to changing of impedances externally of the PA. So, the first priority is overall K factor.

K is not good enough for all cases, such as balanced amps and other structures. Look up "The S-Probe Approach to Amplifier Stability Analysis" from a guy at Triquint. Ansoft also developed a method; remember control theory class and drawing the spiral stability graphs.