The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> Stability for cascaded amplifiers
https://designers-guide.org/forum/YaBB.pl?num=1049348496

Message started by Thomas_Hansson on Apr 2^nd, 2003, 9:41pm

Title: Stability for cascaded amplifiers
Post by Thomas_Hansson on Apr 2^nd, 2003, 9:41pm

Hello
Does anyone have any experience from simulating stability for cascaded RF amplifier stages? The probem is that K-factor (as part of s-parameter analysis) is only valid for one stage. The classical bode diagram approach could be used if the feedback loop could be opened and the loopgain could be estimated. In my case the frequency is around 5GHz and there is a lot of unwanted "parasitic loops" but no regular feedback loop. So far I have used a impulse respone test with varying load using the port-adapter in Cadence/spectre.

Thomas

Title: Re: Stability for cascaded amplifiers
Post by Raccoon on Jun 13^th, 2003, 7:14am

Hi Thomas,

first of all i am quite new to this, so if my answers are not correct i hope someone will correct them.
Can´t you simulate every stage for itself and measure the K-factor? If you can ensure that every stage is unconditional stable, your whole system should be too, since you have no feedback loops. You can even simulate together with the other stages if you cut the line between the two stages and insert a port which represents the input impedance of the second stage as load to the first and another port with the output impedance of the first stage as source for the second stage which also delivers the power that ´d be delivered by the first stage. (uh, what a sentence i hope you understand my English)...

Maybe that helps.

Regards, Raccoon

Title: Re: Stability for cascaded amplifiers
Post by pancho_hideboo on Jun 15^th, 2003, 9:26am

See the following paper.

"The S-probe-a new, cost-effective, 4-gamma method
for evaluating multi-stage amplifier stability".
Wang, K.; Jones, M.; Nelson, S.;
Microwave Symposium Digest, 1992., IEEE MTT-S International , 1-5 Jun 1992
Page(s): 829-832 vol.2

Such analysis is used for PA design which have multiple stages. :P

Supplement.

In the above paper, author implemented this method using legacy touchstone netlist, so many of you might not be able to understand netlist.
In this netlist, we can calculate four impedances, Zs, Zs11, Zs22, Zload at just one s-para simulation.
Here to realize this calculation, author are using 10 ports for sp-analysis.

But if you can permit two simulations instead of just one,
you can do same calculation more easily.

For example, if you use Spectre of Cadence, insert current source for evaluation of impedance using ac-ana, then using Ocean, switch this current source to evaluate impedance for another side.

:o