Load-Pull Simulation in ADS - Convergence Problem -

Hello There

I have some problems and concerns running Load Pull simulation in ADS:

[1] I run the Load Pull simulation (after doing it several times in order to get the optimum area to scan) and I get satisfying results, but still I have some problems. One of them is that the run seems too sensitive to the changing of the parameters, for example if I increase the radius of the circular area just a little bit the embedded harmonic balance simulator inside the Load Pull simulation template gives convergence error problems and finally the Load Pull contours get casual. Is it natural? Should I add stabilizing circuit(s) before running the test? while device is unconditionally stable.

[2] Second problem of mine is that I'm running the Load Pull test for a device and in it's datasheet the Load Pull counters and results for five frequency is given, I'm trying to run the Load Pull in order to get the exact results stated in the datasheet but my simulation gives me different results and even much better ones! while i run the test in the exact conditions stated in the datasheet and I also have valid model of the device provided by the manufacture.

Thank you so much in adance

Seems modeling problem if the datasheet says that the transistor is unconditional stable whatever the load is.

Thank you BigBoss
But it's sure the model is 100% measurement based valid model. But let me add this also that although the device is unconditionally stable over whole frequency range but it's stability factor "mu" is just slightly bigger that 1, like 1.01 and like that. So regarding that these stability factors I mean "mu" as well as others like "k" and "delta" are small signal stability parameters and Load Pull and harmonic balance are basically large signal simulations, is it possible that I have unconditionally small signal stability but not necessarily large signal stability? As we know it's quite possible that small signal stability doesn't insure large signal stability too.

Regrads

lastquark said:

Thank you BigBoss
But it's sure the model is 100% measurement based valid model. But let me add this also that although the device is unconditionally stable over whole frequency range but it's stability factor "mu" is just slightly bigger that 1, like 1.01 and like that. So regarding that these stability factors I mean "mu" as well as others like "k" and "delta" are small signal stability parameters and Load Pull and harmonic balance are basically large signal simulations, is it possible that I have unconditionally small signal stability but not necessarily large signal stability? As we know it's quite possible that small signal stability doesn't insure large signal stability too.

Regrads

Click to expand...

Practically yes, small signal stability cannot principally be a figure for the large signal stability due to OP of the device is changed in a large swinging area.I think you should take precautions first that make the device "evidently" stable then try again.A small AC coupled resistor ( around 100-300 Ohm more or less ) across D and S or maintain a small series feedback at Source or a light parallel feedback may improve the stability problem.

[1] Is the convergence of harmonic balance caused by the saturation of the transistor. If it was in saturation, it often takes a long time for the harmonic balance simulation. If you change the bias voltage, does it help?

[2] Have you terminate the input and output to 50 ohms?