What's your tbest technique in Stabilizing your Power Amplifier?

[Wall-eee]

hello,

my raw transistor PA model seems to be unstable for all my desired frequency range at my bias point.

2 questions:

1. based on ur experience, would u suggest doing stability analysis after matching networks of before? because the matching networks will certainly present losses to some band of frequencies.

2. what's your recommended technique in stabilizing a power amplifier ? (please exclude feedback RLC network as my supervisor won't allow it). Some articles say selective frequency mismatcput hing or putting a shunt RC or RL network at the input to make a high gain freq. lossy....any further comment or advise here?

 

[jiripolivka]

No details are given so it is difficult to offer an advice.
Due to usually high gain, do include a good screen between input and output, and use adequate blocking of DC bias for the input gate or base bias. Make sure there cannot be any coupling back to signal source.

 

[Wall-eee]

No details are given so it is difficult to offer an advice.
Due to usually high gain, do include a good screen between input and output, and use adequate blocking of DC bias for the input gate or base bias. Make sure there cannot be any coupling back to signal source.

hello, my thesis is on a 50W broadband class AB single ended PA design in UHF 480-860MHz.
when i simulated the geometric stability as well as K and B, the device is unstable.

but i proceeded in designing my matching network...another common stability solution is putting small value resistor towards the gate input path.

what design solution did you employ when you encountered real stability issues on your PA design?

 

[vfone]

Playing with the stability of a PA is not as simple as it look.
First step would be choosing the input and output matching network topology. Usually a high-pass network for the input, and a low-pass for the output, get the best performance in terms of stability.
On the other hand, placing any degenerative resistors on a Power Amplifier would decrease linearity, efficiency, and gain. These should be taken into account.

 

[Wall-eee]

Playing with the stability of a PA is not as simple as it look.
First step would be choosing the input and output matching network topology. Usually a high-pass network for the input, and a low-pass for the output, get the best performance in terms of stability.
On the other hand, placing any degenerative resistors on a Power Amplifier would decrease linearity, efficiency, and gain. These should be taken into account.

i totally agree with the issues of adding degenerative resistors at the gate...

i have a 4-section LC network. do you recommend all HP network at the input and all LP network at the output?

right now, my input network from source (50-ohms) to gate (1.5+j1.6) of the PA is HP-LP-LP-LP with a lumped series (CAP) as my reactance cancellation.

for my output network from drain (1.5+j0.3) to load (50-ohm) is first a lumped shunt (CAP) reactance cancellation followed by LP-LP-LP-HP network..

so would you rather recommend to make it all HP for input and all LP at the output?