Matching Circuit Tool generates wrong results

[tomhajjar]

Update:

The device seems to be marginally stable. Project updated with all fixes

Follow up to #401, @andresmmera

In testing Matching Circuit Tool for an "L-section match" case, I found the component values to be unrealistic.

The input match L1 shunt inductor value of "0.000H" should be interpreted as a short to ground or possibly using "creative license", an open circuit to ground. The output match L4 series inductor value of "2.556H" can only be interpreted as an open circuit.

I didn't test any other cases since having to copy and paste S-parameter values becomes tedious and error prone.

BFP193W_amplifier_prj.zip

[andresmmera]

Hi @tomhajjar , thanks for pointing this out.

The device is not unconditionally stable and it cannot be conjugately matched. Equations 12.40a/b in [1] don't have a solution for this case since the argument of the square roots is negative.

I think that the best solution would be to check if the device is unconditionally stable here in order to avoid wrong results like this. I can write a fix for this later.

Schematic: BFP193W_spar_analysis.zip

[1] Microwave Engineering. David M Pozar. 4th Edition. 2012

[tomhajjar]

Not often a device's Spice model correlates so well with it's S-parameter data...

[tomhajjar]

@andresmmera

One minor issue with the tool is if any entry or setting is "wrong" or out of bounds, a error message is shown and the tool is closed. Is there a way to allow the user to fix the wrong entry and not close the tool?

[andresmmera]

Good point. I'll take a look later.

[ra3xdh]

Fixed by #907. Closing as completed.

[andresmmera]

@andresmmera

One minor issue with the tool is if any entry or setting is "wrong" or out of bounds, a error message is shown and the tool is closed. Is there a way to allow the user to fix the wrong entry and not close the tool?

Fixed in PR #909

[tomhajjar]

This Qorvo tool is useful for seeing the effect of R on K.