I pushed up a copy of the RevD schematic under PacSatDocs->Hardware. This is not complete or finished but does have the changes I have been working with here relating to the PA and other fixes.
I am currently biasing the gate of the PA with an external voltage supply set to 2.7 volts and 100 mA. This can be done with a voltage divider and a voltage reference later. Tenths of a volt mater here so best to have a good reference in the long run. This node pulls almost no current.
The drain supply and current are setup by using a 50 ohm resistor in series with the 5V regulator. I see the voltage has dropped to about 4.9 volts due to series inductor and trace resistance. This can be improved on the next PCB layout. There is about 3 volts on the drain with this combination and the resulting overall current low enough so the current monitor MAX4995 does not shut things down when set to its max of 600 mA. This may need to be eliminated when we push for more power and get the drain bias up to 5V. More testing is needed.
The gate is being driven with about 0 dB or RF and the AX5043 is capable of 16dBm according to the data sheet. There is currently a filter in this circuit to reduce harmonics which may need to be eliminated or optimized to get the power level up on the gate of the PA. The final output filter is doing a good job of stopping harmonics. However feeding harmonics into the PA can cause some heating and power loss.
Having the TMS570 switch the PA power on only when transmitting will result in some power savings. When on the PA will have a substantial bias current. Some delay for power to settle may be needed.
My goal here was to document my thought and work.
Bob