Rob and John's hypothesis about oscillation may be right, a
resonance could generate massive voltages.? This is different to
thinking of an impedance mismatch.
It could be far more variable, far more difficult to replicate.?
This could explain Hans not seeing it in his tests. Maybe at
fairly low SWRs.
Perhaps diode(s) could protect against high voltage problems
while monitoring SWR and reducing power protects against high
current problems.? No reason not to have both, belt and braces.
A diode across the inductor in the power to the PA FETs? would
conduct if the voltage went above the supply plus a diode drop,
could that protect as well as Zeners across the PA FETs?? Or would
it conduct too early?? 5W into 50 ohms has a peak voltage of
22.5V, with the 1.5:2 output transformer this gives a peak voltage
of 16.9V at the BS170 drains.? This is for a 12V transformer, less
for the 1.5:3 9V transformer.
I'm afraid I'm thinking of QMX and it's additional options with
SWR measurement and power control, but it all seems relevant to
both.
Chris, G5CTH
