[hpsdr] Alex RF Leakage

[Jim Gailer]

Hi
Before connecting Mercury to the T/R switch I decided to check the RF
leakage on transmit (don't ask, I learned the hard way on another project!).
On the lower bands, the results are acceptable but I'm nervous of the 6M
result. Here are the measured receiver pk-pk input voltages measured on a
60MHz digital scope (Tektronix TDS210) terminated in 50ohms). The filter
boards are grounded to the case as advised (though that made little
difference).

Band, millivolts pk-pk
160, 22
80, 600
40, 1200
20, 1500
6, 3000

These were all measured with a transmitted power of 12.5 watts into a 50 ohm
load, a mismatch could make it worse. I expect that the 6 metres figure is
an underestimate because of the 'scope bandwidth but that it would, in a
worst case, be limited by the low noise amplifier saturating.
At 100 watts on this basis, I would expect to be stressing the Mercury input
protection somewhat, but I'm not very familiar with the protection device
(it's too small to absorb power for long!) - is it OK to go ahead, or have I
got a fault condition?
I have measured the leakage with power off (all relays deactivated) using a
N2PK VNA and get about 95 - 100 dB isolation. Would it be a good idea to
switch the selected high pass filter relay off on transmit? This is an
in-band effect; out of band I measure 80 - 100 dB attenuation. The composite
band pass characteristics look sensible.
My default approach will be to fit a SPDT reed relay in the receiver input
driven by the PA bias control (unless there is a transmit sequence control
line that is active before the bias control and de-activated after it).

Regards

Jim G3RTD