Need help with figuring out why 10 meters doesn't work properly on a
homemade GS35-B amp. when the pi network input circuit is used.
Symptoms: without the 10 meters pi network input circuit connected,
the amp. will put out ~1400 Watts on 28500 KHz with ~65 W from the
IC746 (the IC746 internal ant. tuner used to match the amp.) The 10
meters input pi network output capacitor, C2, a120 pF mica, is
permanently mounted from the cathode to ground at the tube. When the
rest of the10 meters input network is connected and the IC746's ant.
tuner bypassed, only ~800 Watts output can be obtained before the
output drops back to less than 400 Watts (fold back?) and the SWR on
the IC746's SWR readout increases. When the output power is turned
down to just below 800W, normal conditions return until power again
is adjusted to the 800 W level.
Conditions: The input networks C1 is a mica, variable, compression
capacitor and has been adjusted for minimum SWR on the IC746. The
network was designed using a Q of 2. A toroid is used for L1 in the
input network. The amp works okay on other bands with the tuned
input. A separate pi network relay selects the individual pi network
for each band. The 10 meter input C2 is added to all the other
bands' C2 value. 5000 pF mica capacitors are used to couple the
input circuit to the cathode and across the tube's filament. The
input board's output is less than an inch away from the
cathode. Using a B&W FC15 as the filament choke. The amp input is
being fed directly from the IC746 rather than go through the ant.
transfer relay in an attempt to find the problem. Liberal
application of 0.0015 uF disc capacitors has been made. All control
lines have been bypassed with the 0.0015 pF capacitors at both ends.
Chokes have been inserted in several of the lines going from the grid
compartment to the control area.
Need some more ideas to try to isolate the problem. What's causing
the amp to not work properly with the input network?
Amps mailing list