I conducted a test today that I haven't done before. I keyed the amplifier with no input RF from the exciter for 45 seconds into a 50 Ohm dummy load. In that time, the plates of the new, matching 572B tubes began to glow orange-ish red. For the duration of this test, the meter showed flatline relative power, 80mA of plate current, 0 mA grid current and 2100 volts HV. This makes little sense to me. I would assume that if something were wrong with the bias circuit, the plate current would be higher. And if the red plating was due to a mismatched load this test would not have yielded red plates because there was no RF being input.