But.. To get the correct input power (required in broadcasting) you must measure plate current not cathode current. Properly insulated meters are safe and reliable.
The way to get around that is to return the grid and screen circuits to the cathode of the tube, and place the plate meter in the the lead from that return point to ground. Easier with triodes. Otherwise, the plate meter can go in the negative power supply lead. In that case the filter caps and bleeder resistor are connected to a common point, and the meter goes between that point and ground. You want the plate meter to read plate current only, not plate plus control grid plus screen current, or plate plus bleeder current, or filter capacitor charging current.
In my homebrew rigs with triode finals, I place the plate meters between the common point for grid leak/bias supply and PA filament, and ground. The Gates BC1-T does it the same way, with no HV on the meter.
Dave, on your problem, I would place a milliammeter between the 4 kv and the meter terminals and see if it shows the same current reading as the plate meter. If it does, that indicates a leakage current to ground somewhere around the meter. Are you sure there is no carbon path on the bakelite meter panel?
Don K4KYV
