I measured the voltage at pin 2 of the 6M5 as 9mV with the 544K resistor in the circuit.

Measuring the grid voltage can be fraught with inaccuracy, due to loading effect. If there is a cathode resistor measure across it, as that will tell you bias volts. It also tells you what current the tube is drawing (diagnostic).

Also note, that the Back Bias (if used) is a diagnostic, as for it to be correct, all of the current passing through it, has to be correct.

The 9mV was measured across the 544K resistor on the grid. My Fluke DMM has an input impedance of 10Mohm.
The 220R cathode resistor has 6.4V across it.

Unfortunately I know nothing about back bias. Brad has uploaded my schematic to the first post of this thread. Does that clarify it?

I've replaced the 544K (marked 0.5M) with a 470K 1W.

I just realised there's an omission on my schematic. Pin 1 (grid 2) of the 6M5 is supplied with 225VDC. I'll amend the schematic and email it to Brad. Sorry for any confusion.

The 35Ohm resistance between Centre tap and ground is back bias. However, the 6M5 clearly has a cathode resistor for self bias. So Ohms law will give you the tubes current draw; which will be a nominal 41mA 250 plate and screen volts (AWV data).

There's no 35 ohm resistor between the centre tap and ground. Is it the transformer secondary winding resistance you are referring to? I'm confused.

6.4V across the 220R cathode bias resistor gives me 29mA.

The usual bias for 6M5 is 5 volts, but with the extra HT in your radio, the 6.4v that you have seems fine to me. Nothing to worry about.

If the grid resistor is too high the grid itself may glow red, which can be seen by looking through the holes (if they exist) in the anode. Sometimes the plate can glow red a little bit too. Obviously not good for the valve. It's ok for G2 to glow blue though.

Connect your oscilloscope to the 6M5 grid and look for any sign of the 455kHz IF frequency at that point. If there's a lot of it, instability can result, as obviously the valve will attempt to amplify it along with the audio. If you look at some other radio circuits there will often be a PI filter of 2x 100pf caps and a resistor to filter out these artifacts. Also, most radios will have a 47k resistor on the grid to act as a blocker to reduce the tendency to oscillate. Look at other schematics to get the idea.

Thanks Robert.
I'll have a look at the 6M5 tonight for any glowing of the grid or plate.

My oscilloscope shows a 454.5kHz sine wave of 70mV p-p at the grid of the 6M5. I don't know if that's acceptable or not. Should I add a PI filter?

100pF from the anode of the 1st audio to ground should take care of that. It can cause cranky behaviour of the radio if it's radiated by the 6M5s anode wiring.

Thanks Ian. I assume the first audio stage is the 6BD7? There's a 'domino' type mica capacitor from the anode of the 6BD7 to ground that's marked '0001'. Whether that's 1pF, or something else entirely , I don't know. I'll try adding a 100pF ceramic and see if that attenuates the 455kHz signal.

Duplicate post

A modern ceramic cap should do a better job (lower ESR at MHz) than the old mica.

Great. Jaycar only have ceramics rated at 50V. Altronics have them up to 3kV, so it will have to wait until Monday.

Just to clarify- it's 455kHz that we want to attenuate? Or is it everything above audio frequencies?

Both! But that will mainly be the 9kHz inter-channel beat and the IF at 455kHz.

You could hang the low voltage ceramic off the other side of the coupling cap to 6M5 G1, where it will do the same thing without the applied voltage. Modern polyesters have a low enough ESR at these frequencies for this to work.

Sometimes very high voltage ceramics aren't so good at MHz frequencies.

"You could hang the low voltage ceramic off the other side of the coupling cap to 6M5 G1..."

I just did that and it dramatically reduced the 455kHz signal at the grid of the 6M5 from 70mV p-p to under 10mV p-p. Thanks for the suggestion.