Changing R2 & R3 has put the screen voltage for the 6AN7 and 6N8 back in spec. The valves are now happier with increased current draw hence the higher bias voltage which makes the 6M5 happier as well, drawing more current and producing a bit more heat. Power valves in general and 6M5's in particular do run very hot although I have no idea of what a normal temperature would be or indeed any thing to measure it. If the B+ and other voltages are OK and it is genuinely running hot the only other thing I can think of is that its drawing grid current which I believe 6M5's are also noted for, though I think that would lower the plate voltage. Someone else might have some more ideas on this.

Cheers,

Warren

I'm not sure that you have a problem. As was mentioned above, power output valves do run hot.

I would have to pull my Krielser Duplex apart to measure the temp of the 6M5 in operation, and it's fiddly to pull apart.

Perhaps someone else here has a 6M5 in a set that's easier to get at.

I've found this notice from the Kriesler Manual of Technical Instructions regarding issues with faulty batches of 6M5's and how to solve them:-

I don't know if the Fleetwood/Philips sets use pin 9 on the 6M5 socket as a tie point for other components. If it does it may well be wise to follow the above advice. I think the 6M5's giving the most trouble are the earlier "Innoval" types. Innoval valves are distinct from normal miniature 9 pin valves as they have a flatter base.

Thanks Monochrome625 for the info. I checked the radio that pin 9 has nothing connected to it.

Yes, you are right. The base of the tube is kind of flat compared with 6V4's.

That technical bulletin is interesting . I've read in a couple of places that scratching between the pins can sometimes resolve the problem but if the silver migration is internal no amount of scratching will fix that!

I did list the voltage & current for that valve, so if is is operating within specs & the plate is not red I do not see cause for panic. You may find that the rectifier is running at a similar temp. You could always measure the cathode current, but with back bias in the majority of cases, all of the cathode current passes through the back bias resistor. Therefore if the set is not running within the correct parameters the voltage across it will be wrong.

So if the B+ is 250V and there is 7V across it & the 6M5 grid leak is on the CT of the transformer, all should be well. The Pentode that forms 6M5 appears to be in several other valves,

Marc