Divining intermittant cathode-heater leakage?

[N_Cook]

Any tips for doing so?
My tester goes only to 25Megohm , consistent leakage usually shows itself in
that range , would going higher by some modification elicit intermittants?
Any other tips? heating the envelope with hot air gun? holding valve
horizontally while testing ? testing at higher plate/anode voltage ?

 

[N_Cook]

Any tips for doing so?

For intermittent tests, you'd like to tap the tube (use a wooden
stick). My preference would be to test with cathode
warm, using a neon lamp and limit resistor to a plate
supply (200V or so), returned to the heater.

An intermittent will make the lamp flash, but it isn't
clear what, if any, motion you'd see on a meter needle.
Leakage (like a gassy tube) might cause some flicker, too.

reply

you wouldn't happen to know what effective leakage resistance in say ECC83
starts to make its presence felt audibly on the output of an amp? I find
neons rather temperamental and not reliable in DC situations. There is a
mains neon on my bench ps that rarely lights unless the room light is very
dim or off , very odd. If the ps packs up then I will sort that and the neon
while in there

 

[mike]

Any tips for doing so?
My tester goes only to 25Megohm , consistent leakage usually shows itself in
that range , would going higher by some modification elicit intermittants?
Any other tips? heating the envelope with hot air gun? holding valve
horizontally while testing ? testing at higher plate/anode voltage ?

Can't help you with a TEST, but I've had success blowing out shorts,
intermittent or otherwise, with a stun gun.

At >25M, you're gonna need to look for issues with the socket and the
circuitry hooked to it. Doesn't take a very big spider web or chunk
of damp dust to make
25M.

I once had a spider build a tidy web across a CDROM read lens...but
that's another story.

YMMV