Page 1 of 2

#1 Negative grid current WTF?

Posted: Tue Jul 03, 2018 4:11 pm
by Cressy Snr
I have negative grid current on the grid of one of my sockets for the PCL86 amp.
The fault does not move with the valve.

I get a consistent -3V on the grid of the pentode half which subtracts from the cathode bias, giving 4V instead of 7V.
I can understand positive grid current if the thing is starting to run away but I've never seen negative grid current.

Could it be oscillation linked to poor layout?
could it be leakage between the socket pins?
Grid resistor is 100K on the pentode side and there is a 4K7 grid stopper.

#2 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 4:16 pm
by Nick
Do you mean the grid is at -3v WRT ground, so the grid end of the 100k is at -3v and the ground end 0?

If so, would that not make the cathode more +ve.

#3 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 4:23 pm
by Cressy Snr
Nick wrote: Tue Jul 03, 2018 4:16 pm Do you mean the grid is at -3v WRT ground, so the grid end of the 100k is at -3v and the ground end 0?

If so, would that not make the cathode more +ve.
Yes the grid end of the 100K resistor is is at -3V WRT ground and the cathode is +4V WRT ground, so what is going on I don't know. So where the negative voltage on the grid is coming from I don't know either. There must be current flowing from somewhere.

#4 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 4:27 pm
by Cressy Snr
in fact it's now gone to neg 6.8 and the cathode has gone to 2V.

#5 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 4:31 pm
by IslandPink
Was this the left channel when you look at the amp from the front ..... :D
Cos it did seem it was clipping a bit early when I was using it for the FX120.
I just put it down to the difference in efficiency of your stereo 96dB speaker compared to my 1-off 89db/w speaker.
That alone may be enough to explain that, though.

#6 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 4:39 pm
by Nick
Well, if it doesnt follow the valve it must be the socket or parts around it. What are the voltages on the pins without the valve in the socket.

Have you checked that what you are calling 0v may in fact have become 3v wrt the rest of the amp?

#7 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 4:49 pm
by Cressy Snr
Hmm there's 6.7mV WRT to ground on the pentode control grid pin, with no valve in the socket. I would guess that points to leakage somewhere.
The anode pin of the triode section is right next to the control grid pin of the pentode section and it has 243V on it.
OV is right WRT the rest of the amp.

Yes Mark, that was the channel in question.

#8 Re: Negative grid current WTF

Posted: Tue Jul 03, 2018 4:58 pm
by Cressy Snr
It's looking like I'll need to rebuild with ceramic sockets. Luckily, the chassis Ed gave me is perfect for the job as it has ceramic B9A sockets fitted and with a bit of tarting up of the top plate, should make a decent job.

That's me sorted for something to do.

#9 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 5:04 pm
by IslandPink
Ah... interesting .
It's a minefield , innit ?
Hardest game in the world.... etc.
What are the sockets atm then ? - Bakelite ?

#10 Re: Negative grid current WTF?

Posted: Tue Jul 03, 2018 5:09 pm
by Cressy Snr
They're some sort of brown plastic.
http://www.hotroxuk.com/belton-9pin-w-c ... 9-st2.html

Never come across leakage before.
Mind you, I've never used a triode/pentode in a nine-pin miniature bottle before.
Looks like, with these things, your attention to layout and care for/choice of socket has to be spot-on.

These valves are definitely worth the trouble though.

#11 Re: Negative grid current WTF?

Posted: Wed Jul 04, 2018 4:52 pm
by Mike H
Any way you can check if it's oscillating at RF?

Negative grid charge is the kind of thing an oscillator valve can do, e.g. radio local oscillator, and which shuts down the gain so is self biasing. Might account for the clipping too. (?)

#12 Re: Negative grid current WTF?

Posted: Wed Jul 04, 2018 5:35 pm
by Cressy Snr
I'll have a probe about.

#13 Re: Negative grid current WTF?

Posted: Wed Jul 04, 2018 6:06 pm
by Cressy Snr
What a screamer! :shock:

L channel with the negative grid current is off the scale.
Image

R channel with no negative grid current is nevertheless still at it.
Image

The bloody thing is festooned with grid stoppers on every orifice too.

I don't know the frequency. I'm crap at working it out from the info on the screen.
It is obviously not happy though.

Something to watch out for when rebuilding.
Going to have to limit the bandwidth with 10s of Ks on the pentode grid, like Rozenblit does on his SOB OTL.
He used 100K on his triode strapped EL509s.

#14 Re: Negative grid current WTF?

Posted: Wed Jul 04, 2018 7:58 pm
by Cressy Snr
Cured it :)

68K grid stoppers on the pentode section and reduced the grid to ground resistor value on the triode stage.
Both changes needed to be done to completely stabilise the system at RF. Merlin Blencowe and Bruce Rozenblit provided some useful info on tracing and removing RF parasitics.

Now there's only a hint of ripple off the AC heaters, that can be seen on the anode of the output stage, with the scope set to max sensitivity.

My excuse is that it was a back-of-fag-packet design. It worked OK and sounded nice as we heard at Owston, but unknown to me it was oscillating like buggery.

Thanks for the head's up Mike.

#15 Re: Negative grid current WTF?

Posted: Wed Jul 04, 2018 8:11 pm
by pre65
Glad you have it sorted Steve.

Does it sound any better ?