Problem: Really Bright glowing 6CA7

[gary f]

Yesterday, one of my 6CA7 tube in my amplifier started to glow really bright while I was listening to music. It happened after maybe 1 hour of listening, the gray cathode-plate assembly became incandescent with a red-yellow light, as if the heather was too hot. The amplifier was still working OK at that time and music seemed fine. I turned off the amplifier, and waited a bit for tubes too cool down. I swapped tubes, and turned the amplifier On again, and everything was OK. I listened to some music for at least one hour, and everything was normal. The amp is a Dynaco ST-70 serie 2, from 1990.

So right now, I don't know if it is tube related or a problem with the amp.

-Anybody seen something like this?
-I read the archives, and someone mentionned maybe a bad coupling cap problem. Is this possible? And why does it seems intermittent. It is surely related with heat...

Thanks for any input.
F

 

[SY]

It could be an intermittent coupling cap, but with that amp, if you still have the original tube sockets, the problem may well lie there. I'd replace all the output tube sockets (it's a pain in the ***, but Dyna used some really cheap hardware). If the coupling caps are original, you probably want to replace them, too, just to be sure.

 

[gary f]

SY: How a socket problem could produce excessive heating of the tube and glow?

If it is a socket problem, I can change sockets and caps. I will try to reproduce the problem to be sure

Thanks
F

 

[Nordic]

The other night while playing around, I forgot to put back the heatsink on the regulator to my heater... after a while my wife called me, because the music was starting to go softer... on looking a the valve it was brightly glowing, as the heater was getting something like 10V because the regulator had a thermal shutdown of some sort.....

 

[testlab]

You have lost bias or lost a tube my friend. The other possibility is mentioned above, a bad coupling cap after the phase inverter.

 

[richwalters]

Some power tubes used in fixed bias mode go through the red hot to melt down phase at end of life. It can happen quite quickly and dramatically.

Check Pin 5 (g1) octal skt for continuity i.e contact condition,and neg volts if fixed bias is used. Check the soldering in the valve pins....(once I had a 6550B with a suspect duff heater, found to be poor soldering in tube pins)
A fuse in the output transformer ct will protect tube and tranny and if it blows will give the hint something's wrong. The fuse has no sonic effect.
Where's best place for ST2 schematic ?

richj

 

[SY]

Gary, if the socket pin corresponding to the tube's control grid loosens, corrodes, or otherwise has intermittent contact, the tube will lose bias and run away. That produces a nice light show...

While you're under the hood, check the cathode resistor (the 15.6 ohm) to see if it is cracked or changed value. Tube runaway stresses that part and I've seen more than one go blooey shortly after an over-current incident.

 

[gary f]

Schematic

Thanks everyone

The amp seems fine since yesterday. It may well be a bad contact. I had problems with these sockets. I will check resistance value and caps anyway. If the problem arise again, I will change sockets.

Here is the schematic. This is Dynaco Serie II by sound valves, redesigned in early 90's
F

EDIT: hum, the file is way too large at 2,4 MB. Email me if you need the schematic

 

[testlab]

I used to work with the engineer who re-designed that piece. He was really gratified by the review it got in The Absolute Sound.

 

[gary f]

Do you have a copy or link of the article?

This amplifier is very good. I like the sound very much, although I don't have much experience with expensive amplifiers so I can't compare with other products.
I have a custom chromed version, and the look is very very nice.

Thanks for all the information everyone!

 

[commsysman]

intermittent tube problems

My experience is that 90% of the time your problem is a bad /intermittent coupling capacitor connected to the grid, allowing bias to become lost or modified and plate current to increase dramatically. A bad connection at the grid due to a faulty socket is another possiblity, of course; same result.
The first thing I would do is replace all of the coupling capacitors with new ones rated higher in voltage than the originals. In most cases you will no longer have a problem.

 

[richwalters]

Just had a new JJ88 fail on me for no apparent reason and yes the anode glowed red hot. Swapping tubes sorted the problem out. A sharp rap on the suspect tube reverted it back to normal but only for a short period. Obviously a poor contact within tube. It was drawing 100uA of bias which is way too high.
Don't aways blame the coupling cap !

richj

 

[Johan Potgieter]

I once serviced such an amplifier, and found an ancient selenium rectifier supplying the fixed bias voltage. Do throw that out if it has not been done! Otherwise I would concur with SY (was it SY?)about the tube sockets. Especially for power tubes they are below standard.

 

[gary f]

Update:

I want to confirm that I seriously think that the problem is the sockets as SY guessed. I swapped tubes, and the problem is not tube related, nor channel related. (i had the problem with both channels at different times, and not the same tubes)

The sockets are corroded a bit, and they hold some tubes loose. Some tube don't fit tightly, but seems to be loose and move a bit, not good.

I will change to gold sockets. They are PCB monted. I guess it's will be hard to unsolder them. I have desoldering braid or something.

I'll post after it's done

Thanks
F

 

[kevinkr]

In most cases you can usually tighten the contacts with a jewelers screw driver inserted between the socket body and the pin. (There are some pin types where this doesn't work too well, these look sort of like a tuning fork as an analogy..) Make sure of course that the amplifier is disconnected from the mains and the supply capacitors are all completely discharged. - Use a meter to verify.

Replacing the sockets is never a bad idea, but unless you purchase substantially better quality sockets it is not going to make a big difference. The gold plating on current Russian and Chinese made ceramic pcb mount sockets is very thin and the contacts are otherwise no different than what is used in the cheaper socket.

Good luck!