runaway bias problem

[deeval]

I have a AB763 SuperReverb,and one power tube will start to redplate after amp is played about 20mins.
I replaced tubes with several new and NOS tubes that are perfect cond.
and re bias to 33ma.
I have 492v plate voltage.
and I also checked for coupling cap leaks and they measure the same.
I swapped the redplating tube with 4 other tubes and they all redplate after 20min.
Also tube redplates faster if I use an Overdrive pedal to push the amp.
So I re tensioned the tube socket pins and problem still there.
Any ideas will help.
Thanks

 

[Member 1963]

You gotta look at parts that are NOT common to both tubes. First thing i'd probably look at is the 220k resistors where the bias is split to each tube and the 1k screen resistors. And check the

 

[donnyjaguar]

Agreed. I think you are losing bias voltage as something warms up. Guessing dried up capacitor in bias circuit could aggravate this too, but likely pot or resistor in bias circuit failing as it warms up.

 

[phsyconoodler]

I'd suspect the bias supply or components as mentioned.As it follows the same socket,it can likely be one of the 220k resistors,assuming it's not a bias balance system,or one of the coupling caps.
The bias balance silverface system can be a factor.
Also check the screen resistors.At 492v on the plates you are obviously using diode rectification.That will bump the screens up pretty high.

 

[deeval]

Using tube rectifier.

 

[phsyconoodler]

The voltage is way too high with a tube rectifier.You may have a bad reading with your meter or low batteries giving you a false reading.
The picture shows a bias adjust system.
You need to measure the DC voltage on pins 5 of each power tube socket and balance them to be equal.
The voltage is straying on one tube and not the other.Change the bias supply components,preferrably the capacitor and diode.
Then change the two-220k resistors on the board and the driveline coupling caps. .1uf/630v .

 

[VaughnC]

I have a AB763 SuperReverb,and one power tube will start to redplate after amp is played about 20mins.
I replaced tubes with several new and NOS tubes that are perfect cond.
and re bias to 33ma.
I have 492v plate voltage.
and I also checked for coupling cap leaks and they measure the same.
I swapped the redplating tube with 4 other tubes and they all redplate after 20min.
Also tube redplates faster if I use an Overdrive pedal to push the amp.
So I re tensioned the tube socket pins and problem still there.
Any ideas will help.
Thanks

Sounds like to me that the non-red plating tube has a grid to cathode short which, in turn, drags down the bias voltage to the red plating tube. Did you try replacing both output tubes or testing them in a tester for shorts?

 

[phsyconoodler]

He did mention he tried multiple tubes.

 

[hasserl]

Sounds like to me that the non-red plating tube tubes has a grid to cathode short which, in turn, drags down the bias voltage to the red plating tube. Did you try replacing both output tubes or testing them in a tester for shorts?

This

He did mention he tried multiple tubes.

No, he said he replaced the redplating tube with several others and had the same condition. The problem is it's not the redplating tube that has the failure, it's the other one. Try swapping tubes, see if the problem stays with the socket or transfers with the tube. I'm betting it transfers with the tube, because the problem, as Vaughan pointed out, is with the tube that appears to be OK. I've had this happen to me on two different occasions.

At least he needs to verify that isn't the problem, the tube swap will do that. A bias voltage check at the tube grid would also help.

 

[deeval]

Maybe not clear but I put in 3 pairs of tubes and every time it was the same socket near power trans,even switch tubes everytime within the same pairs,still did same.
Thanks for all the pointers guy,s I will try replacing resistors and some caps.Also the voltage at pin 5 was 482 on bothe tube sockets.

 

[hasserl]

Well see, somethings not right there. The voltage at pin 5 should be a negative voltage, in the double digits, not hundreds.

 

[deeval]

Yes sorry I, need to slow down,I will recheck again and then I will post,
Thanks

 

[hasserl]

Yes sorry I, need to slow down,I will recheck again and then I will post,
Thanks

Check the voltage all the way thru the bias circuit, find out where you're loosing it.

 

[CDA]

The voltage is way too high with a tube rectifier.You may have a bad reading with your meter or low batteries giving you a false reading.

Not so fast.

What voltages were specified back in the sixties are all too commonly out of whack now.

An extra 5 volts going into the power transformer can equal allot.

Grid load resistor can be bad.

Grid # 1 stopper resistor can be bad. (Only there for RFI, I remove them) comonly 1k ohm, or 1k5 ohm

The diode in any vintage amp should always be replaced.

But if it were to fail, both tubes would get pissed. As well as the entire amplifier.

A leaky decoupling cap could be the cause as well because if it were to pass DC, it would pass positive voltage, not negative, and a grid biased power tube amp is looking for -voltage to the grids, with respect to ground, or cathode.

Other then that, maybe you have a ****** tube socket. But I would suspect that last.

Replace your screen grid resistors as well, 470 ohm 1 watt carbons. Put a high quality 2 watt 470 ohm part in it, metal film what have you.... Carbon if you're a believer....

Oh, make sure you actually have a bias control and not a bias balance control.

If it's a control, designed to "Split" the bias supply current capability to both tubes. aka bias balance...

Correct it and run the amp with a simple bias voltage control, this way you know both power tubes are receiving the same amount of bias voltage so they will draw the same amount of current.

 

[VaughnC]

Ok, if the problem follows the socket rather than the tube, the first thing I would try is to retension the offending tube socket's contacts with a tiny screwdriver. You could have bias voltage at the tube base but, if the tube pin isn't making good contact with the socket due to sprung (or broken) contacts, its possible that the bias voltage isn't getting to the tube...causing it to redplate.

 

[donnyjaguar]

I must've misread this one too. If the fault is only in one tube location then yes, its either a loose tube socket receiver pin or blown Cgrid bias resistor (the 220k one). In my experience, grid bias resistors are not prone to failure as they are never called upon to supply much (if any) current. Hence, I'm with VaughnC.

Sometimes the receiver within the socket is loose within the hole and pushes down when the tube is inserted making a poor connection. This is impossible to see from the outside. Sometimes you can re-seat them from the rear but if they're worn its prudent to replace the entire socket.