Help diagnosing Tube Amp PLEASE

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Thanks again for all your replies :)

So I checked the 6V6s screen resistors and one was gone (470R). I replaced it with a new one I had but it's a 5w. Maybe that's a bit overkill and 1 or 2w is enough? Also if the screen resistor is supposed to protect the tube maybe its rating shouldn't be more than the right figure?

In any case, both 6v6 sockets now read the same (still WITHOUT tubes).



V1 ECC83:
1: 422v
2: 0v
3: 0v
4: 41v
5: 41v
6: 423v
7: 0v
8: 0v
9: 41v

V2 ECC83:
1: 428v
2: 0v
3: 0v
4: 41v
5: 41v
6: 427v
7: 423v
8: 0v
9: 41v

V3 ECC83:
1: 422v
2: 0v
3: 0v
4: 41v
5: 41v
6: 423v
7: 0v
8: 0v
9: 41v


V4 6V6:
1: 0v
2: 40v
3: 440v
4: 440v
5: -40v
6: 0v
7: 40v
8: 0v

V5 6V6:
1: 0v
2: 40v
3: 440v
4: 438v
5: -40v
6: 0v
7: 40v
8: 0v

There's still that high voltage on V2p7 which I don't know if that's normal...

Filaments are defo elevated, otherwise there’d be 0VDC on them.

Excuse my ignorance, I'm a hobbyist really, but what does "filaments are elevated" mean? Does it mean that the 0v reference is like "higher"?

Trace the wire from Pin 4 of the 6V6’s back to the board. You should end up at a pair of 220k resistors (Red-Red-Yellow, or red-red-black-orange bands) that are tied together at the opposite end. Measure the voltage at that tied point and see what you get. It should be negative and somewhere in the -25 to -35 VDC range ish (precise value is unimportant right now) but it should be present.

I can measure -42v there:

IMG_20240323_165833.jpg


Would that be ok? Is it safe to put the tubes in?


This is the amp:

IMG_20240323_112032.jpg


What seems to be the PSU:

IMG_20240323_112043.jpg


I believe this is the preamp board:

IMG_20240323_112048.jpg


Power Tubes Board. The blown screen resistor on the right:

IMG_20240323_112058.jpg

IMG_20240323_112104.jpg


Thanks a lot for all your advices :)
 
The ab763 is not cathode biased. It’s a fixed bias amp. Which means even with no tubes you should be seeing some kind of DC voltage on Pin 4 of both 6v6’s and it should be identical.
6V6 grid is on pin 5; OP measures -40V there, on both sides.
However, voltage on pin 5 of V4 is not correct, and it's the screen grid.
 
It is hard to determine the fault with incomplete information.
Did OP find the schematic?
Testing w/o tubes, why? Loaded voltages tell more. A series power resistor could have failed to a higher resistance.
The black resistor on output PCB looks overheated.
Besides power ratings metal film resistors have voltage ratings, may work for years, then gradually fail.
Other possible causes:
Bad solder joints.
Failed potentiometers.
Shorted caps, measure potential across.
Bad tube sockets.
Etc.
Find out WHAT works, inject a signal at the input and trace it thru the chain, where does it go missing?
Measure AC heater voltages, do the tubes even light up?
I see solder joint on boards without nice fillets. For thru hole parts I'd want to see this on both sides of PCB.
Get one of these kill-a-watt wall plug power meters, cheap. Can give clues on major malfunctions.

Understand the circuit and what every part does.
 
6V6 grid is on pin 5; OP measures -40V there, on both sides.
However, voltage on pin 5 of V4 is not correct, and it's the screen grid.
This is what I get for responding to a post right out of bed. You are correct.

But with the updated chart the OP sent after replacing screen grid the voltages are now correct. V2B pin 7 is way high tho. None of the preamp tube grids should have 400+ volts on them.
 
Last edited:
Excuse my ignorance, I'm a hobbyist really, but what does "filaments are elevated" mean? Does it mean that the 0v reference is like "higher"?
In almost all guitar amps, the tube heaters receive AC (alternating current) power while the rest of the amp receives DC power.

DC heaters are implemented in some very sensitive circuits (like microphones and mic preamps), because it’s lower-noise. It’s also more expensive to implement as a strategy, and more complex (with more potential failure points)

For these reasons, it’s not a very good trade-off in a guitar amp, as the levels of gain typically just aren’t high enough to make filament-induced hum an issue. There is little practical benefit, which doesn’t justify the extra engineering complexity in most cases.

As a result, if your meter is set to read DC volts, you would usually expect to read nominally 0 VDC on filament pins, as a DC meter won’t read AC voltage.

However, there is a sort of “middle path”—elevated heaters.

With elevated heaters, a DC voltage is applied, and an AC voltage “floats on top of” the DC

The tube filaments only care about the difference across them—if the entire filament is elevated by 40VDC or so, it will still only “see” the 6.3VAC across it.

Since voltage is potential energy, you can think of a stone on top of a hill. As long as it’s held in place and not allowed to roll down, that potential remains just that—potential—even if the stone is attached to an oscillating bungee cord that bounces it up and down atop the hill.

Perhaps a tortured analogy, but…

In any case, why would we do this?

Well, the normal method by which filament hum couples to audio relates to the heater’s proximity to the tube cathode. For reasons beyond the already TL;DR scope of this post, in most tubes a high potential difference between heater and cathode will reduce the amount of hum coupled

It won’t be as quiet as DC heaters, but it’ll be better—and it’s way way less complicated (since the elevation voltage can be derived from elsewhere in the amp, often the bias supply)

That strategy seems to have been implemented here, based on your measurements

Hope that helps
 
Thanks again for all your replies :)

So I checked the 6V6s screen resistors and one was gone (470R). I replaced it with a new one I had but it's a 5w. Maybe that's a bit overkill and 1 or 2w is enough? Also if the screen resistor is supposed to protect the tube maybe its rating shouldn't be more than the right figure?

In any case, both 6v6 sockets now read the same (still WITHOUT tubes).

Agree we need to see loaded voltages. I'm still not convinced those output tubes aren't toast, esp that one on the bad screen grid side. A failed tube is one of the prime suspects for dead screen grid resistors. The V2b grid reading is very high however, and if correct AND it's doing the same job as V2 in an original Deluxe, that's not good and could mean you have a bad cap in the tone stack bleeding DC from the plate of the previous stage.

Also no, the screen grid is designed to protect the tube screen from pulling too much current- it is NOT imo a good design practice to make them small enough rated to act as "fuses" in the event of a fault or of the grid drawing too much current a few too many times. There are better ways to execute that than using a power resistor that is not meant to be fusible.

In old fender circuits they call for 1W resistors which imo is way under rated, esp for the carbon comp designs in use at the time. I've seen more than my share of burned-to-cracking grid resistors in old fenders. 5W may be overkill, but it's a common value to use there, and is quite fine.
 
It also looks like it has been replaced. Solder joints looks better. The resistor next to the "input" cable has a sketchy looking solder joint.
Signal tracing with an o'scope and a signal generator helps. Rigol scopes are good and low cost
 
Hello again guys,
Sorry for not replying yesterday but I was gigging.
Hope that helps

Thanks a lot for your explanation soapfoot!
I had never came across this. I build 18w amps based on the Bluesbreaker and I've always used transformers with a 6.3v secondary. Twisting the wires properly and taking care of the lead dress never gave me any hum...


Agree we need to see loaded voltages. I'm still not convinced those output tubes aren't toast, esp that one on the bad screen grid side.

I got a new set of tubes for the amp. I did the measurements without the tubes before plugging them in just to be sure the amp wouldn't toast the new ones.


The V2b grid reading is very high however, and if correct AND it's doing the same job as V2 in an original Deluxe, that's not good and could mean you have a bad cap in the tone stack bleeding DC from the plate of the previous stage.

I'll check the caps in the tone stack, which is going to be a pain since I'll have to lift one of their legs... ;)


5W may be overkill, but it's a common value to use there, and is quite fine.

Ok, I'll leave it in then and I'll change RG2 to another 5w :)


RG1 looks burnt out.

It's the resistor I have replaced with a 5w one moamps :)

So does rg2

Actually, RG2 reads fine, Pucho :)


Thanks again to all for your input. I'll plug the new tubes in and report back!
Cheers
Sono
 
Ok, so WITH the tubes and new 5w screen resistors these are the readings:


V1 ECC83

1: 175
2: 0
3: 1,6
4: 39
5: 39
6: 227
7: 0
8: 2
9: 39
V2 ECC83

1: 170
2: 0
3: 1
4: 39
5: 39
6: 313
7: 170
8: 170
9: 39
V3 ECC83

1: 226
2: 43
3: 67
4: 39
5: 39
6: 227
7: 37
8: 68
9: 39
V4 6V6GT

1: 0
2: 40
3: 422
4: 422
5: -40
6: 0
7: 39
8: 0
V5 6V6GT

1: 0
2: 39
3: 422
4: 422
5: -40
6: 0
7: 39
8: 0


That seems more like it, right?
Cheers
Sono
 
It seems to be connected as a voltage-follower, with V2b's grid connected to V2a's plate. With the tube missing, they are both at B+.
Or, there is a leaky cap.
If it’s connected that way then it’s a big departure from an ab763 deluxe. That configuration doesn’t occur in a deluxe reverb and definitely not on V2 which should be the first two gain stages of the Vibe channel.
 
If it’s connected that way then it’s a big departure from an ab763 deluxe. That configuration doesn’t occur in a deluxe reverb and definitely not on V2 which should be the first two gain stages of the Vibe channel.
Agreed. That's why I mentioned the possiblility of a leaky capacitor. In order to answer that, the OP must trace the circuit. Without it, we're lost withoutt a map and no GPS.
 
"Actually, RG2 reads fine, Pucho"
I would also replace this resistor with a 5W
I would also resolder the pin connections to T2 (the one you changed the Rg on) as it looks like those joints are questionable.
 
Back
Top