Weird problem with MK-III amplifier

So while exploring my monoblocks for cause of humming (surprisingly apparently bad ground connection between PCB and chassis), I screwed something up with one of them.
One of them now has serious distortion on the output. I reflowed all solder joints and cleaned the tube sockets - that's all I did.

Anyway, output of first amp cascade looks good, goes to input is triode phase splitter. Here is where it gets weird. Pin 3 output looks great, whereas pin 1 output of tube is all messed up - bottom of the sine wave is distorted significantly. Swapping the tube doesn't help - output of pin 1 is weird and different with another good 6AN8 tube.

Tube socket looks good visually. I desoldered it, all 9 pins have good continuity through the socket body. I don't have any spare new tube sockets to try unfortunately.

What could be causing this problem? Thank you.

ViperZ wrote:So while exploring my monoblocks for cause of humming (surprisingly apparently bad ground connection between PCB and chassis), I screwed something up with one of them.
One of them now has serious distortion on the output. I reflowed all solder joints and cleaned the tube sockets - that's all I did.

Anyway, output of first amp cascade looks good, goes to input is triode phase splitter. Here is where it gets weird. Pin 3 output looks great, whereas pin 1 output of tube is all messed up - bottom of the sine wave is distorted significantly. Swapping the tube doesn't help - output of pin 1 is weird and different with another good 6AN8 tube.

Tube socket looks good visually. I desoldered it, all 9 pins have good continuity through the socket body. I don't have any spare new tube sockets to try unfortunately.

What could be causing this problem? Thank you.

What are the DC voltages compared to the voltages in the manual ?

peterh wrote:
What are the DC voltages compared to the voltages in the manual ?

All within spec (slightly higher due to higher line voltage). All DC voltages from the manual are spec'd around the output tubes though. Would be interesting to know expected anode and cathode voltages on triode part of the tube, but they do look normal.

ViperZ wrote:

peterh wrote:
What are the DC voltages compared to the voltages in the manual ?

All within spec (slightly higher due to higher line voltage). All DC voltages from the manual are spec'd around the output tubes though. Would be interesting to know expected anode and cathode voltages on triode part of the tube, but they do look normal.

The expected DC around the cathodyne is depending on tube used and normally not
critical.
You could expect the plate(t) to be at app 300V and the cathode (t) at app 100V,
Penthode cathode is expected to be at 1V

If plate ( triode) pin1 is still unsymmetrical it could indicate a problem with teh corresponding 6550 tube ( grid current )

I tried different output tubes, it's not them.

There is something really wrong on triode side. Looks like I have my triode completely open. My cathode voltage in both amps is 145V. Anode voltage on good amp is 300V, anode voltage on this amp is 175V. I'm losing half the voltage across 47K resistor connected to anode, so the vacuum tube must be fully open for some reason.

ViperZ wrote:I tried different output tubes, it's not them.

There is something really wrong on triode side. Looks like I have my triode completely open. My cathode voltage in both amps is 145V. Anode voltage on good amp is 300V, anode voltage on this amp is 175V. I'm losing half the voltage across 47K resistor connected to anode, so the vacuum tube must be fully open for some reason.

Check plate(triode) resistor and the .25uF coupling cap

So far all resistors and traces belled out ok. Maybe it's the 12pF cap screwing everything up? I never changed 12 and 750pF caps on these amps.
O.25 uF cap is almost new, it doesn't seem to be shorting out or open.

ViperZ wrote:So far all resistors and traces belled out ok. Maybe it's the 12pF cap screwing everything up? I never changed 12 and 750pF caps on these amps.
O.25 uF cap is almost new, it doesn't seem to be shorting out or open.

There is a reason the plate(triode) is lower. Disconnect the cap and test he voltage again
Measure the plate(triode) resistor, it should be 47k and close matched with the cathode(triod) resistor.
You can also check the eyelet 5, it's not mentioned in the manual but it should have app 440Volt

The 12pF has nothing to do with the plate(triode).

I do have correct voltage on pin 5, and 47k resistors check out fine, one is 52k in circuit, the other around 48k.

ViperZ wrote:I do have correct voltage on pin 5, and 47k resistors check out fine, one is 52k in circuit, the other around 48k.

Change them. The are supposed to be exact matches ! But first, disconnect the 0.25uF and
measure again.

Eyelet 5 has 325v on it. If I lift 0.25uF cap, that voltage goes up to 465v. No change to pin 1 problem.

Also no change if I lift or swap 12pF capacitor.

ViperZ wrote:Eyelet 5 has 325v on it. If I lift 0.25uF cap, that voltage goes up to 465v. No change to pin 1 problem.

Also no change if I lift or swap 12pF capacitor.

Did you lift the coupling 0.25uF cap ?
What cap did you lift ?
Lifting a working cap ( disconnecting) should never give DC changes. That cap might be leaking.

The 12pF cap HAS NO RELATION to the problem and besides it's orignal and good quality.
Don't mess with it!

Was the 47k resistor unequal after lifting the 0.25uF cap ? Then replace them with good matched ones please!

I replaced both 0.25 coupling caps, both 47k resistors and tube socket. No change. I'm out of ideas ....

ViperZ wrote:I replaced both 0.25 coupling caps, both 47k resistors and tube socket. No change. I'm out of ideas ....

Are the two resistors equal when you ohm then in place and power off ?
plate(t) to eyelet 5 and cathode(t) to eyelet 11 ( ground) ?

If they are equal continue with ( with power on ) measure DC voltages
- Plate(t) vs ground
- cathode(t) vs ground
-grid ( pin5) on 6550 relative ground, measure both

Better be slow and methodical that making mistakes.

Both resistors are equal to eyelets that you mentioned. -59v on each pin 5 of kt88. Triode plate 2 to ground is 143v, cathode 3 to ground 144v, anode 1 to ground 174v

Compare pin 1 to pin 3 output

ViperZ wrote:Both resistors are equal to eyelets that you mentioned. -59v on each pin 5 of kt88. Triode plate 2 to ground is 143v, cathode 3 to ground 144v, anode 1 to ground 174v

You make me confused. What are the above numbers ( 1 2 3 ) ? Are you telling us that
the 6an8 connections 1 2 and 3 ( 1=plate 2=grid 3=cathode , all referring to the triode part) ?

Where are your scope readings taken ?

peterh wrote:

ViperZ wrote:Eyelet 5 has 325v on it. If I lift 0.25uF cap, that voltage goes up to 465v. No change to pin 1 problem.

Also no change if I lift or swap 12pF capacitor.

Did you lift the coupling 0.25uF cap ?
What cap did you lift ?
Lifting a working cap ( disconnecting) should never give DC changes. That cap might be leaking.

The 12pF cap HAS NO RELATION to the problem and besides it's orignal and good quality.
Don't mess with it!

Was the 47k resistor unequal after lifting the 0.25uF cap ? Then replace them with good matched ones please!

Lifting a 0.25uF cap resulted in a huge DC change !
Exactly which cap was this ? And what is the voltage in the 6an8 triode plate ( pin 1 ) when this cap
is lifted ?

1-2-3 are 6AN8 pins as marked on MK-III manual, all refer to triode part of the tube. 2 is input/grid, 1 is anode on the top, 3 is cathode on the bottom. Scope readings were taken relative to ground - good reading is on pin 3 cathode, bad reading is on pin 1 anode.

The 0.25uF cap was the one connected on the cathode side to 6AN8 pin 3. I changed both 0.25 caps after that on the board, even though they were almost new, with no change. Lifting that cap did not change the problem on pin 1.

ViperZ wrote:1-2-3 are 6AN8 pins as marked on MK-III manual, all refer to triode part of the tube. 2 is input/grid, 1 is anode on the top, 3 is cathode on the bottom. Scope readings were taken relative to ground - good reading is on pin 3 cathode, bad reading is on pin 1 anode.

The 0.25uF cap was the one connected on the cathode side to 6AN8 pin 3. I changed both 0.25 caps after that on the board, even though they were almost new, with no change. Lifting that cap did not change the problem on pin 1.

Which cap resulted in DC changes :
"Eyelet 5 has 325v on it. If I lift 0.25uF cap, that voltage goes up to 465v. "
This is strange and should be examined. Which cap was lifted when this happened ?

As for the readings on the triode plate, the voltage indicated that something is
either loadng down the plate OR
the voltage on the other end av 47k resistor is too low

What is the voltatge on eyelet 5 on the two amps ?

The 0.25uF cap connected to pin 3. I think it is strange too, maybe it's because the loading got removed off the tube, so it just shut off. I have not measured the voltage on eyelet 5 of a good amp. On this amp, with all parts installed, it is 325V on this amp.

Peterh, I just wanted to thank you for being patient and very helpful.

I started troubleshooting outside the board and found the cause. It sucks to admit, but wires coming off stacked capacitor to eyelets 5 and 6 were swapped. They must have been swapped even before I rebuilt these amps, as I was replacing like for like everything.. still embarrassing. But the amp works great now.

I will install my new 5AR4 rectifiers later today instead of 5U4s and rebias them.

ViperZ wrote:Peterh, I just wanted to thank you for being patient and very helpful.

I started troubleshooting outside the board and found the cause. It sucks to admit, but wires coming off stacked capacitor to eyelets 5 and 6 were swapped. They must have been swapped even before I rebuilt these amps, as I was replacing like for like everything.. still embarrassing. But the amp works great now.

I will install my new 5AR4 rectifiers later today instead of 5U4s and rebias them.

Glad to hear. Now you are in business again, one learns from experience !