After having the resonance feedback type scenario through the cab speakers at high volumes and thinking the issue may have been sound feedback and there is nothing wrong, I looked for other issues possibly affecting the output tubes just to be sure before re-assembly and found a burnt resistor R83 at the cathode of one EL34.
The heat here has been enough to show the PCB gauze under the resistor.
It may have been changed already as it was a different type from the grey coloured types in the rest of the build, though a marked value of 1 ohm as the other at R84, but it measured 200 ohm on removal.
This is pretty insignificant compared to a 3k4 ohm valve across 480V, the cathode is still effectively 0V, can be at ground in many designs anyway, but I bought two new 2.2 ohm, 3W wire wounds to replace these two 1W cathode resistors, more to make me feel better really about balancing both sides again, and nothing should get so hot it burns the PCB anyway. Not having a schematic is a real pain.
I checked the bias pot values from 0 ohms to 21k ohm for a 47k pot.
This is because it is in parallel with the 47k next to it:
I put the DMM across it and fired up to read -26V across it – a bit high for other amp values I’ve seen and read of I thought, so set it to max to get the EL34s as cold as possible to try and stop them oscillating at full volume – if there is even an issue with the amp – just to test really. This max pot value gave -36V which was close enough to the value of -37V that worked well for the Tremolux EL34s which runs at a similar HT, so tried that – it’s on the max safe side of red plating anyway!
Finding this table from WikiP
Typical class AB amplifier operation
An old Post table helped reassure me this is a good value too for EL34s – a 31mA current running at about 480V HT:
All this still didn’t help the runaway oscillation at near full volume through the cab though.
Next day I thought about checking the NFB links for possible LFO PFB issues.
I traced the link from the 8 ohm tap back to the first resistor. I reversed the feed to check the wires were the right phase, and got a positive squeal so changed them back as they were right already.
I then thought about removing the NFB cable to see if the amp ok worked without it – which it does – through an old 50W stereo speaker so I don’t damage my cab speakers, and didn’t have the resonance issue – just a lot of distortion at the OT secondary as expected if the input volume and the output Master volumes are full up. This 8″ speaker doesn’t have the bass response that a 12″ guitar speaker does and it was not pointing back at the valves which the cab was before, so I’m pretty sure I was being paranoid and the resonance issue was volume feedback, and there is nothing wrong with the amp. I’ll try again with the new Celestion on arrival to be sure before putting it all back in the cab.
I don’t know why the original Celestion was blown in the first place.
To remove speaker resonance from the issue completely I used a 200W, 16ohm wire wound load resistor I have had for ages but never used, and did not get the resonant freak out on the scope either.
It was only when experimenting with this I thought the output was too deformed after me altering the bias, as it looks like crossover distortion at the speaker now.
I returned the bias value to close to the original by guesswork, thinking a 15k value was a good start out of the available 20k. This gave a bias voltage of -28V across the pot – it was -26V before – close enough:
This showed nice symmetry at 100mV input and 5V non distorted output volume levels for channel 1 so happy with that. Only a guitar sound check will tell if it’s ok.
No crossover distortion and good symmetry between input and output.
Here is channel 1 distortion at full volume:
After a quick sound check I put it all back in the cab and have effectively, a very new condition amp, so very pleased overall after my initial disappointment and concern about fixing a PCB chip amp. There is only a few scuffs on the left side, other than that it looks almost unused.
These retail at over 300 quid, so I’m pleased. I should be able to sell it no problem on Ebay as this amp had some good reviews.
I expect it to be loud like mine.
I will give it a full work out at DBS next week and put a video up.
I just need to get the reverb tank back in. The channel switching works via the rear jack.
The clean tone is great as expected, but I still don’t like the OD – it’s transistorised to me on full gain, but there seems to be nicer soft clipped tones to investigate as you want from a valve amp.
I’ll add it to the test sound library here, when done – a new one needed now it’s back in the cab:
Once back in the cab and recorded: