Large Channel Imbalence KEL84

[mudvin]

Hi All,

My newly completed KEL 84 has an issue where the Left Channel is much lower in volume than the right. The issue is a little strange in that the extent of the imbalance varies a lot and can be changed by quickly moving the potentiometer. Depending on the extent the issue can present as one channel being almost completely mute, or just a slight imbalance of the stereo image.

The issue remains in the same channel if I switch the valves, and if I switch the inputs (So I'm pretty sure that the problem is not in the valves or input wiring) and Voltages all check out as per the manual. Other than the channel imbalance the amp is silent, with no popping or hiss. I have attempted to poke all the joints with a chopstick and nothing presents as a bad solder joint.

So with that in mind I tried the nuclear solution, removing the main PCB and re flowing all joints (and by extension all the cables leading to it, however there was no change in the issue and this was pretty much a complete rebuild of the amp/)

So at this stage the only thing I have left is that the the pot may be faulty, and I will replace it as soon as I can. In the meantime though, I was wondering if there was anything else I can try. The was a previous post with a similar issue suggesting to check negative feedback and ground connections. However I believe I have addressed those with the rebuild, anything I've missed ?

Thanks in advance, Rob

[A Stuart]

I am not au fait with this circuit, but your comment that the gain of a channel appears related to the rate of change of position of the potentiometer must lead to the centre of the problem.
Check if there is a DC voltage across the potentiometer.
As I said, not familiar with this particular circuit, but in general I think it would be true to say there should be no DC across a pot.

[BOONDI]

Try a meter set to resistance on both channels of the pot.
Attach leads to input of pot and wiper of pot.
Then turn the pot from side to side.
This should give a smooth rise or fall from 0 to 50k.
Then do the same again on the other channel.
Again you should see a smooth rise or fall between 0 and 50k

You will need some clip leads or 3 hands to do this.

If you get anything other than a smooth rise or fall of resistance i would suspect the pot.
If you swap the channel wiring on the pot and the fault follows then it is the pot.

If the pot tests good, try swapping the speaker leads over. If the fault is still in the left channel then you have a faulty speaker. If it moves to the right then it' s in the amp. Wouldn't be the first time ive seen this either.

Let us know how it is going just in case you need further assistance.....


Alternatively, you may have the left. Channel wire directly across the pot instead of wired to the wiper. This would give same volume on left channel no matter what position yoh have the pot in, so check this carefully just in case.... Also, your pot may be 100k. It would still be the same as above.....

[A Stuart]

My thinking was that the only way the amp can "know" whether the pot is being moved is if there is a voltage across the pot and moving the wiper alters the standing bias on a grid.
It does depend exactly what Rob means by "moved quickly". If he means eg tapping the knob, then what I have said does not apply.

Otherwise, only do what Bob says with the amp off and discharged, unless you have already ruled out the presence of any voltage across the pot. Depending on the design, putting a voltage into an ohmmeter could wreck it.

[BOONDI]

Yes, power down first justin case.

That said, i dont see where there would be dc at the pot. The wiper goes direct to the grid via a 3k3 resistor with 1M to ground. The input to the pot is direct from the source inputs. Sometimes the wiper is not making good contact with the track, so turning the pot can push the wiper onto it causing it to work properly. Also, sometimes the track is not fully coated properly and again moving the pot will cause differing output. Further, if the grid is incorrectly wired to the other end of the pot then limited volume and scratchy sound when the pot is rotated. Dont laugh but i have done this myself way back in the day.

If the pot passes the tests, then i suspect either the wiring or the two resistors. I cant see how it would be anything else as the op has said swapping and etc had no discernible effect to which side faulty was on.

Also try swapping speaker wire around. If fault now on right hand then its the amp. If still on left then u have a dodgy speaker.....

Try powering up the amp and wiggling the ecf80 gently in its socket. Any chages to the sound could mean a loose pin in the valve socket.

[A Stuart]

It would certainly need something very strange to get DC to it, as you say there are no potentially leaky interstage capacitors involved.
I think that we need more detail on this comment about moving the pot quickly altering the amp's characteristics.
A simple bad connection would surely be very noisy, but I agree with you that isolating and confirming the pot's behaviour would be good.
Alastair

[BOONDI]

Originally Posted by A Stuart

It would certainly need something very strange to get DC to it, as you say there are no potentially leaky interstage capacitors involved.
I think that we need more detail on this comment about moving the pot quickly altering the amp's characteristics.
A simple bad connection would surely be very noisy, but I agree with you that isolating and confirming the pot's behaviour would be good.
Alastair

Esp as the op suspects the pot as the culprit.
If it's not the pot then maybe poking the 3k3 and 1M with the aforementioned chopstick, just to see if they are faulty. Ive seen some where they look fine but have a hairline crack or the endcap is not soldered correctly. I cant recall the way the board faces in 84 so this may not be possible without taking it out the case.

Derek.

[mudvin]

Hi All, and thank you for your suggestions.

I have been out poking around, and have narrowed down the following.

Firstly the Pot position no longer influences the sound (after I re-flowed the entire amp (possibly this was a separate build issue that I fixed accidentally). I also measured the performance of the pot using the meter and both channels seem to track well, Thus I don't think the issue is with the pot either.

DC between Input and output on the pot measured a peak of 0.62mV on the right (working) channel and 0.40mV on the Left (Broken) channel. That seems pretty low to me ?

During the process I accidentally broke the L/IP cable working in the tight space, so that gave me an opportunity to use a jumper cable to the pot. Jumpering to the Pot Input resulted in no change (So the original cable was fine. Jumpering to the Pot output (bypassing the pot on that channel I assume ?) raised the volume, however it was still very quiet, and the other speaker outperformed it on 1/4 vol

I will go and check those resistors that were suggested, and thanks to everyone that has helped.

Rob

[mudvin]

So if this is potentially a wiring issue, any ideas where I should focus given that the bad side makes clean sound but at a reduced volume.

I will happily swap out components, however this sort of board does not seem very resilient to repeated work.

My other thoughts are that it is a component that that is soldered properly on one side but the through hole plating to the other side is bad (I did try to do both sides where possible). I think that is this case the output would have noise rather than just being quiet ?

Anyway thanks again - Rob

[BOONDI]

Hi Rob.
I am assuming you have swapped the L and R channels of your source?I

Have you swapped speaker leads over too?

Do you have a spare pot that you can try?

Plate through holes mean that only one side has to be soldered properly, so if as you say you have good joints on one side then you should be ok.

Check all component values very carefully just in case.

Have you tried the other suggestions ???

Derek