kel84 channel imbalance

[chambers67]

This is my first post and concerns my KEL84. There is a discernable channel imbalance, with the right being noticeably quieter than the left. The valves have about 100 hours on them. Swapping the valves from left to right has no effect. The problem remains when I use my Revolver speakers or Heybrooks. My knowledge of electronic circuitry is sparse to say the least, but any suggestions, in abc form, would be most welcome. I can use a soldering iron, although my technique is not perfect.
I purchased the amp from Ebay, to be used in another room. I also have a Pure Sound A10, which performs faultlessly, making the KEL fault even more frustrating.

[Richard]

Hello and welcome
Faults like this are usually a dry/poor solder joint, a broken pcb track, or a failed electrolytic capacitor.
How big is the channel difference? Is the R channel just a bit quieter or is it almost not working?
Do you have the circuit and build sheets?
Do you have a digital multimeter?
If so check the voltages around the circuit and compare each channel as they will often pinpoint the area of fault.
Check also ohms reading from pin 7 of the ecf80 to chassis looking for 750R. This is made via the feedback wire and is sometimes broken.
If you're not sure about how to do any of the voltage or ohms checks or need the circ etc post again. There are high voltages involved and it's far better to ask than wish you had...

[chambers67]

Cheers for that. The quieter channel is such that, to achieve anything like a stereo image, I have to be substantially off centre. I have trawled through many of the posts on the forum. I have upended the amp with the bottom panel removed and tapped the pcb with a pencil -- no crackling. I have copies of the circuit/schematic, but not the build sheets. I own a basic meter, but am clueless as to how I would check voltages. I believe I can test the ohm reading on pin 7 and will do this tomorrow. Many thanks for your advice.
Kind regards

Brian

[Richard]

Hi Brian,

If you send me your email address via personal message I'll send the build instructions inc test voltages etc.

Some faults will show in the voltages and some not.

For example, does R channel distort at high vol? If so it could be an el84 not working which would show in voltages.

Otoh, if it doesn't distort and is simply quieter it may be the cathode capacitor and resistor of the ecf80 is out of circuit for reason of failure or bad solder joint. That would give lower gain but would not show in the voltages, so worth going over the joints with a hot iron and fresh solder.

[chambers67]

Many thanks, Richard, for the pdfs.
Now, this will indicate my ignorance when it comes to valve amplifier testing ---
when testing, for example, the resistance between pin 7 of the ecf80 and ground, should the amp be switched on, with either the speakers connected or some form of dummy load? Same applies when testing the valve voltages? Should the valves be in situ?
In the meantime, I have found another multimeter, a Clarkes, that I was using when working on some old Roberts and Hacker radios. Of course, transistor radios do not pose a risk of potentially lethal shocks and I could prod around pcbs willy nilly.
I am hoping for your forbearance.

[Richard]

Hi Brian,

Yes no power and unplugged for ohms testing, continuity checks, soldering etc. Be aware the main electrolytic capacitors may hold charge for a few minutes so check across them with a volt meter before starting work.

Power on for voltage testing. Attach speakers or dummy loads to the outputs. Valves in place (as they will load the supply and affect all voltages) then select an input and attach shorting plugs or a CD player to it (to block possible rf input). The most dangerous shock would be across the heart so get into the habit of just using one hand when probing. Attach a clip probe to the chassis for example then you just need one hand to probe the voltage source.

Mark the pcb for L and R as it can get confusing upside down. Check it's been wired correctly by the builder. Mark your test points first, especially valve pin numbers, and follow through the voltage test list methodically

[chambers67]

Hi Richard.
Opened up the amp and the first thing I did was to re-examine the joints on the Alps volume pot that I soldered rather hastily some months ago. Looking carefully, I could see some stray braids that were touching neighbouring tabs. I desoldered the lot, trimmed and tinned the wires and soldered all connections back. This has improved things, although it has not eliminated some vague scratchiness when the pot is turned back and forth. This was a new pot and replaced what looked like the original and one that was noisy in the extreme.
Anyway, I decided to just enjoy the amp, before testing valve voltages, for experience. I have bought an auto-ranging multimeter and tested the resistance reading between ECF80 pin 7 and earth and both sides read 740R or thereabouts.
Then, this morning, I turned the amp on to warm up and returned about a half an hour later to find that none of the valves were lit up. I could hear that the transformer/s were on, but obviously no sound from the speakers. A search of the symptoms on Google brought me to the forum and a suggestion to check the solder connections from the main transformer to the pcb. Visually, all looked ok. Then some prodding around, I found that one of the white wired leading from the tx was merely leaning against the tab and the solder connection was lost. Re-soldering the wire has brought the glow back to the valves. I cannot find a post icon that fully depicts my smugness!