KEL84, one channel intermittent - where to start?

[Richard]

Originally Posted by david

Hi Richard,
Maybe a lesson learnt with the ECF80s in the Kat34s, as I had to add a Zobel betwork to mine to stop it oscillating at hi frequencies!I'm not sure of the chronology of these amps tho.....
David

Hi David, Greg, all,

April 2000 Kit88
Dec 2000 Kel84
April 2000 Kat34

Kit88 is included in the timescale as it could be unstable as launched, which was settled with increased grid stoppers, and showed some ringing on square wave. Kel84 then appeared to have been carefully tweaked to ensure it would be stable and the launch article included a very good SW picture.

Kat34 followed and uses similar feedback network and zobel on the OP but misses off the step network across the anode load. Presumably that's the one you added David and similar was used in Mullard 5-20 which used a pentode first stage. (Pentodes are able to amplify HF without the roll off that's common with a triode.)

Soon after, Kit6550 came out again with the ECF80 first valve and we had one here on the scope which burst into HF oscillation at 19W (from memory) on just one channel which may well have responded to that network like yours but we didn't try it. Pic below.

Valve base and the valve itself would be first suspects for a rustling and Shane has done them but it wouldn't hurt to try another ECF80 as they can be variable with Mullard being available and perhaps the best.

Dry joints also can be a pain to find so go over them all.

Then the loss of power, it could be that it's going unstable at HF as it warms up (or it could be a dry joint again). As suggested I'd bypass the psu caps C2 and C3 with film caps to ensure a decent HF path to help the anode RC network do its job.

Most Kel84 are fine but we have had a handful report noise and even squealing and it could be the excellent HF gain of the pentode first stage that's amplifying spurious noise from outside the system. One builder here had one that did this and I checked it over and listened to it in my system but couldn't find a problem so it could be local noise or unshielded inputs.

So these things aren't black and white, try everything, a scope helps to check oscillation but check for unshielded inputs (cables and arm wiring etc) fit the film caps, change valves, re-solder the joints etc...



blobby dry ground joint was cold on the thick wire it was hooked to



this Kit6550 broke into oscillation at half power

[Shane]

Thanks Greg, I was beginning to think the same, but Richard has given me a useful useful information to work from. I'm a bit suspicious of the input wiring (this one doesn't have the relay board, which may be a useful mod for later), and it seems that that could be relevant, so the first thing I'm going to try is to disconnect the lead from the input selector to the tape switch and run it as single input to see if that makes any difference. Then I'll try replacing the coupling caps as I can do that without any outlay, then I'll start looking at the bypass caps on V1. Should keep me going for a day or two...

[Greg.]

Hi Shane,

Good that you've got some options to explore. If you want to convert at a later stage to the relay input board, proceed with caution. Someone else here, it might have been Phil Y, found that the dimensions of the board did not correspond with the holes in the chassis. I think John Caswell suggested the way with this being to wire the existing sockets to the relay board mounted inside the chassis with short wire lengths.

Regards,

Greg

[Shane]

Yes, I've heard that as well. Seemed like a good idea even if it fits, because it would be much easier to remove should that be necessary.

I just realised I neglected to thank Richard for all his detailed info above. Sorry Richard, and thanks!

[Shane]

Having decided that the problem could be cap related, I decided to start by replacing the coupling caps as I have suitable substitutes. So, out with the PCB to remove the offending items and lo and behold, it's not the caps at all. One cracked track and two dry joints. So now I've put in PCB pins, wired in jumpers over the weak or damaged tracks and harmony is restored. Job done. It sounds amazing.

[Greg.]

Well done mate! Glad you've got the result needed. Nice one

[pre65]

Nice one Shane. Glad you got a result.

[justin brown]

Congratulations, Shane. When I attempted my first build (a phono 3s with XL caps), I too experienced teething problems. To be honest, in some ways I was surprised to get it working (a big Thank you to Matthew), but, looking at the build logically, it shouldn't have been a problem. My problem was was that I only had a tuner as my source, so although I tried my best to be patient, I did become a little too hasty as I neared the end of the build (as I was desperate to play some records). And that's where the problems lay. As I had no experience I couldn't trust my instincts... It is sometimes surprising what can be overlooked.