86-Special ready with noise problem

[lord preset]

I tried some of your suggestions last night. Using a. 12at7 in v1 allows me to increase the volume before buzzing occurs but it still very much there. Swapped the 120k resistor for a 100k trim pot set to ~20k. No real change.

I don't have the caps that you recommended I substitute handy but I can certainly get them. But I wonder - are they likely to improve things given that the tube substitution and resistor change did not? I also did some chop sticking to no effect.

Another clue emerged in testing last night. In addition to the fact that the buzzing does not happen immediately upon start up, I realized that it seems to be centered on lower frequencies. With the bass turned all the way down there is very little buzzing on the clean channel. Engaging the mid switch or PAB switches both bring back lots of buzz as they bring in more lows and mids.

I spent quite a while tracing through the wiring vs the schematic and am still not seeing anything wrong. I'll repeat tonight.

Some random thoughts/questions.

1. If I have a bad cap which ones would be the most likely culprit given the symptoms?

2. Would it be worthwhile to try to bypass as much as possible between the volume control and the PI to try to isolate the problem? If so what would be the best way to do that?

3. Is it possible that there is an error in the schematic? If there is I don't know enough to spot it. Obviously Mat's amp works but AFAIK I may be the 1st one to try this exact build based on the schematic

Thanks for the help.

[Tillydog]

I think you have an 82k resistor in the PI tail, rather than 820R. Could give you those sort of symptoms.

If that's not the case, I suspect you have a wiring error, or an (ultrasonic) oscillation problem. Other things to try:

You have a level control between V1a and V1b, and also between V1b and the PI: do both of these behave the same, or can you turn one up with the other down without getting a problem?

Do you still get the problem if you pull V1?

The amount of wire snaking around makes me uneasy, particularly where there are grid wires mixed in with plate wires. If you don't get any joy, see if you can run the grid wires, especially, clear of other stuff, and ideally tight against the chassis. Make sure that plate wires aren't near any components, etc.

I couldn't find the cathode resistor for the EL84s for a long time - are you sure that yours has a high enough power rating? It will likely get pretty hot, so probably best create some air space between it and the electrolytic cap. Out of interest, what voltage are you getting across that resistor?

I haven't checked everything!

Andy

[Tillydog]

I think you have an 82k resistor in the PI tail, rather than 820R.

Hmm... maybe not - I get confused as to which orange wire I'm following.

[lord preset]

I think you have an 82k resistor in the PI tail, rather than 820R. Could give you those sort of symptoms.

If that's not the case, I suspect you have a wiring error, or an (ultrasonic) oscillation problem. Other things to try:

You have a level control between V1a and V1b, and also between V1b and the PI: do both of these behave the same, or can you turn one up with the other down without getting a problem?

Do you still get the problem if you pull V1?

The amount of wire snaking around makes me uneasy, particularly where there are grid wires mixed in with plate wires. If you don't get any joy, see if you can run the grid wires, especially, clear of other stuff, and ideally tight against the chassis. Make sure that plate wires aren't near any components, etc.

I couldn't find the cathode resistor for the EL84s for a long time - are you sure that yours has a high enough power rating? It will likely get pretty hot, so probably best create some air space between it and the electrolytic cap. Out of interest, what voltage are you getting across that resistor?

I haven't checked everything!

Andy

Thanks for the suggestions.

I just measured the 820R resistor - it measures at 813R. As for the Vol and Master, it doesn't matter which is set high or low provided the actual volume is high enough. I will some rerouting of wires as you suggest. As for the cathode resistor it's rated at 1/2w. What should it be? The voltage across the resistor is 8.2v.

Thx

[cbass]

You are pretty m ch disapatng a half watt at idle it will be more when you are playing you could get aaybwith 2 watts though a 5 watt would be better.I have a 2 el84 amp with a 10 watt cathode resistor and it gets pretty warm

[lord preset]

You are pretty m ch disapatng a half watt at idle it will be more when you are playing you could get aaybwith 2 watts though a 5 watt would be better.I have a 2 el84 amp with a 10 watt cathode resistor and it gets pretty warm

OK I'll definitely change that. I was wondering if the underrated resistor could be culprit. I took one of my 150R 1/2 watt resistors and hooked it to the DMM while cooking it with my soldering iron Not sure if that's a valid test but resistance didn't move much so I would think that its ok in the circuit for the moment.

[Tillydog]

I'd confused your 820R PI bias resistor with the 82K split load resistor on the EF86 by following the wrong orange wire!

In fact, I'd pull the EF86 for now (if you haven't already).

If either volume makes the difference, it tentatively suggests the problem is after V1 - i.e. PI or output, althought I wouldn't be too wedded to that idea just yet. Can you measure all the voltages on the PI? (i.e B+ supply at node 'C', both plates, the cathodes & the voltage at the junction of the 4 resistors)?

The EL84 cathode resistor won't be the immediate problem - it's just by-the-by. It's at its limit now, at idle (8.2V^2 / 150R = ~0.5W). You could probably get away with a 3W, but I would stick at least a 5W in there and make sure it's got some air around it.

The schematic looks OK to me.

:-/

[10thTx]

Would it be worthwhile to try to bypass as much as possible between the volume control and the PI to try to isolate the problem? If so what would be the best way to do that?

Yes, I think that would be worth while. Bypassing can definitely isolate problems at times. Forget about the OD and the EF86 for now.

The bypassing may or may not show where the issue is. I think it would be worth doing.

Have you tried different tubes for the PI and the EL84 power tubes?

I think I'd hold off on changing the bass & mid caps for now.

With respect, 10thtx

[lord preset]

Would it be worthwhile to try to bypass as much as possible between the volume control and the PI to try to isolate the problem? If so what would be the best way to do that?

Yes, I think that would be worth while. Bypassing can definitely isolate problems at times. Forget about the OD and the EF86 for now.

The bypassing may or may not show where the issue is. I think it would be worth doing.

Have you tried different tubes for the PI and the EL84 power tubes?

I think I'd hold off on changing the bass & mid caps for now.

With respect, 10thtx

Perfect. I had a chance to bypass V1b as per your 1st schematic. And the amp sounds absolutely fine. Buzz is gone. This helps a lot. Still more work to do but this narrows the possibilities down nicely. Thanks!

[10thTx]

OK, good that is some progress. Now, it could be that eliminating a gain stage reduced buzz ................... but let's pursue the idea that the problem is around V1b triode.

I have on rare occasion had a "new" silver mica cap be bad. It could also be that there is something wrong with the LNFB (local negative feedback)

So, what I would try next would be to disconnect those and safely clip from the volume pot directly to V1-7 pin. Does that eliminate the buzz and still sound right?

IF so, ........ then I would try replacing the silver mica cap and the 470k resistor and hook those back up without the LNFB hooked up.

And if that still sounds OK, then I would suspect the LNFB is wired wrong or has bad components in it?

With respect, 10thtx

[lord preset]

Had some time for more testing. The 470p cap/220k resistor and LNFB are not the culprit. Nor is it the presence of an additional gain stage.

With both V1a and b active and with V1B going directly to the PI (.047 cap direct to 270K resistor) there is no buzz. Volume is on a hair trigger with no master volume but the distortion is smooth as I go up the dial.

So I speculate that something is hosed re the Master volume/OD Switch. Upon my quadrillionth check I saw that I had a 1M pot on the master rather than 500K. I just corrected that and put in a different 70p cap but no change. Next I plan to stare at the OD/Clean switch until it runs scared.

[lord preset]

I spoke too soon. For whatever pilot error reason the experiment of running v1a and v1b at the same time is not repeatable. So my theory about the master volume OD/Switch wiring is a bust. I think

In case it might help I did a short clip of the buzz problem. Cell phone makes the buzz stand out more than it does in real life but it pretty much does sound like this.

[10thTx]

the experiment of running v1a and v1b at the same time is not repeatable

Does that mean it did work running V1a & V1b into the LTPI bypassing the DPDT and master volume? And there was no buzz

But now doesn't? And there is buzz still bypassing the DPDT and master volume?

Kind of hard to tell from the sound clip but it sounds like an oscillation problem (maybe a layout issue?), bad solder joint or wiring mistake?

Can you offer some clarity and whether there is some consistency here?

With respect, 10thtx

[lord preset]

Indeed clarity would be good. I work from home and was stealing a minutes here and there today to test and I'm sure I got sideways about what I was testing. This is humbling to say the least.

I will take a few steps back and repeat the tests and report back when I have my story straight .

[10thTx]

Don't give up on this amp. Mat comes up with GREAT ideas and I experience his sound clips as knowing what superb amp tone is.

Once the issue is resolved, I think you will have a great sounding amp.

IF the problem is intermittent, that again would leave me to consider that there is a bad solder joint somewhere in there although it could be a combination of things including oscillation.

Yes, please post again what your results are when you have some clarity and consistency in trouble shooting.

I would like to know how it sounds with the master volume and DPDT bypassed and with V1a & V1b going directly into the LTPI?

With respect, 10thtx