Need help fixing a power amp oscillation on JCM800 clone

[mbader]

Hello all,

I've been trying to solve an oscillation problem on my 2204 build for the past 2 months with no luck.

The Problem
At low volume (below 2 or so) the amp sounds perfect; once the Master Volume passes that point there is a bad unnatural distortion/buzz when a note is hit - it is noticeably worse on lower strings.

I have already tested/replaced all components (except for the PT - but all voltages are in spec), tried about 5 different grounding schemes, chopsticked all leads, as well as redressed them to the shortest possible length, swapped all power and preamp tubes, and temporarily swapped OTs with no change.

I have also tested the preamp by jumping into the power section before and after the phase inverter to another 2204 build's power amp and everything sounds perfect.

The schematic I built by was the one from dreamtone.org - which I know for a fact is a correct working schematic (I'll attach below) and I am currently using that exact grounding scheme (I've tried a bunch of others with no change at all in the problem. Here's some pics...

[IMG:800:600]http://i16.photobucket.com/albums/b48/b ... board2.jpg[/img]

[IMG:800:600]http://i16.photobucket.com/albums/b48/b ... sis2-1.jpg[/img]

[IMG:800:600]http://i16.photobucket.com/albums/b48/b ... ots1-1.jpg[/img]

I even did a quick test pulling one of the EL34's for a second to see what happened, and the oscillation disappeared. Also, the negative feedback circuit is having some affect on it also - I can disconnect it and it improves quite a bit, however there's still a high end oscillation present

If anyone has any ideas at all I'd be extremely grateful. Thanks!

[mlp-mx6]

You mentioned replacing components... did that include tubes? (sorry, I have to start with the easy one)

[Allynmey]

If it is a new build and it has been doing this since first fire up, try reversing the OT/plate leads.Sounds like when you turn it up it's giving you positive feedback oscillation. Does it howl and squeal when you turn it up higher? If so then it's probably that. If it's just a buzz on the not it's probably not. Try it anyways. If afterwards, it squeals put it back. and check your lead dress.

[paulster]

If you pull V1 does it stop?

The schematic you are using has the grid resistor for V1 (R3) on the board, as early Marshalls had. It's often better to have it at the tube socket itself.

Also, the schematic calls for RG174 screened cable to/from this resistor.

It's difficult to see from your pictures whether you're using screened cable at this point or not. I can see you are from the wiper of the gain pot back to the other half of V1. Assuming you are then where have you connected the shields on each piece of the cable? With that layout I'd probably use the ground connection at the input jack with no screen connection at the R3 resistor end, and then the plate (pin 1) at V1 and no screen connection at the resistor end (with it heatshrinked as well since you've got plate voltage there).

That's the obvious area to look at so far. I'd also be inclined to re-route the HT supply from the standby switch away from the front panel and towards the board as it looks very close to your tone stack pots.

The problem with the NFB is likely lead dress as you've got the NFB resistor in a position which causes you to drag the signal across the amp twice and right by your tone stack pots. Try connecting the same value resistor directly to the presence pot and routing the lead direct from the impedance selector to the presence pot and set the presence to 10. That should see whether it's a lead dress issue there causing part of the problem, and also allow you to establish whether your OT primaries are reversed (you'll need to have NFB to establish this).

[mbader]

Thanks much for the replies! In response:

I have tried different power and preamp tubes

I have tried reversing the OT leads - it squeals like mad when I do that. I've gotten my lead dress as short and neat as my chassis will allow also

I've tried grounding R3 at the tube socket with no change. Also have shielded cable from the input jack through the first gain stage

The one thing that I haven't tried that was mentioned is re-routing the NFB circuit to shorten it up - I'll give that a shot today if I get a chance.

thanks again for the responses and if any more ideas come to mind, please let me know!

[paulster]

Try diming the controls and remove the preamp tubes one at a time, i.e. remove V1 and leave all the others in, then replace it and remove V2, and again with V3. It'll help identify which stage is causing the problem and therefore where to focus on first.

[paddy]

Hi there

I had a similar problem with an 800 I finished last week.I solved the
problem using the following method.
Clip a jumper lead to the ground bolt that the main filter caps are attached
to,then place the other end of the lead at the other ground points throughout the circuit,one by one.If you get any change in the noise,then
unsolder the offending ground point and rewire to the main ground.
This worked great for me,I still haven't worked it out really,as I had metered all grounds with a low ohms meter and they were all perfect
and lead dress was good.
Also have you tried moving wires around with a chopstick?
I remember reading somewhere,that all wires should be dressed close
to the chassis apart from the grid wires which should be pulled up
from the chassis.

good luck
paddy

[tictac]

Looks like you need to change the way your input to pin2 of V1 is wired

Try a single shielded cable from the input jack to pin 2.

Mount a grid stopper (33K) resistor right on pin 2 with zero lead length.

The other end of the resistor connects to the shielded cable, I like to put heat-shrink over the end for extra strength

[mbader]

i had to let this project sit a while to keep myself from getting too aggravated, but here's some more info finally:

It seems that the problem is somehow tied to the negative feedback circuit - I can either disconnect the NFB wire or cut the presence pot to 10 and the oscillation goes away. I'm guessing that means that I'm somehow getting positive feedback when the presence pot is grounding the signal.

the question is - how can i fix that? The OT primaries are definitely connected correctly, because if I swap them it squeals as soon as i flip the standby switch. The amp actually plays pretty well without the NFB circuit, but it really lacks the smoothness I like without it. Would changing C17 to a larger value help control the oscillation? I don't really understand the background theory on NFB so I'm not completely sure what the next step is.

Any ideas? Thanks much

Mike

[paulster]

It looks like you've grounded one of the secondary taps from the OT at the impedance switch. If this is the case then you're having to use your OT primaries round the wrong way and hence you're getting positive rather than negative feedback.

Your OT secondary should be grounded from the speaker jack.

Try removing the ground connection from the impedance selector, connecting the shield of the speaker jack to ground and then flipping the OT primaries.

Paul

[mbader]

the OT is actually grounding at the speaker jack - the white wire you're referring to is the NFB wire

[paulster]

I meant the white wire which looks from the second photo as if it's connected between the impedance selector and the ground point on the cap can to the right. I guess this passes under the impedance selector then rather than being connected to it.

Have you got a wiring diagram with your OT? The PDF looks as though either grey or green is the common (I'm guessing grey but they use two shades of green) and black is 16ohm, in which case you'll still have the OT connected wrong from what I can see, as you appear to be using black for your common.

A confirmed wiring diagram for your OT and a closer photo of this area should be enough to get this thing up and running properly.

[mbader]

yeah, that's just the camera angle making it look like the NFB wire is grounded.

The OT secondary ground is the black wire going to the left side of the speaker jack - the grey is the 16 ohm tap, green is 8 ohm, and yellow is 4 ohm