problem with marshall clone: presence control HELP!

[flatart]

I have a problem with a JTM45 clone I've just finished to build, I've already built 5 of them, but I never faced this problem before.
When I set presence control to 10 (max) I have a very bad sound with a very strange kind of sustain feedback.
In attachment I put a short mp3: in the first part you can hear the amp with presence control at 10(max). In second half you hear the amp with presence at 0 (min).
Has someone already faced such a problem?
Once I had a similar isse (but not the same) and I solved it putting grid stoppers on output tubes, but this amp has already grid stoppers so... HELP!

[Bob-I]

It sounds like oscillation to me. Use a chop stick to move the NFB wire coming from the speaker jack to the board which someone is playing that note. I had a similar issue with a JTM 45 a few years ago. I simply ran the NFB wire away from the grid wires to the PI. Mine sounded a tab more "wooly" than your, but similar.

If you have a scope, put it on the PI outputs and see if you see and AC like output with no input to the amp. That's how I found both of mine.

[Structo]

At the Metro forum there have been many posts such as this about squealing and in the majority of cases it was lead dress of the NFB wire.

BUT,

OK, I just listened to the clip.

I think that is what the Marshall guys call the "Cymbal Crash" syndrome.

Try putting a 220K resistor on Pin 2 of V3.
This acts as a grid stopper on the input of the phase inverter.

You may be able to get away with low as 100K.

So I would first try a 100K, if you still hear it, move up to a 150K, listen, then a 220K if you still hear it with 150K.

Un solder the wire to pin 2 and solder the resistor right to the pin.
Then solder the wire to the resistor.

Slide some heat shrink tubing onto the wire before you solder the wire to the resistor and only shrink it around the wire and resistor when you are satisfied with the result.

[stoo]

Could this be the infamous case of the "fizzies" ? ...where you put a 80-100pf across the 2 plates of the LTP?
sTEW

[jjman]

Turning up the presence essentially removes the NFB for high frequencies.
Since the presence can reveal/mask the problem I would suspect oscillation within the circuit that's within the NFB loop.

I had a similar sounding problem on my Vibro Champ after I removed the stock anti-oscillation grid cap on it's 6v6. Switching to a shielded grid wire solved the problem.

Of course it could be other things.

[flatart]

I've tried a lot things, some with sense and others (I admit) with no sense, but really this issue is driving me crazy:
- I changed presence cap
- I changed 0.1u PI caps
- removed feedback
- I removed 47pF cap between anodes of V3 (PI)
- Moreover I did some other tests on output tubes section: I removed 8K before screen resitors and so I burnt 470 ohm output screen resistor (only 2W). replaced with 5W
- I burnt my tester (left on output plate while playing)
......
Anyway I found a trick that solves the problem but to me it's unexplainable.
Look at the picture in attachment: I connected one long white cable to pin 1 of V3 (phase inverter) and I put this cable near the input side of the chassis, but the other lead of the white cable is not connected to anything. Consider that if I move this white cable towards the power section the problem comes back.
Maybe this trick makes sense for someone of you.

[jjman]

Moving your antenna wire around can cause you to inject either positive or negative "feedback". Positive could manifest as oscillation which could sound like your problem sound. Negative could mask other problem areas, reduce signal, or do nothing.

Are those OP grid wires now running parallel with other wires under the board?

I would try moving or changing the grid wires on the OP tube sockets to shielded.

[flatart]

wires from PI to output tubes are twisted. I also tried another thing: I swapped OT wires in the output tubes and obviously also wires coming from the PI inverter. I moved ground connections of 16+16 can near preamp chassis zone, I tried to move every wire inside the chassis but notjing changes: the only way to have the amp playing in a decent way is the trick I told you before: cable to pin 1 of V3 (phase inverter) and the other lead near the input side of the chassis, also after swapping wire from OT...
For today I have had enough of this fuc...ing amp and I dont want to burn it or destroy it with a hammer
goodnight everybody

[Structo]

Ah, but did you try the 220K resistor on the grid of V3?

Don't feel bad about using that white test lead.

Have you ever looked inside a Fender silverface amp?

It's a wonder those things don't squeal all the way from the factory.
For some reason, after CBS bought Fender, the wiring became very messy.
Sometimes requiring the tech to wrap pigtails around grid wires and either run the wire to ground or to B+ in an effort to quiet the beast.

Here is a late 70's Princeton Reverb.

[IMG995]http://i14.photobucket.com/albums/a322/ ... 009005.jpg[/img]

This is a 64 blackface Princeton Reverb.

[IMG:640:480]http://i4.photobucket.com/albums/y122/n ... MG1675.jpg[/img]

[MBD115]

If moving a wire thats only connected on 1 end fixes the amp then my conclusion would be lead dress. I can see 1 wire thats burned in a couple places, and at least 3 wires that have been extended. Wires that looks to be coming straight up from the board and also wires that looks to be too long. I can see a few iffy solder joints like on the ground buss at the volume pot and on the input jacks.

But I don't know. All I can do is look at the pic and take a guess.

Good luck man
Hope you get it figured out

[MBD115]

After looking at the pics Structo posted " I take back everything I said!!!!! "

[Phil_S]

I find it difficult to tell, but I think you may have a wiring error on the pots, particularly on the presence pot. These are pics of a real JTM45 I found somewhere when I built mine. Maybe they will help. In particular, I think the placement of the cap on the presence pot might be suspect, but I'm not sure of anything.

[M Fowler]

Metro JTM 45 wires theirs cap from middle (wiper) to left pin using that lead as a ground to pot as well.

[tubeswell]

wires from PI to output tubes are twisted.

I would try using shielded cable nevertheless. Even a twisted pair of signal wires can couple to stray EMR. I had a similar issue in a recent JTM45/5F6A clone which shielding resolved.

[flatart]

someone on Metro forum seems to have a similar problem and fixed it removing 1ohm bias resistor (!!!???) of output tubes.

http://forum.metroamp.com/viewtopic.php ... ng#p275076

It really makes no sense to me but I will try this too. After that I think I'll have to replace all signal wires with shielded cable.