JTM45 Build

[guitarmike2107]

I don't really have any more ideas either, been a hard thread to follow what has been tried and what has not, did you bypass V2 altogether as suggested above, that would at least tell you if the problem if defo there or not.

Also you have rebuilt the board, so unless there is a faulty component on the board, though unlikley now, then the problem is off the board, have you tried reflowing all your solder joints, did you have a duffer pre amp valve in the socket when you soldered the connections, always a good idea....

could just be a bad joint that is fine at lower levels but can't handle the higher signals... complete guessing though

[Littlewyan]

Yes sorry the thread is rather messy. Basically I removed the power valves and V1, pushed the signal into the mixer resistors and still get the issue. Replaced all of the capacitors between V1 and the power valves (except for the V1 coupling caps but I ruled those out by grounding the anode on one side whilst using the other) and all of the resistors measure up ok.

I didn't have a valve in a socket when I soldered the connections but all of the wires to the sockets are new so all of the connections are new. I did accidentally bend the grid pins when trying to lift the grid wires into the air, wonder if that did anything. I haven't yet tried bypassing V2.

Everything in that area has been replaced except for resistors and pots which all measure up fine. All thats left that is viable are the sockets.

[guitarmike2107]

ok, bypass V2 and the tone stack and see what happens.

you could also bypass V2 and attach the tone stack to the V1 Plates...thus eliminating V2 but still testing the tone stack

[Littlewyan]

Ok I'll do that at some point this weekend. I'll connect the mixer resistors to the tonestack, bypass the bright channel mixer with a jumper lead and up the input signal to 4/5V. Basically until I see the tonestack being pushed hard enough or until I possibly recreate the issue. Hopefully I wont recreate the issue as that will narrow it down a LOT.

[Littlewyan]

Bypassed V2, same issue. Going to try bypassing the tonestack.

Edit: Ok bypassing the tonestack. If I push a 1Khz signal into it its fine, if I push a 3Khz signal into it and turn the volume up to 5 the issue starts and the anodes go 60V apart. Then as I turn the volume up even more it sorts itself out. This is with the tonestack completely disconnected and with the PI input cap connected to the coupling cap for V1B.

If I leave the tonestack connected but connect the V1B coupling cap direct to the PI input cap I can still stop the issue by turning the middle up.

So its not V2.

Ok just tried this. I connected the bright channel volume pot directly to the PI input cap, nothing else is in the circuit now and the issue still occurs. However it only occurs with the volume on 1, I suspect its because the bright cap is allowing the high frequencies through and then as you turn up the volume the lower frequencies get larger and even out the PI. The issue is worse with a 3Khz signal than it is a 1Khz signal.

Also the issue only occurs when I use the normal channel if I put a bright cap on the volume pot. I am certain the issue is in the PI as I have driven the mixer resistors with the oscillator before and that has recreated the issue.

[Littlewyan]

I've ordered a new valve socket to see if that fixes it. Just at thought though, is there anyway it could be the GZ34?

Another thing I need to add. When I connected the mixer resistors to the PI Input Cap BUT left the tonestack connected, I could still fix the issue by turning the middle control up.

[guitarmike2107]

ok so at least you know it is not the tone stack or cathode follower, and not the power valves or bias circuit.

I don't think the rectifier would have anything to do with this

You did have a tube failure in V3, so that may point to something with the socket. for all the issues you have had I would replace everything to do with that socket/valve that you haven't already replaced.

What is the biggest cap you have tried between the plate to kill HF? if this is oscillation I wonder if it would be worthwhile trying a large cap there just to see ...500p to 1nf even

[Littlewyan]

I tried 250pf in parallel with the 47pf but it didnt make a difference. Socket is on order so that will get replaced. I did find some dirt between pins 6 and 7 so I wonder if that caused an arc at some point. Whatever it is its going to be something odd.

Edit: Another thing I'm going to try is lifting both of the output coupling caps to make sure its nothing to do with the grid wires going to the output valve sockets.

[Littlewyan]

Update.

Lifted both PI Output Coupling caps, same issue. Disconnected NFB and grounded PI at 10K Tail Resistor, same issue. Removed V1, lifted both V1 coupling caps and fed signal into mixer resistors, same issue. So the PI is definitely at fault. Both the Caps in the PI input circuit have been replaced and all resistors measure up fine. I've also tried replacing the PI filter cap.

At this point the only part left is the valve socket. But one more thought, the mains filter wouldn't affect it would it? As surely I'd have serious issues in the output stage if that was making noise.

[ValveStorm]

This is starting to remind me of a thread on the Tube Guitar Amp Builders facebook page where an experienced tech was chasing weirdness... it ended up being the choke! Look for "just don't play an F#".

[Littlewyan]

I joined the group and read through that. Got a few more things to try now.

1. Remove the PI Coupling Caps altogether. All I've done in the past is lift them and every time I've left them connected to the PI Anodes.
2. Test the PI again by removing V1, V2, disconnecting the tonestack and disconnecting the power for V1 and V2 so the PI is completely separated from them.
3. Replace choke with resistor (I'm sure I've tried this but will give it another go).
4. Replace mains and screens filter caps (again I'm sure I've tried this).

Thanks for posting ValveStorm, I see you've only made 4 posts, I must be special!

[pdf64]

Sorry, I've got too bogged down in all the thread posts to find where the current issue is detailed.
Maybe start a new thread as this has become rather unwieldy / daunting?

[Littlewyan]

Yes I was thinking that. The amp is with a local tech now anyway. I had enough of pulling my hair out over it so I thought I'd just pay £35 to get it sorted. Although so far the guy hasn't found an issue, he said it tests fine but he didn't try turning just the volume and treble up. He was testing it with all controls on 0 or 10.

Plus he pointed out that I should of used a 30mV input signal to test and not a 250mV as I'll be saturating the amp. Which is true, however later Marshalls are the same as these amps just with an additional gain stage on the front end so I didn't see an issue with it.

[pdf64]

I think it's normal for the 'old' 5F6A /JTM45 / JTM50 etc tone stack (250pF treble cap and 56k slope resistor) to go pretty weird with vol, treb and presence up high, mid and bass low.
It's a horrid hollow tone, and on the scope the wave seems to shift phase and change shape massively.
Change to the 500pF and 33k slope and it seems to stop doing it.
I like lots of 'chime' and because of the issue, can't seem get a straight JTM45 circuit to a tone setting I like with my LP.
The answer for me is to increase the bright cap value to 220 - 470pF, and avoid turning the treb much past halfway (linear taper).

[Littlewyan]

Funny you should say that. I tried the 33k/500pf stack and the issue got worse!