JTM50 Blocking Distortion?

[Littlewyan]

Ok I unfortunately actually had 2 issues in the amp. I found recently that I could sometimes get it to motorboat temporarily on turn on. I assumed this was linked to my original issue but actually it turned out the 12AX7 in V3 had pin 3 strapped to pin 4 internally but once the voltages have finished rising it would act normal -_-. Anyway that is now sorted but means a lot of my troubleshooting went to waste as I assumed that issue was linked and was trying different things to resolve that issue. So I may have possibly fixed the other problem but didn't realise -_-.

Anyway what I did find when playing last night was if I have the middle and bass up then you don't hear the issue as well. Its like you can sort of hear the treble still having the problem in the background but hear the middle and bass signals clear as day. I tried using a jumper cable to go from the treble cap to the pot to see if that would sort it but it didn't. I'm going to get a scope on it to see if I can determine where the issue occurs. I did try grid stoppers quickly to try and determine which stage this issue occurs at but the 33K stopper didn't make a difference at the PI, V2a or V1B. Completely stumped.

Edit: It does seem by the way that the treble doesn't work at all. As if I turn the treble down to 0 and just have the middle to 10 its fine. If I raise the treble then the tone doesn't get anymore trebley as its sort of blocked. It also only starts to occur when the amp starts to really break up, when clean it sounds ok or it isn't anywhere near as bad.

[Littlewyan]

Would moving the NFB wire to the other side of the V4 socket help instead of running it with the OT Primaries and secondaries then across to the 27K NFB resistor?

I've also scoped the amp by the way, I'll upload photos in a bit. There is definitely an issue that appears to start at the PI.

Edit: Just to give you an update. Whilst testing the amp with a signal generator and measuring voltages I found that the PI voltages go a bit......weird. When I turn the treble up to 10 the 82K Anode swings up to 300V from 200V and the 100K Anode swings down to 160V from 200V. I removed the power valves and it did the same thing. So with the power valves out I measured the other end of the coupling caps and found the bias voltage on both sides will go a few volts positive and then back to normal as I turn the treble up. But the voltage on the bias circuit stays constant. Would a leaky coupling cap do this? I already tried replacing the 82K Coupling cap but not the 100K coupling cap.

[Littlewyan]

Another update - I kept measuring a very small voltage in the tonestack that I could vary with the controls that would go away once I turned the amp off. So I disconnected the treble cap from it's turret, disconnected the wire for that turret from the pot and then measured the turret on it's own and measured 60mV. I also found that when I touched the turret I would hear a noise through the speaker as if I made an electrical connection. I did the same thing with the turret for the bass control capacitor in the tonestack and this also made a noise but it wasn't as loud.

I also measured the same voltage on the empty turret that I left near the V1 cathode (in case I ever went split cathode) and this turret would also make a noise through the speaker if I had the volume up. Bear in mind these turrets were NOT connected to anything when I did this!

Anyway after speaking with the supplier I bought the board from it turns out I probably overheated it as the board can only take 120 degrees and I had my iron set to 400 degrees and made the board bubble when I was soldering. So I've removed the board, still couldn't find anything a miss so the supplier has kindly sent me a new board. Its an extremely slim chance that the board has become conductive as I couldn't find any other evidence of this but I'm replacing it just in case. This time however its made of a material that I used in my other amps which I know can handle my soldering skills . Will let you know once the amp is rebuilt.

[Littlewyan]

Update.

Swapped the board over and rewired the amp, same problem although it has improved. The amp now plays fine as long as the treble doesn't go above 5 or if it does I need to make sure the middle is at 5 or more. Also the small voltage I was reading on the treble cap has disappeared.

I redid the grounding today so the pre amp and PI/tonestack grounds are seperate and also connected the speaker ground to the PI ground but still same issue. I also tried connecting a jack socket to some shielded wire, sticking a 100k grid stopper on it and connecting that directly to the grid on V1b but it didn't help. Its got to be parasitic oscillation but question is where :/.

[neikeel]

Just a thought!

I had almost identical problem with open circuit on one half of the preamp filter can on a 70s superlead, I had just done a recap and assumed all caps good but one was DOA.

Might be worth checking?

[Littlewyan]

Swapped the cap over already and it didn't make a difference. Still doesn't explain why turning the middle control up stops the issue as thats obviously letting more signal through! Turning the bass up doesn't make any difference. Also turning up the presence seems to make the issue more prominent.

After listening to the issue tonight its very much like blocking distortion in that you'll hit an open E and the volume will go down a bit and then after a few seconds it'll come back up again. Plus its like its only blocking certain frequencies which then moves us onto parasitic oscillation. All I know is its definitely between V2 and V3 as both channels have the issue which in my mind rules out V1 and I can see the issue happening on the PI without the power valves installed.

[neikeel]

Worth checking the silver mica betwen the PI plates too.

Does it do it if you lift the NFB resistor (as you would do if the primaries reveresed) normally constant scream or motorboating but all things need considering now

[Littlewyan]

I did but I haven't tried that since I rebuilt the amp. I did check the pi input and speaker output with a scope before the rebuild and they are out of phase. Amp has been wired up again the same way so its still out of phase. Its improved a lot since the rebuild as before I couldn't have the treble above 0. Must be wires coupling somewhere. No parts have changed except for a HT dropper resistor.

I also checked the PI cap before the rebuild, maybe I should check it again.

[neikeel]

Its improved a lot since the rebuild as before I couldn't have the treble above 0. Must be wires coupling somewhere. .

That would fit with what you described before, PO with a blocking effect but would have expected the suggestions first made to have solved the issue, like you trying to think laterally!

[Littlewyan]

I do wonder if it has something to do with the treble cap. As it only occurs when the treble is over 5 and the middle is on 0. Now if we think about how the tonestack works, if the treble is over 5 then that means there is less resistance between the treble cap and the pi cap than between the pi cap and the bass cap/pot. So the signal is mainly coming from the treble cap. When we turn the middle up the signal on bass cap/bass pot goes up. So now the signal is coming from another direction as well.

[Littlewyan]

Another thing to add is it also occurs on the normal channel. As I have the mixer capacitor installed most of the high end on the normal channel is drained so the oscillation surely cannot be in the first stage.

[neikeel]

Are your tone pots the correct values?

[Littlewyan]

Yes all pots read within tolerance. I got myself a new scope yesterday, went through the amp and found the issue is at the PI. It doesn't occur before the PI and it still occurs in the PI if I remove the power valves. So its either the PI or the Bias Circuit. I measured all of the resistors and they all read within spec so it must be a cap causing the issue. I'm wondering whether or not the bias capacitors are bad as I saw last night that during the rebuild I had swapped them around so wondered if that is why the issue has improved. The only other parts I have swapped around are the pre amp valves.

Anyway the issue appears to be that when I hit the PI with only high frequency content the 82K side will basically cut off and stops drawing current almost completely, whereas the 100K side draws a LOT more current. This could be why the sound goes bad as V4 is basically being cut off or hardly driven at all. As for the red plating I think that happened previously because A) Its an old Marshall and B) Because the last set of valves were a bit dodge.

Also I should point out that the amp is playable now as the middle control is turned up but I can tell it doesn't sound as good as it should. Still sounds kinda awesome though . I will investigate more either tonight or tomorrow night.

[Littlewyan]

Update.

I disconnected the PI from the tonestack and fed my oscillator into it. I sent a 4Vpp signal in which is plenty large enough and sweeped frequencies from 100Hz to 50Khz and tried sine wave, square wave and triangle wave. PI was stable the whole time. So its not the PI.

I monitored the output of the PI as well as the input and found that when I turn the middle up which stops the issue the following frequencies get squashed. 7Khz, 16Khz, 23Khz and 30Khz. They can also vary a bit so sometimes I'll pick up 32Khz instead of 30Khz. Now the PI swinging the way it does it caused by something upsetting the bias on one side, most likely on the 82k side. I highly doubt its a leaky cap as the PI input cap would have to leak and a tonestack cap would have to leak. Its not the NFB as I tried disconnecting that. So it must be oscillation causing it. I can trace the 30Khz back to V2A's anode, although its very small its still present and STILL gets squashed when I turn the middle up.

[VelvetGeorge]

Try shireded coax from the mix resistors to v2a grid. A lot of leakage happens between v1b and v2a.

If you suspect blocking distortion, install a high value resistor in series with grid. 100k-470k is fine.

George