Marshall JTM45 Problem

[steppenwolf]

Hey did the bridging rectifier ever get replaced?

This amp uses tube rectification. I swapped the tube and tried solid state rectifiers, but that didn't help...thanks again!

 

[CJ]

you don't need to run the amp to sweep the xfmr.
just pull the pwr cord and spk jack.
cold tubes are an open circuit, so just clip the sig gen leads on the tube sockets.
you will get a less noisy waveform if you connect to the output and step up above the noise floor, but i don't know if your generator will drive a spk winding.

sweep from 10K on up.
you should see the voltage ratio change, ie, you injct 30 v p-p and get out 300 mv p-p, all the way up the freq band.
at some point, the xfmr output voltage will begin to rise, then fall back to zero as you keep increasing the freq.

 

[Junction]

I don't have the energy to go back thru this thread and catch up with progress, but I have a JTM45 in hand for the next 2 days, if there is anything you would like me to check I can, I have an oscilloscope but no sine wave generator.

Michael

 

[steppenwolf]

Hi!

I don't have the energy to go back thru this thread and catch up with progress, but I have a JTM45 in hand for the next 2 days, if there is anything you would like me to check I can, I have an oscilloscope but no sine wave generator.

This is very kind of you! I don't know if this works, but could you measure the BIAS voltage at each of the PI 0,1uF while turning up the power and a signal through it. (Do you have a dummyload?)
Do you get different reading at higher voltages?

I don't know what else to check without a signal generator or a cd player with a sine on it.

But thank you very much!

I'm going to exchange the sockets plus all the resistors around to see, if that changes anything but I doubt it. There most be a positive feedback somewhere or a great unbalance but it gives me a hard tie as I can't find the source...

I recently read an article (I think on the Bogner site) that says, that every wire and every component radiates a magnetic field that can cause positive or negative feedback and can dramatically change the sound even though it is not a real malfunction...
I even read that the earliest Marshall used small wires that caused a capitance to ground. They said that could have contributed to the mystical JTM sound.
That's to much voodoo for my right now... :shock:

 

[CJ]

hold off on ripping it apart, i have a bucket load of stuff you have not tried...

 

[AnalogPackrat]

[quote author="steppenwolf"]
I don't know what else to check without a signal generator or a cd player with a sine on it.[/quote]

Google around. There are some wav files out there with various test signals and I'm sure a 1kHz sine is among them. Send a pointer to Junction when you find it. Hopefully he can burn a CD with the wav and play it...should work fine.

I recently read an article (I think on the Bogner site) that says, that every wire and every component radiates a magnetic field that can cause positive or negative feedback and can dramatically change the sound even though it is not a real malfunction...
I even read that the earliest Marshall used small wires that caused a capitance to ground. They said that could have contributed to the mystical JTM sound.
That's to much voodoo for my right now... :shock:

All of that is why we keep harping on you about lead dress. The problem is not necessarily something that's on the schematic--parasitics are different from amp to amp due to lead dress. When I got my Deluxe fixed I can tell you that the changes I made were very minor.

Don't exchange sockets and parts willy nilly. You're just wasting your time. You need to cut those cable ties and start rearranging the wiring from the board to the sockets in the PI and power tube area. Start with moving the grid wires on the power tubes up and away from everything else (above the board, not around or underneath it).

A P

 

[CJ]

ok, weird story, guy grabs tube socket, osc stops, he thinks replacing socket helps, but his hand, when it touched the tube to wiggle the socket and check for looseness, it added capacitance, so he changes socket, nothing haoppens because his hand aint there, shunting hi freqs to ground.

ok, we need to either:

1) improve grounding
2) alter layout
3) reduce high end response
4) reduce gain

1. install grounding buss, bare solid wire that goes from each pots ground lead and then the input jacks.
remove jacks and clean any corrosion while you are in there.

check rersistance of all grounds. set onmmeter to lowest range, tie one end to chassis with clip, go around and check every ground. anything over 1 or 2 ohms is suspect.
don't forget to dis regard test lead resistance.

check output trans grounding. check sec ground connection.

check speaker ground(see weird idea in earlier post)

2) move output transformer. you might have coupling going on. move it while playing, see if you can find a better location.
even hor to vert can help.

reinstall series grid resistors closer to socket.
shorten up leads, so only resistyor body is between socket pins.
1/4" can make a difference.

shorten wires in all grid circuits. go thru the amp and shorten every wire.
(pins 2 and 7 of 12ax7, pins 5 on 6v6,6l6,el34, etc.

use shielded wire from input jack to first grid

re position coupling caps. these can sometimes cause problems via the input jack, mostly 5f3 delux, not marshall.

3) install series grid resistors. starts with smallest value that stabilizes amp. if 2.2 k stops it, try 1.5 k.
go all the way up tyo 10 k.

mount resistor body directlly up against the solder lug, as mentioned before.

add grid to ground capacitance.
this is the last resort, silver face fender mod.
it works, but kills tone, so i won't even tell you how.

shunt a plate load resistor with a cap.
works in plexi marshalls.
idea is to bypass hf around plate resistor.
5 pf to 2000 pf, use lowest value, either from plate to grid, or plate to catheter

4) add more neg feedback. hook the feedback line to the 16 ohm term, not 8 ohm.

change preamp tube biasing by increasing the cathode resistor value. for instance, if you have 1500 on pins 3 or 8, 12ax) try 1800 or 2200. just a little reduction might help. solder in a trimmer pot to find the right value.

lower plate voltages on preamp tubes-this reduces gain, and curbs high end response. this involves changing the power supply resistor to a higher value.

make sure you use the same wattage or above.

special cases: spk inductance is not right, causes phase shift= osc.
change spk type.

conductive circuit board, sometimes the old fender black boards, the dye they used combines with humidity to cause leakage currents, this only applies to old fenders.

hope this helps...
cj

 

[Junction]

[quote author="steppenwolf"] I don't know if this works, but could you measure the BIAS voltage at each of the PI 0,1uF while turning up the power and a signal through it.
Do you get different reading at higher voltages?
[/quote]

When measuring at the junction of the 0.1uF caps and the 220K resistors, whilst varying the volume with a sine wave source, I get the following;
-47V DC to -50V DC on one side
-47V DC to - 65V DC on the other side
A little surprised this was not more balanced

A JTM45 is dirty at the best of times, the sine wave starts to distort and kinda twist around half volume and then squares off at high volumes.

I dont get the high frequency crap riding on the sine wave, that you have seen.

Let me know if there is anything else you want me to check.
Cheers
Michael

 

[steppenwolf]

All of that is why we keep harping on you about lead dress. The problem is not necessarily something that's on the schematic--parasitics are different from amp to amp due to lead dress. When I got my Deluxe fixed I can tell you that the changes I made were very minor.

Don't exchange sockets and parts willy nilly. You're just wasting your time. You need to cut those cable ties and start rearranging the wiring from the board to the sockets in the PI and power tube area. Start with moving the grid wires on the power tubes up and away from everything else (above the board, not around or underneath it).

Question. Should I also try to move ground wires or DC wires around?
I already opened those cable ties and moved the signal cables around without any effect...

1. install grounding buss, bare solid wire that goes from each pots ground lead and then the input jacks.
remove jacks and clean any corrosion while you are in there.

check rersistance of all grounds. set onmmeter to lowest range, tie one end to chassis with clip, go around and check every ground. anything over 1 or 2 ohms is suspect.
don't forget to dis regard test lead resistance.

check output trans grounding. check sec ground connection.


check speaker ground(see weird idea in earlier post)

There is already a really big solid copper wire installed at the board, all the ground connection of the pots go there and on wire goes from there to a starground. I'll add some pictures at the end...

Checked all ground connection, all of them read 0,2 Ohms, tips together 0,2 Ohms , too...

The only thing that comes to my mind is the grounding of pins 1 and 8 of the powertubes. They are directly grounded at the sockets screw. Also the OT 0 Ohm connection goes to one of those srews, not to the star...

2) move output transformer. you might have coupling going on. move it while playing, see if you can find a better location.
even hor to vert can help.

I have a big Mu Metal can. I could place that one over the OT and ground it...Bad idea?

reinstall series grid resistors closer to socket.
shorten up leads, so only resistyor body is between socket pins.
1/4" can make a difference.

I'll try that! I'm going to isolate the wire to the resistor...

shorten wires in all grid circuits. go thru the amp and shorten every wire.
(pins 2 and 7 of 12ax7, pins 5 on 6v6,6l6,el34, etc.

They are pretty short already. There is no excessive wire anywhere, erverything is as short as possible!

3) install series grid resistors. starts with smallest value that stabilizes amp. if 2.2 k stops it, try 1.5 k.
go all the way up tyo 10 k.

mount resistor body directly up against the solder lug, as mentioned before.

I have 5,6k installed on the tubes...should I try 10k?

For the rest I wouldn't dare to mess with the circuit design it self :wink:
But I'll try, if nothing else helps...

When measuring at the junction of the 0.1uF caps and the 220K resistors, whilst varying the volume with a sine wave source, I get the following;
-47V DC to -50V DC on one side
-47V DC to - 65V DC on the other side
A little surprised this was not more balanced

So you have this phenomena, too...Strange

A JTM45 is dirty at the best of times, the sine wave starts to distort and kinda twist around half volume and then squares off at high volumes.

I dont get the high frequency crap riding on the sine wave, that you have seen.

Let me know if there is anything else you want me to check.

Where did you scope it, at the output or at the anodes?

Here some pics:

 

[Junction]

Where did you scope it, at the output or at the anodes?

I scoped at the speaker output with speaker attached, had the neighbourhood complaining about the 1k tone!

I see you have a Ceriatone PTP board, but PT and other components from elsewhere. Mine's actually a full Ceriatone build and very happy with it, I've done 3 of these recently with great success.

Have you checked their layout, it works, link as follows;
http://www.ceriatone.com/images/layoutPic/marshallLayout/JTM45ceriatone.jpg

I suspect, like others have, that your problem may just be down to lead dress and earthing etc

Michael

 

[steppenwolf]

Hi Guys!

The next 7 days I'm in Italy on vacation :grin:
I'll post back as soon as I'm back!
Thanks again for your help!!!

Hear you soon!
Have a great week!
Stefan

 

[walter]

That amp looks nice. I like to connect the ground of each pot to the case of that pot, as well as the ground buss to ensure each pot is shielded. And your speaker negative, it must be connected to the output transformer and the chassis. If the OT is grounded to chassis, ground the speaker jack at the same point. When the amp is mounted in the cab, is there a metal plane on the guts side to shield rf? you may need a sheet of metal, or foil on a piece of wood placed over the chassis when testing out of cab. For the O-scope, use a ground lift on the power plug for working on equipment with multiple grounds. Gerald Weber does have some good info, damn good salesman aswell. I used to be on page 100 of one of his books.

 

[steppenwolf]

Hi guys!

. When the amp is mounted in the cab, is there a metal plane on the guts side to shield rf? you may need a sheet of metal, or foil on a piece of wood placed over the chassis when testing out of cab.

I put a thick grounded aluminum plate over the amp when measuring, sadly it doesn't change anything

For the O-scope, use a ground lift on the power plug for working on equipment with multiple grounds. Gerald Weber does have some good info, damn good salesman aswell. I used to be on page 100 of one of his books.

I don't think the scope gives me a wrong reading as I can actually hear the problem as well...

This thread was dead for a few months now. I started my classes at university again and just had no time and no motivation to fix it...

Now I decided to give it another try.

For all those who don't know whats going on:

My JTM45 (clone) sounds VERY buzzy on distortion and lacks headroom. It begins to distort on volume setting 2 and the highs sounds like playing a guitar through a overloaded mic pre...

That is exactely what I can see on the scope.

Putting a 1kHz sine through it you can see how it squares up nicely at around volume 5 as it should be. But you can see some breaking through at very high levels.
Increasing the sine to 2kHz,it begins to look weird at volume 3 and there is a strange oscillation going on when increasing the volume.
With a 4kHz sine I nearly can't get it clean even at low levels.

When monitoring the B+ with the 1 kHz sine it gradually sags. That should be normal due to the rectifier tube...
On 4 kHz the voltage drops down considerably when the oscillation begins...

Yesterday I recorded some tracks with it in direct comparison with a friend's Marshall on the same setup, same mic pos and same cabinet.

The differences where obvious!

I already tried to swap tubes, rebias, changing OT, fiddling with the layout and really nothing changed a bit.

I'm going through it again, maybe I'll find something wrong or maybe someone has a suggestion in which direction I could look, although you gave me a LOT OF ideas already....

Thank you!
Stefan

 

[Sorr]

Stefan,

Didnt you mention in an early post that you discconected the pre to the PI and fed another preamp signal into the PI and it worked OK?
Sorr

 

[steppenwolf]

Didn't you mention in an early post that you discconected the pre to the PI and fed another preamp signal into the PI and it worked OK?
Sorr

Hi!
No, sadly the problem persisted even with another signal fed into the PI. Meanwhile I changed the KT66 sockets and all the resistors around that area.
I located the trouble and it defenitly should be around there, so I hope it is a bad socket. (Maybe arching or whatever...)
I came to that conclusion because the problem happens mainly on the right KT66 tube. Now if I swap the tubes and put the right in the left socket and vice versa the problem stays at the right anode...

I really hope it is just a bad socket but I'll have to test it!

The only thing I could think of is bad grounding (checked, it is good), bad layout, (moved wires, broke up bundles.., and a faulty OT, swapped that one too with no change!)

Thanks again! That amp is really bad patient.

 

[steppenwolf]

After having changed the sockets and some resistors with them, the problem persists. But I anticipated that...
It begins to oscillate with frequencies higher than 1kHz.

The only strange thing I notice measuring voltages are the unbalanced BIAS voltages. On idle on both tubes I read around -50V. When pushing the volume up, one grid goes up to -55V max, the other one to more than -65V.
I thought that could be a badly matched tube but even if I change the positions of the two KT66, the voltages read the same and on the same sockets as before although I swapped the tubes around... :?
That tells me that it comes from somewhere else...
Any idea from where?
Thank you!

 

[AnalogPackrat]

Glad you haven't given up on this yet. Since you have data from another person with a functioning amp that has the same bias thing, I think you should be looking elsewhere. I was thinking about your earlier experiment where you broke out the preamp to PI connection and tried to isolate the problem that way. Isn't it possible that the problem is parasitic and disappears when you break that connection?

When you get the thing to misbehave do you see any unusual behavior back in the preamp stages?

A P

 

[steppenwolf]

Glad you haven't given up on this yet. Since you have data from another person with a functioning amp that has the same bias thing, I think you should be looking elsewhere.

You might be right...But that was the only problem I could actually see...

I was thinking about your earlier experiment where you broke out the preamp to PI connection and tried to isolate the problem that way. Isn't it possible that the problem is parasitic and disappears when you break that connection?

When you get the thing to misbehave do you see any unusual behavior back in the preamp stages?

When I broke up the connection between Preamp and PI and fed the PI with an external signal, the problem was still there unchanged. When that nasty oscillation thing is happening, there is no change in the preamp waveform (except some drop in level when I push the power amp but that should be due to B+ sag). The Waveform after the PI has this oscillation, too, but when I disconnect the negative feedback, the waveform is clean there but stays weird at the anodes and at the output.

I'll try to describe the sound of the amp:
On really low volumes, it sounds good but as I push the volume higher than 2-3 (with a strat...) the highs start to distort and I'm talking transistor like here...You now the sound of a guitar through a overdriven DI in. that gives you the idea. The speakers limited response calms it down a little but it's still just buzzy. That is very noticeably when the signal is about to break up. When pushing it harder, things just get more compressed and lack power but the sound stays the same...
That might sound like a bad OT, but that amp actually puts out a lot of volume. Moreover I changed the OT and the problem was still there!

That is pretty much what I can see on the scope. On lower frequencies (1kHz and lower) everthing is fine and it stays clean to about 5 at the volume knob.
On higher frequencies I get an increased oscillation and the higher the frequency the earlier it happens...

That is so STRANGE!!!

 

[Johan]

perhaps try and rotate the three 0.1 uF decoupling caps at the P.I and see if anything changes...perhaps one is leaky or the one from the P.I right side grid to junction of precence-pot and feedback isn't holding its expected value?..
johan

 

[cd27]

Hi, new member here. Read all 6 pages of this thread this evening with interest...  I wonder if you ever got to the bottom of this mystery?

I have a tubedepot jtm45+ build that was working fine but seems to have developed the same 'sound'. I've swapped out tubes and OTs and redone the Power Tube sockets once already. Lead dress had been fine originally and shielding foil kept hum down. The voltages all read within tolerance but the sound has the same issue - distortion appears early, sustaining notes have a sputter/interrupted sound, chords go into complete fuzz. I have a decent multimeter but don't own a scope.

What's odd is that the amp seems to behave better on a Variac rather than plugged into the wall power either at home or at my studio. My Furman Power conditioner has no effect in this regard. When plugged into the wall socket (reading 122VAC before rectification) my amp has around 483VDC on the Power Supply Caps and the tube sockets. However, it reads higher when going through a Variac (same as Isolating PT?) The Power Supply Caps are JJ, rated at 500VDC, but produce as high as 540VDC when the Variac is all the way up and the solid state diode rectification option is engaged.

Also when I turn the standby switch on the sound continues to pass through the speaker and decreases in volume over a period of around 5 seconds. I'm thinking a cap is discharging, but which cap would still have a path through to the speaker?