Marshall JTM45 Problem

[steppenwolf]

Hi guys!

I just scoped the directly at the anodes of the KT66 (btw I'm only using a 100:1 probe...) and that's what I got:

Left anode:



Looks ok, but you can see the notch, maybe that is allright...

But here the right anode:



and zoomed in:



What do you guys think about that?
The volume was identical around 7-8 on both readings, I didn't touch a knob....

Is this hf oscillation? It should have around 50kHz

Thanks for your support :grin:

Have a nice evening!
Stefan

 

[plexi]

I assume the output at the speaker has the "ringing" on it too. Lead dress and layout, around the power tubes, maybe PI too. Too much leakage ind in the OT, maybe, but you swapped and it stayed the same, so I still suspect layout.

 

[steppenwolf]

Hi!

I'm posting two pics of the layout around the power tubes, I tried to move some wires around, but nothing really seems to change the oscillation...

 

[VacuumVoodoo]

The blue wires from PI to grid stopper resistors on pins No.5 of power tubes are running in a bundle with other wires. Move them away from bundles and make them as short as possible. You might even replace them with shielded wire.

 

[steppenwolf]

:wink:
I already had those wires shielded and removed the shield to verify, if capacitive coupling was the problem, it obviously wasn't...

The notch I talked about much doesn't seem to be cross over dist...
I can't get it go away even with the bias maxed...
But when I look at the anodes, with increasing volume this notch appears and "climbs" up the sine to become part of the oscillating frequency...

:?:

 

[VacuumVoodoo]

Well, it's not a sine but a square wave, so you're withholding an essential piece of information: is the squaring already happening in the preamp or is this power stage saturation? In power stage saturation this "riding notch" happens in output transformer when plate current exceeds what the transformer can take. Going up and down on the square wave flanks the magnetic field in the core lags behind. I've seen this in a couple amps with undersized and cheaply (i.e. single section primary, high interwinding capacitances etc) wound OTs. On the other hand this can produce an interesting "swooshy-flangy" sound that some guys like.

What OT is installed in the amp?

Since you're in Germany: you can call me on Skype using my forum nick if you'd like to talk.

 

[steppenwolf]

It is a sine with saturating power amp...

The OT is a Mercury Magnetics, not really cheap :roll:

What about the oscillation happening only on one side of the poweramp?

With a 1kHz sine, the oscillation is less than with 2khZ or 4kHz...just wanted to mention that
Greets

 

[AnalogPackrat]

Can you move the power tube grid feeds so that run as directly as possible between the PI and the power tubes? If they must cross the plate connections, keep as much air between as possible and make them cross at 90 degrees. What size grid stoppers do you have on there? <edit--never mind on the orange cap question--bias supply--duh>

Could you post another picture a bit down and left of the ones you already posted? The pics are helpful.

BTW, you said you have heard friend's JTMs--are they built from the same kit? Can you compare their lead dress to yours?

A P

 

[steppenwolf]

Hi everybody!

Can you move the power tube grid feeds so that run as directly as possible between the PI and the power tubes? If they must cross the plate connections, keep as much air between as possible and make them cross at 90 degrees. What size grid stoppers do you have on there? <edit--never mind on the orange cap question--bias supply--duh>

I'll try that, but I doubt that could solve the problem...
The reason why I first put a sine through the amp was that I didn't like its sound compared to other JTM45 I heard. It breaks up to early (I can't get a clean sound out of it at moderatly high volume and there was no inbetween sound between clean and distorted sound. It just gets mushy und undifined and the distortion sounds fuzzy and weak. I'll post a sample soon!
I'm not very experienced how to interpret what I see on the screen, that why I ask you guys as you are the experts!

I just thought I could get to the problem fast with the scope, but that was obviously a mistake...

BTW, you said you have heard friend's JTMs--are they built from the same kit? Can you compare their lead dress to yours?

More or less, yes, one of them had his kit from ceriatone, too...

Question: Could such a problem also be caused by a damaged grid stopper resistor or screen resistor even though it measures within tolerances?
I have 5,69k on one side and 5,79k on the other?

Thanks again!
Stefan

 

[CJ]

thanks for the pics, now we can fix this.
swap output transformer leads, the ones that go to the pwr tube sockets, re scope and report back immediately.
did the damped sinusoidal follow the output transformer around?
when you use full tilt boogie (set amp on 11), can make both tubes display the damped sine wave?

if i was the amp tech for buddy guy, and he was screamin at me somewhere on a stop near winemucca nev, i would beg borrow or steal a different output transformer and try that.
the roadie for SRV used to carry a whole sitload of stuff around with him, what was his name?
is he RIP?


i think mercury screwed up.
that champ xfmr i took apart from soundguy was, well...of "questionable" quality.
(pos)
they probably went cheap and wound pri-sec, no interleaving, which meazns one half of the pri is different from the other, it has more capacitance, measure capacitance from sec (any wire, disconnect spk)
each side of the pri leads that feed B+ to the pwr tubes.
i bet my dead granny you get two different readings.

what are you feeding the amp again?
music or sig generator?

 

[steppenwolf]

thanks for the pics, now we can fix this.
swap output transformer leads, the ones that go to the pwr tube sockets, re scope and report back immediately.
did the damped sinusoidal follow the output transformer around?

Should I also swap the wires coming from the PI going to pin 5 on the pwr tubes to avoid positive feedback?

when you use full tilt boogie (set amp on 11), can make both tubes display the damped sine wave?

On full tilt, the other side stays pretty clean but also shows a tiny bit of oscillation but on the negative side of the sine (down the sine...)
But it isn'T paragonable to the one on the other side!

they probably went cheap and wound pri-sec, no interleaving, which meazns one half of the pri is different from the other, it has more capacitance, measure capacitance from sec (any wire, disconnect spk)
each side of the pri leads that feed B+ to the pwr tubes.
i bet my dead granny you get two different readings.

I measured between primary leads: Get a reading of around 45nF
Between primary and centertab (B+) I get 193nF on one side and 194nF on the other...

what are you feeding the amp again?
music or sig generator?

I'm feeding it with a signal generator...
thanks again!

 

[Johan]

what happens if you lift the feedbackloop?
should be easier to trace the problem if it isnt going round and round...and how big are your bias feed resistors?
j

 

[steppenwolf]

what happens if you lift the feedbackloop?

I allready did that. I think all that just happened earlier on the volume know, but I'll check that again!

and how big are your bias feed resistors?

Are the bias feed resistors the ones attached together and the other ends going to the 0,1uF coupling caps?
Those are 0,5Watt carbon combs, one measures 228k, the other one around 240k...I measured a voltage drop across it when driven the power stage hard. On one the drop was around 2-3V, on the other up to 11V...Strange...

 

[CJ]

see if it will run with zero feedback and not go crazy.

come up slow with the signal strength.

then scope out those two plates again.

BTW, that is transformer ringing, not zero crossing.
The second I saw that pic, I knew it was the iron.

 

[steppenwolf]

I just wanted to add an information.
The pictures posted and the comments I made were based on a 4kHz Sine, as the effect I described gets more evident with higher frequencies...
But I think that 4 kHz are in the electric guitar spectrum, I mean, 20kHz would be exaggerate, but 4kHz should be a frequency that my guitar produces anyway...
The pics I'm posting right now are 1kHz sine...

see if it will run with zero feedback and not go crazy.

I didn't go crazy, but all that oscillation happened earlier...

come up slow with the signal strength.

then scope out those two plates again.

Here are the pics of the right anode first with feedback:


and without feedback:


and here the right, bad :wink: anode, with feedback:



and without feedback:

BTW, that is transformer ringing, not zero crossing.
The second I saw that pic, I knew it was the iron.

Is ringing normal? I mean, I tried a different OT tranny, it had around 4k Primary and is meant for a 50Watt JMP, it was physically bigger than the MM but it also showed that notch...
I think that tranny should easily take the mighty 30 Watts, or maybe 45Watts at full throttle of the JTM45... :?

 

[CJ]

are you feeding it a square wave?

 

[Johan]

with the feedback detached, can you spot the problem earlier than the output anode?..what about on the grid? ( pin 5 )
..with the feedback still detached, can you spot it on the PI anodes?...if so, what about the cap that goes from PI "right side"-grid to ground? (usualy 0.1uF on Marshalls)
what about the caps that goes from the PI to the bias feed..maby one is leaking DC ever so slightly...it is at least odd that one bias feed resistor drops 11volts when signal passes, when the other only drops 3...
j

 

[CJ]

looks normal to me, i think this guy is either pulling our chain, or he has never heard a marshall before.
i sense a lot of in experience, and its starting to make me snap!


sad. would take any good tube tech 5 minutes to sort, but we need three weeks.

 

[VacuumVoodoo]

CJ, he feeds sine to the input and drives the power amp into full saturation. What he sees is pretty normal, the differences in left/right plate waveform are due to wiring and transformer asymmetries. I pointed this out a few post earlier but, alas, it went unnoticed :sad:

To verify this I just made same measurement on a 1972 JCM I just renovated: very similar results visually on the scope. Since Stefan doesn't tell what the scales are in his scope pics I can't say if his amp does this at expected power level or not.

 

[steppenwolf]

looks normal to me, i think this guy is either pulling our chain, or he has never heard a marshall before.
i sense a lot of in experience, and its starting to make me snap!
sad. would take any good tube tech 5 minutes to sort, but we need three weeks.

Sorry, I didn't meant to bug anybody...and yes, I have played Marshalls, especially JTM45 and none of them started to distort at 2 on the volume and none of them had an ugly distortion...
Maybe you think I'm just a bugging newbie, but I know what I am doing...
I couldn't find the reason, why this JTM45 sounds the way it does, and I tried everything, changing tubes, measuring bias, swapping the OT, changing loudspeakers...As nothing helped I thought you guys could help me, especially as I'm not experienced in Scopes. Somebody spotted breaking through on a sine wave I posted earlier, and I trusted that comment and searched in that direction...

Sorry again for bugging, I leave that amp as it is, may it rest in peace :sad:

Best regards,
Stefan