Marshall JTM45 Problem

[AnalogPackrat]

Don't give up on it Stefan! CJ's stash must be running low or maybe he bumped his head on a rock last weekend or got tangled up in a hornet nest hanging over the river. Your amp is not breaking any laws of physics so you can and will find the fault.

Painful as it may be, how about trying this. Hook up your signal generator and use the speaker instead of a dummy load. Feed the amp a 1Vp-p sine wave at, say, 350Hz. Now turn up the gain until it starts to sound wrong. Turn off the amp and swap the speaker for a dummy load. Now trace backwards starting at the power tubes and note where the signal goes from "decent" looking sine to something else. Now you've narrowed it down. Be methodical.

A P

p.s. My Deluxe Reverb is doing the spitty/farty thing again, so I'm in a similar situation even if the problem is different. I'll be pulling the chassis in a few minutes to track it down.

 

[lotus]

Hello I'm no marshall expert, but i thought i would post a similar experience. I had a 73 50 watter doing the same thing your describing "distortion at 2 on knob and ugly sounding" turned out to be a failed diode. .

 

[AnalogPackrat]

Stefan,

I hope you're still working on your amp--maybe getting some help offline. I found and fixed my problem. I had parasitic oscillation back in the preamp. It did not make it through the PI intact, but did cause kinking in the input to the PI which was audible as nasty sounding distortion on loud, low notes even with the volume only on about 5. The distortion was raspy as the note faded away.

Anyway, I traced it back to the grid of the second triode in the vibrato channel so I started moving wires around and got it to stop. The amp now sounds great at any volume. Lesson learned: just because you don't see oscillation near the output doesn't mean that it isn't happening further upstream and still having an audible effect.

Good luck with your amp!

A P

 

[steppenwolf]

Hi!

I stopped looking for the problem for some days to go to work with fresh ideas...
You said your Problem was oscillation in the preamp, could you discribe, how it looked like on the scope?
I tried to swap the primary wires with the feedback detached, now I have oscillation on both pwr anodes...

I noticed, that the signal before the PI looks pretty good, but at the PI it distorts early, even with the powertubes pulled out.
I don't now, if its normal, I hope you can tell me that.
On 1kHz, I can get a decent sine at the PI output to about 5-6 on the volume..But on 2kHz the PI clips at 3 at the volume knob and at 4kHz it starts to clip around 2 on the volume...
Is that ok?

BTW, I brought it to an amp tech once and he said he couldn`t do anything for me, as it wasn't an original marshall and he wouldn't put his hands on it. Good tube techs seem to be in short supply over here...

The sine looks defenitly nasty at the anodes of the powertubes and a bit less nasty at the speaker out, but I really can't get a decent sine out of it at a moderate level and thats what I noticed playing it...

Thanks for helping out!
Stefan

 

[AnalogPackrat]

[quote author="steppenwolf"]
I stopped looking for the problem for some days to go to work with fresh ideas...[/quote]

I did the same and solved my problem when I came back to it.

You said your Problem was oscillation in the preamp, could you discribe, how it looked like on the scope?

I neglected to take any photos...stupid. Since my problem seemed to be worse on low notes, I set my sig gen to 100 Hz and about 1Vp-p. I monitored the output (dummy load) with one channel of the scope and used the other to look for problems. I traced backwards through the amp.

Here's a sketch of the output of the triode where the osc appeared:

The sine would be clean at low volume settings, but above a certain level the HF would appear on a portion of the wave as I've shown. This would survive into the last triode stage before the PI, but not at the input to the PI. It would, however, cause a kink in the waveform there which was audible.

I tried to swap the primary wires with the feedback detached, now I have oscillation on both pwr anodes...

I noticed, that the signal before the PI looks pretty good, but at the PI it distorts early, even with the powertubes pulled out.
I don't now, if its normal, I hope you can tell me that.
On 1kHz, I can get a decent sine at the PI output to about 5-6 on the volume..But on 2kHz the PI clips at 3 at the volume knob and at 4kHz it starts to clip around 2 on the volume...
Is that ok?

That sounds wrong to me. Which channel are you testing? Are you sure you have your tone stack wired correctly? If you input a very small signal, say 0.1Vp-p or less, can you sweep the frequency from 100Hz-10kHz in octaves and plot the output voltage? Just set the volume to something like 2 or 3, bass and treble to 0, mid to 5. It sounds like you might have a treble tilt somewhere in the preamp. You should have a roughly flat response over that range with

If your response isn't basically flat, trace backwards to find out where it's getting skewed.

BTW, I brought it to an amp tech once and he said he couldn`t do anything for me, as it wasn't an original marshall and he wouldn't put his hands on it. Good tube techs seem to be in short supply over here...

Yeah, there are some questionable ones around, though. Better to learn how to do it yourself, I think.

Thanks for helping out!

No problem. I'm no expert at this--I've only messed with half a dozen or so tube amps in my life--all Fenders.

A P

 

[CJ]

ok, sorry, back from sf med, everything good again...

the orig jtm 45 had the B+ taken after the choke
the reissue has the B+ on the upstream side.
how is your amp wired?

also, remember that guitar puts out a tiny signal, so you might need to build a voltage divider for your signal generator to lower the output.

i have a jtm 45 with kt66 im my closet, if we have to drag it out and do a step by step follow thru comparrison, thaen damn the torpedos, we will.

first off, anybody know the output level for 1 kc git with singal coil strat pickups?

me guess 10 mv max.

there is a good book out by doyle, you get stuff like this:

and after you get the amp workin, try and spank this guy:

 

[walter]

I rebuilt a fender vibrochamp, now it has an effects loop between pre and power amp, I use it for troubleshooting. Divide and conquer! Lift the cap feeding the phase inverter. Take the pre-amp signal and run it into the effects return of another amp. If you hear the problem, it's in the pre-amp or power supply. Now take signal from the effects send of another amp, go into the Phase inverter coupling cap (to bypass the pre-amp) if you hear the problem, it's in the output section or power supply. If you hear the problem in both situations, the problem is in the power supply.If the problem is in the pre-amp: Do you like Chinese food? get a chopstick to move wires around. I like to use shielded wire for the pre-amp grids, connect the shield on one end only, either to ground, or to the plate.

 

[steppenwolf]

Hi Guys!

@A P Thanks for that sketch, that gives me a good idea of how it shouldn't look like!

That sounds wrong to me. Which channel are you testing? Are you sure you have your tone stack wired correctly? If you input a very small signal, say 0.1Vp-p or less, can you sweep the frequency from 100Hz-10kHz in octaves and plot the output voltage? Just set the volume to something like 2 or 3, bass and treble to 0, mid to 5. It sounds like you might have a treble tilt somewhere in the preamp. You should have a roughly flat response over that range with

If your response isn't basically flat, trace backwards to find out where it's getting skewed.

I'm testing the "high treble" channel, but I don't use the treble bypass cap on the volume pot, so I shouldn't have a too trebly response, will check that, thanks for the idea!

he orig jtm 45 had the B+ taken after the choke
the reissue has the B+ on the upstream side.
how is your amp wired?

Hi CJ, you're back, thanks! Nice pic!!
The B+ is taken after the choke to a 1k resistor that feeds the 2 470Ohm resistors on PIN 4 of the powertubes...

also, remember that guitar puts out a tiny signal, so you might need to build a voltage divider for your signal generator to lower the output.

i have a jtm 45 with kt66 im my closet, if we have to drag it out and do a step by step follow thru comparrison, thaen damn the torpedos, we will.

first off, anybody know the output level for 1 kc git with singal coil strat pickups?

me guess 10 mv max.

I was told that I should use around 200mV because thats the average output voltage of singlecoils..can somebody confirm that?

I rebuilt a fender vibrochamp, now it has an effects loop between pre and power amp, I use it for troubleshooting. Divide and conquer! Lift the cap feeding the phase inverter. Take the pre-amp signal and run it into the effects return of another amp. If you hear the problem, it's in the pre-amp or power supply. Now take signal from the effects send of another amp, go into the Phase inverter coupling cap (to bypass the pre-amp) if you hear the problem, it's in the output section or power supply. If you hear the problem in both situations, the problem is in the power supply.If the problem is in the pre-amp: Do you like Chinese food? get a chopstick to move wires around. I like to use shielded wire for the pre-amp grids, connect the shield on one end only, either to ground, or to the plate.

Thanks Walter for that!
I'll try to scope the preamp stage after stage, if there are some problems further upstream, I also noticed a slight kink in the waveform at the PI without the powertubes installed...
Than, I will load the signal before the PI into the return of an other amp and look, what happens with it, and I'm feeding the PI of the JTM45 with the signal of a known working amp and see, if the problem is still there, this should get me even closer to the problem...

I measured the B+ voltage at the anode on the powertubes. I got around 435V idling. Turning up the volume knob the B+ drops to around 310V to ground and than strangely rises fastly to more than 1000V...I get funny readings...maybe thats because of the strong Signal...That sould be normal I think...

Another thing I noticed:
I posted that I have a unbalanced bias voltage after the 220k resistors before PIN 5 at higher volumes. (60V on one side and 72 on the other).
Moreover, one 220k drops 3V max and the other one 11V max...
I thought that would be due to matching of the power tubes, but I switched those around but that voltage drop didn't follow. After swapping the primary of the OT, the drop is still there, nothing changed, thats funny...

I'm sure I'm gonna win the battle :grin:

Have a nice sunday!
Stefan

 

[lofi]

WOW! sorry nothing to add, but from an educational point of view this is superb meta on trouble shooting valves!!!

this place rocks at times ..... but i think that us new kids get a bit worried arround this stuff so we take it one step at a time, thats probubaly realy annoying to the 'gurus' as they do ALL the logical steps in a sequence then report back, but we dont know when the logic stops and the dangerous starts so we play safe, no amount of reading replaces hands on after all :grin:

once again, sorry i cant help, but a big big thank you to all the guys teaching me for free, have some on me .. :sam: :sam: :guinness: :guinness: :green:

 

[lofi]

OT ~ sorry

Hameg HM203-7 20Mhz Oscilloscope

it seems all of you guys use one of these things, well proper versions of it anyway, so i have gone and evil bayed this, i think it will do till i understand how its working, then i can upgrade. plus its just been calibrated ..... so

any good instruction sites out there?

 

[steppenwolf]

Hi Guys!

Ok, some news...

I tried to split up the signal before the PI and fed my practice combo with the signal before the PI (into the return of the combo, sadly, it has a transistor poweramp...)
I scoped the output and slowly turned up the volume at the JTM45.
The sine is neat, no kinks, no oscillation. It squares out at max power, but it look really clean, pretty much like a normal square wave.

Than, I fed the combo with the 1kHz signal and went from the send of the combo into the cap before the PI of the JTM45 and scoped the JTM's output.
Clean sine up to around 4, than it starts to clip the kink appears and with rising volume, around 6 on the volume the oscillation can be dealt in and looks it's worse on full throttle...

That kind of eliminates the preamp as the source of the problem, am I right?

Have a nice week!
Stefan

 

[AnalogPackrat]

[quote author="steppenwolf"]
That kind of eliminates the preamp as the source of the problem, am I right?[/quote]

Seems like it, but I would still do a quick freq sweep of the pre just to make sure it doesn't have a funny response curve. If it's reasonably flat, then concentrate on the PI and power section. You'll find it--the clues are there.

Cheers,
A P

 

[CJ]

make sure the vol control is grounded
in fact, check all the grounds.
Marshalls always had grounding issues, is this an orig or DIY?
too lazy to retread.

the amp is breaking into ultra sonic oscillation at vol = 6.

how do i know since I can't hear it?
the tell tale signs were the weird readings on the grid resistors.
i should have figured this earlier, but that brain cell was burried in the ceribral hardrive

you don't hear the oscillation, you just hear the results.
this is why the B+ is dropping also.
the amp is using all its power to make rf.

so the chop stick guy is on it, your lead dress needs tweaking.
you need some extra capacitance to ground, or better yet, find out how the feedback is getting into the amp, oscillation is caused by positive feedback, if you ellimiate this, you do not have to do the wrong post CBS Fender fix, which is to add tone robbing caps all over the place.

finding the feedback reason can be a major pain, this is why this is taking so long.

i spent a year on a motorboating problem, only to read that the AT mic motorboats! wasnt the amp at all.

also, maybe use an iso trans on your scope.
i have seen weird things happen because of power line considerations.

you can use a 12AT or 12AU in the PI slot, if you want reduced gain.
this is a major diff between F3nder and MArshall, Fender uses the AT, M the AX.

 

[CJ]

here is a link to an old but cool book:

http://www.pacificrecone.com/JackDarrBook.html

 

[steppenwolf]

Hi Guys!

The last days I tried to move some wires with a chopstick. It was hard, as the wires used are teflon insulated silver stranded and those things are rally stiff! I couldn't get the oscillation go away.
I noticed that I can actually here the sine out of the amp thats normal I think, but I also hear that the sound changes as the osciallion is coming in. There is a high pitched wining added to the frequency that comes and goes while turning the knob. You can see on the scope, when the sound changes...funny...
I again suspected the OT as the last time I swapped it I was focusing on that notch...So, I swapped it again and I had the same result again. Scoped at the output the sine doesn't look too bad, only when dimend, it shows the sinusoidal. But watched at the anodes, only one seem to cause the major problem, the other behaves pretty good.
I think something is really unbalanced and causes the problems. I'm thinking of the BIAS supply that drops different voltages...

I'll try to change the tube sockets and all the resistors with them, maybe there is a faulty resistor or socket itself somewhere...
Is this a bad idea?

Thank you!
Have I nice day!
Stefan

 

[AnalogPackrat]

Have you checked or swapped out the 220k resistors on the power tube bias feeds? What about those 0.1uF coupling caps on the PI output?

A P

 

[CJ]

check your presence control.
it could be the output xfmr has a resonant frequency that makes it unstable.
you should sweep it with the sig gen and sciope, un loaded at both ends.
to see the ultra sonic osc, simple turn your scope frequency control, cm/sec. or whatever, up to a higher frequency, so it syncs with the garbage .

which tap does the feedback come from?
2 4 8 or 16 ohms?

email Gerald Weber.
he always gets back to me within a day or so, but he has been sick, so i don't know.
also, see hios book, 10 ways to stabalize an unstable amp, in particular, might help you out.

 

[steppenwolf]

Have you checked or swapped out the 220k resistors on the power tube bias feeds? What about those 0.1uF coupling caps on the PI output?

I ordered 2 new 220k resistors, but they measure fine...
I tried to change the 0,1uF, didn't change a thing, but thanks for your idea!

PostPosted: Wed Aug 29, 2007 5:22 pm Post subject:
check your presence control.
it could be the output xfmr has a resonant frequency that makes it unstable.
you should sweep it with the sig gen and sciope, un loaded at both ends.
to see the ultra sonic osc, simple turn your scope frequency control, cm/sec. or whatever, up to a higher frequency, so it syncs with the garbage .

Hi CJ, I had the same problems, even with the Feedback disconnected, can I exclude the presence control then?
I swapped the OT and had the same problems, but I'll sweep it again!

which tap does the feedback come from?
2 4 8 or 16 ohms?

I take it from positive tip of the output. 8 ohms....

email Gerald Weber.
he always gets back to me within a day or so, but he has been sick, so i don't know.
also, see his book, 10 ways to stabilize an unstable amp, in particular, might help you out.

Is this guy a forum member? Can't find anything about him...

 

[CJ]

He owns Kendrick.
try that.

long shot: connect a wire from your spk frame to the amp chassis.
no electrical reason why this should work, but sometimes it does.

the noise should be thinner lines, but higher amplitude than the input freq.

 

[pucho812]

Hey did the bridging rectifier ever get replaced?