1968 JMP50 problem - need suggestions

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
I'm having trouble tracking down a noise in my 68 JMP50.

When I got this amp a couple of years ago it was in need of some repair - all 1M pots were completely shot and it had the wrong value bass pot and the tone stack wrongly wired. Power transformer is a Marstran replacement, but everything else except the 1M pots and the power transformer is original. I added 5.6k output grid stoppers and 1k screen grid resistors.

After getting it back to stock it has sounded just glorious, until a weird static noise started to ride on the trails of the notes. It *may* have been there all the while, but I don't think so. I've added a video clip below which hopefully demonstrates it, but it's not easy to hear clearly in the recording as the note itself overpowers the buzz, but it's very clearly there. There's a more low pitched "murmur" at the attack of the note, followed by a more buzzing, static sort of sound on the trail and sustain of the note.

What's interesting is this;
- if I have an attenuator in line between the amp and the cab the noise goes away
- if I disconnect the purple feedback wire from the output jack it goes away. Doesn't matter which output (4,8 or 16 ohm) it's connected to.
- the noise is most evident with presence at 0 and almost entirely gone with presence at 10.

From this I deduce that there is something present on the output which feeds back into the phase inverter. That "something" only presents itself when it is being fed into the phase inverter again. If an attenuator is present, that "something" gets shunted to ground. Same thing when it passes through the presence control - shunting high frequencies from the negative feedback loop to ground by turning up the presence control prevents them from entering into the phase inverter to be amplified again.

I have eliminated the usual suspects; tried known good preamp and power tubes. I have subbed in known good and working potentiometers (all except the middle one, but I am looking harder at the output stage now). I have also eliminated the preamp filter cap and the phase inverter/screens filter cap by subbing in new ones. I also removed the output tube grid stoppers just to be sure. No effect whatsoever.

What next? I am running out of suggestions. I probably need to also try subbing in a different mains filter cap, but it tests just fine and there's no hum or any other problems. I would sort of expect a problem with a mains filter cap to also present itself with the negative feedback loop disengaged.

I would love to hear your suggestions guys. Hope you can get something out of this video:

 
Last edited:

Jon Snell

Well-Known Member
Joined
Apr 12, 2020
Messages
294
Reaction score
322
Location
Jurassic Coast, England
By adjusting the presence control, you are adjusting the NFB slope, proving it is an output issue.
Have you checked the EL34 bias voltage. Don't forgrt to remove the grid stoppers are they may cause the Miller Effect, which is what you are hearing.
 

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
Bias voltage is where it's supposed to be, I think around -40V. I've tested this across several sets of output tubes with the same issue.

I have already removed the output tube grid stoppers. Or did you mean screen grid resistors?
 

Jon Snell

Well-Known Member
Joined
Apr 12, 2020
Messages
294
Reaction score
322
Location
Jurassic Coast, England
Bias voltage is where it's supposed to be, I think around -40V. I've tested this across several sets of output tubes with the same issue.

I have already removed the output tube grid stoppers. Or did you mean screen grid resistors?
Grid stoppers is what I meant.
What is the current draw of the EL34s?
No signal of course.
 

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
Current draw read with a Weber Bias-Rite is about 40 mA each with 425 VDC anode voltage.
 

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
If I reversed the primaries I would get an awful howl, at least with the negative feedback loop hooked up.

In the power amp, the only connections I've touched are pins 5 where grid stoppers were added (now removed) and pins 4 (for adding screen grid resistors).
I redid all the potentiometers when I got it as I needed to replace the 1M ones, and in the process redid all the ground connections in the preamp, removing the buss wire across the pots and adding a common ground point for each preamp stage ("Larry" grounding as it was called on the Metro forum).

But seeing as this seems to be a power amp issue I'm searching there first and foremost. I suppose the mains filter cap is still a suspect.
 

AndyD

Well-Known Member
Joined
Sep 13, 2018
Messages
859
Reaction score
938
Location
UK
Could this be a speaker/cabinet resonance issue?
 

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
No, it happens with all cabs I use. It’s not as obvious in the clip as it is in real life.
 

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
And I really appreciate all suggestions! I didn’t intend to ignore or dismiss anything.

If it was unclear; the output transformer is original, and no connections to the output tubes have been touched except, as mentioned, pins 4 and 5. Which is why I won’t be reversing the primary wires…unless I misunderstand something? If I do that they will be out of phase and the NFB loop will become a PFB loop and it howls…
 

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
27k going to the output selector common lug, in other words similar to having it wired to the speaker jack so it varies according to the output tap used (always 16 ohm in my case). I tried hooking it directly to the 16 ohm or 8 ohm tap, no difference really.
 

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
Sure, several silver micas, maybe ceramics. Shoot out any suggestions.
 

Whatwhatringrang

Well-Known Member
Joined
Aug 5, 2022
Messages
244
Reaction score
293
You don’t need to answer these . More just questions I ask myself about this here. So don’t worry about taking the time to answer this random stuff in question marks.

Try the mains 1st electrolytic??? Then check the 100k and 820r resistors or look for heat damaged resistor (cc ) A few in this one . Maybe one of small ceramic or mica caps? Or even the resistor or the 250uf 6v cap?? Look really close for damage or excessive drift.

Attenuators hide all this stuff I notice. I can hear it now but at first I could not. Does this happen when dialing the voltage up on variac to certain voltage point or just when playing like described? Say it is good does not become apparent till your reach say reach 200-205 Vac? When playing it

That is what i would do . I am not 1/4th as knowledgeable as people on here. Check any carbon comps and look for excessive drift or damage . Just what I would do.

Good luck I hope you get it figured out . I am sure you will . Maybe something right in front of you. Great Cabs and amps ! Just trying to help not act like a expert of trouble shooting lol. Which I am certainly not.
 
Last edited:

shakti

Well-Known Member
Joined
Jul 14, 2016
Messages
368
Reaction score
408
OK, so bypassing the 27k with a 250 pF very nearly removes the noise. Not entirely, but much less noticeable.

So we're on the track of something. I like to try to understand what's going on rather than just go replacing stuff. So when I bypass the resistor, I have a high pass filter across the NFB resistor, allowing all those high frequencies back into the phase inverter but out of phase, so they cancel out.
Is this the sign of a faulty negative feedback resistor? It is a carbon comp. Or could it be a poor solder joint under the board? I have of course tried to tap the wires to see if that elicits any noise, but no success there.
 

Pete Farrington

Well-Known Member
Joined
Jan 18, 2021
Messages
2,043
Reaction score
2,338
Location
Staffordshire UK
There’s nothing wrong per se, just ’non ideal’ design, more negative feedback than the OT can accommodate. Phase shifts within it result in poor stability with NFB, leading to a tendency to oscillate up in the ultrasonics, typically 40kHz - 140kHz.
The added cap brings the phase of the NFB signal up there back into line.
Alternatively, just reduce the amount of NFB by the simply turning up the presence. As the issue only occurs at low presence settings.
Bear in mind that early Marshalls used the same NFB network resistor values as the 5F6A Bassman, but sampled from the somewhat higher signal voltage present at a 16 or 8R output, compared to the 2R output of the 5F6A.
And a big chunk of the available window for the OT secondary was taken up with the windings for the 100V output. So the OT will tend to have worse primary to secondary coupling (and hence high frequency performance) than the one used in the 5F6A.
 
Last edited:

Latest posts



Top