Anyone want to help me troubleshoot a Polytone 104 "Maxi-Brute" guitar amp?

[dmeek]

Back in post #2 you noted the cut leads on U12B. This is important. Pins 8 and 13 will
interfere with the FET switching. make sure they are cut or bent out of the socket.

Also I would remove or clip R50 and R51

[g1]

One thing I noticed on the layout, the fets have the middle leg bent forward, whereas the small bipolars have the middle leg bent to the back.
Since the fets had all been changed out by someone else previously, make sure they are not installed 'backwards'.

[Paul Marossy]

What are the voltages when Stereo/Mono is switched with the fets removed?
The datasheet says -1V to -5V is the pinch-off voltage.

You cannot view this attachment.

I will check that, but let me ask a question first: Is that stereo/mono switch on my PCB drawing wired correctly? It looks like it came from the factory that way but it seems odd to me that in stereo mode it's grounding some of the pins on U11 but in mono mode it's connecting only to +15.6V. I don't quite understand how all that works to be honest. I guess the LFOs are basically being put out of circuit by applying the +15.6V?

I did replace the 6800uF caps in both power supplies tonight, so at least now there is sound in stereo mode but still no tremolo. In mono mode I can hear the LFO bleeding thru, makes a bit of a whirring buzz sound that changes with the speed control. I also got the octave filter to work when I tried different FETs for Q7 & Q8 (2N5952s) but the effect is always on and the footswitch does nothing. Also was not reliable, sometimes on certain notes it would get waaaay louder. Need to revisit that now that the PS filter caps have been replaced.

After the big filter caps in PS were replaced the sound was generally a lot better, more responsive bass. I still have a 120Hz hum/buzz, not horrible but it's there. Wondering if isolated jacks would help with that. Seems like there may also be some ground loop issues going on.

Sure is a lot of strange problems on this amp lol, but it sounds pretty good with the reverb on, although that is kinda noisy. I'm not sure what the purpose of the FET is on the reverb recovery circuit... is that supposed to be like a noise gate or something? I currently have a 2N952 there, I think it sounds better. Will have to play around with that some more.

[Paul Marossy]

Back in post #2 you noted the cut leads on U12B. This is important. Pins 8 and 13 will
interfere with the FET switching. make sure they are cut or bent out of the socket.

Also I would remove or clip R50 and R51

Noted! I will undo my fix for U12B and see what results that yields. That U12B area had me confused for a while. Because output of U12B says "to power amp", I thought maybe it was supposed to be connected to something, and someone had been in there before me messing with something. In any case, I had come to the conclusion that it had been left unused. I wonder why on earth did they do it like that? Perhaps they didn't know any better? 

EDIT 12/28/2024: I lifted R50 & R51 and made U12B as it was when I acquired the amp. None of that helped. :-(

[Paul Marossy]

Yeah that was my thinking but it does not work at all with 2N5457, J201, or the J113s that are in it now! It works sort of with 2N5952s but it's still not right.

That was my thinking, any FET ought to work but that has not been my experience in this case.


J113s are what is in it now and it does not work!

When I started on this project I had no reason to suspect the FETs but thru the process of elimination I have come to suspect them. I'm getting all the right voltages. Already have looked for broken tracks, etc. but maybe I'll do that for a fourth time just for the heck of it. :-D

Attached is the data sheet I have for PN4093. 2N5952 seems closest but still not quite right. I have not found anything that is a direct replacement for a 2N4093.

Just thinking aloud: then maybe you do not have a FET problem but something else.

These FETs do not amplify or create any effect or anything, they just let Audio through or not.

Here is one which is easy to test functionally, Q2.
Switch Distortion  ON/OFF.
Does that work?
IF so, Q2 is working.

Even more, I would inject 100mV 1kHz at input and check for signal level/presence/absence on nodes C21/R19 and on U5 pin 6




If not: does voltage change on nodes R20/D10 or R20/R21?
I would expect 0V (Dist ON) or -15V (Dist OFF)

What dmeek already suggested.

Notice FET is switched ON by merely applying 0V to gate and OFF by applying -15V to it, which is WAY higher than needed to turn ANY FET OFF

I reluctantly suggested shotgunning at first but clearly we are getting back into actual troubleshooting.

Maybe because it works?

One of the reasons why I thought it could be the FETs is because the Vgs off number on the 2N4093s is pretty high compared to others I have on hand. Curiously, I get worse results with the 2N4093s that I bought the other day. There definitely seems to be a connection with what FETs it likes but I don't understand exactly what it is. Seems like some of that was connected with the two power supplies. Some things have improved since I replaced the 6800uF caps in both power supplies (which needed to be done anyway), like at least now I don't have no sound at all when switching to stereo mode - there is sound just no tremolo. I think I know why that may be per one of dmeek's posts, so I'll check into that.

Distortion circuit has always worked with no issues. Reverb driver is finicky about what opamp is in it. Original LM301 in U7 position seems to be dead, worked with an LM741 but not other opamps. Don't understand what is going on with that. Sometimes there is loud pop when turning off amp with reverb level up high, other times not. Not sure what the purpose of the FET on the reverb recovery circuit is... what is that supposed to do exactly? 

[Paul Marossy]

One thing I noticed on the layout, the fets have the middle leg bent forward, whereas the small bipolars have the middle leg bent to the back.
Since the fets had all been changed out by someone else previously, make sure they are not installed 'backwards'.

I'm not 100% sure on whether or not someone changed them all. The soldering looked original on all of those FETs but there was what appeared to be a date code on some of the J113s that looked like 1982 to me. If that was the case, with the amp being made in 1979 and the person I bought it from saying that he bought the amp in 1980, that would suggest that they were changed out by someone some time after he bought the amp. I could be mistaken, maybe that's not a date code and the "82" is coincidental. Were J113s even being manufactured back in 1979?

As far as those legs go, I'm 95% sure this PCB was designed by hand, so those pads probably ended up being wherever they could get them on the board. It almost looks like it was designed around a TO-18 package to me.

[J M Fahey]

Maybe.

Way back then FETs were a newfangled premium component and deserved expensive metallic cases.

Today?

Apparently not worth even a humble plastic TO92 case because these are being discontinued, everything is shifting towards SMT cases, SOT23 and similar.

In any case, what matters is pinout; if you need to do some lead bending, so be it.

Different pinch-off voltage?
Sure, but in any case switching line applies 14-15V to gates, way above any pinch-off voltage, so they are forced to work.

I suggested testing the Distortion FET switcher, basically to *measure* voltages around it and take that as a reference.

All others must work about the same.
Best FET in the World will not switch if proper voltages are not applied, that why I insist you measure them.

J111/12/13 are modern stuff, straight designed fr switching, almost nobody makes FET preamps any more.
Doubt they were around in the late 70s

[Paul Marossy]

Maybe.

Way back then FETs were a newfangled premium component and deserved expensive metallic cases.

Today?

Apparently not worth even a humble plastic TO92 case because these are being discontinued, everything is shifting towards SMT cases, SOT23 and similar.

In any case, what matters is pinout; if you need to do some lead bending, so be it.

Different pinch-off voltage?
Sure, but in any case switching line applies 14-15V to gates, way above any pinch-off voltage, so they are forced to work.

I suggested testing the Distortion FET switcher, basically to *measure* voltages around it and take that as a reference.

All others must work about the same.
Best FET in the World will not switch if proper voltages are not applied, that why I insist you measure them.

J111/12/13 are modern stuff, straight designed fr switching, almost nobody makes FET preamps any more.
Doubt they were around in the late 70s

I measured Q5 & Q6 in circuit this morning. Here is what I measured with 2N5952s (with J113s no sound will pass in "stereo" mode):

In "mono" mode-- 
Q5 G= 0.01V, S= 0.01V, D= 0V
Q6 G= 0.01V, S= 0.01V, D= 0v

In "stereo" mode--
Q5 G= -5.5V, S= -5.5V, D= -4.85V
Q6 G= -5.84V, S= -5.84V, D= -4.84V

On that stereo/mono switch, it is directly connecting +15.79V (red wire, point C on my PCB layout) and -6.07V (brown wire, point D on my PCB layout). The switch connects these two together when in "mono" mode. In "stereo" mode, these two are disconnected. I assume this is what forces the FETs to become switches? It does not seem to work very well as when you flip that switch it is VERY noticeable - not a huge pop but the speakers are moving a lot of air. Don't know how to describe the sound exactly. Like SCHARMF? Anyway, I imagine that should not be happening when everything is functioning properly. Is that switch really supposed to be connecting those two voltages together like that?! Something seems very odd about that to me... but I also don't work on things with bipolar power supplies much.

I've been looking for leaky caps too. I checked all of the 1uF caps on the main PCB, they all seem OK to me. Do you think it's possible that I have some leaky diodes? I checked them all with multi-meter, and none of them are shorted or open. I read somewhere that someone had a leaky diode in a Polytone amp, but I don't recall all of the details. I have never heard of that before. These are all 1N4003s according to the schematics. They certainly look like they are but I don't see any markings on them. just a pink-ish band on the cathode end. I did not test any of these diodes while the power was on however.
       

[Paul Marossy]

Here is an updated PCB layout. Corrected a few minor errors and/or omissions. I also drew up the power supply PCB layout since none exists online and I wanted to have one.

Looking at the LM391 data sheet, it says on page 7 that "the signal ground must return to the power supply alone, as must the output load ground. All other grounds -- bypass, output R-C, protection, etc., can tie together and then return to supply. This ground is called high frequency ground." I'm not sure what that means exactly or if that power supply is doing the grounds the way the the data sheet shows on page 12.  I thought the speaker grounds and the rest of the signal grounds were supposed to be isolated from each other... is that just with tube amps?

Just trying to identify what sounds to me like a ground loop. Wondering if the input jacks, which are not isolated from the chassis, are picking up noise and amplifying it. Could also be the power supplies are a little mis-biased. Need to check on the bias after I've rounded up the appropriate resistors to check the crossover distortion.

[Paul Marossy]

OK so I have made some progress: I put R50 & R51 back in circuit. The I tried a NOS IC chip for U9 & U11 one more time, but this time put the original J113s back in and now the harmonic divider works! It's a very interesting effect... it's not like anything I've ever heard. Closest thing would be the Bigfoot FX "Magnavibe" DIY vibrato I built like 10 years ago now, but in stereo. That DIY guitar effect actually IS a vibrato, not a tremolo, and I verified that via oscilloscope. It's here if you want to check it out: https://revolutiondeux.blogspot.com/2012/01/bigfoot-fx-magnavibe.html
Anyway, I really like this harmonic divider thing. Very different but once you learn how to use it, it's pretty cool.
BTW, it did NOT work with the 2N5952s and the NOS IC chips - it would not pass any sound and only got the very loud thumping. So despite the notion that any old FET should work, this circuit is very picky about what FET it will accept.

So now my remaining major issue is getting that octave divider thing to work. That is still not functional. Another thing that seems to be FET related as it doesn't work with the J113s in it now but sort of works with 2N5952s, but they seem to always be biased on and the footswitch doesn't do anything. I guess I'll eventually figure that out too. It's gotta be the FETs associated with that circuit (or maybe the diodes?) because U4 is verified to be working.

This morning I verified that the input jacks do NOT want to be isolated from the chassis. Did some lead dress stuff - got the long length of twisted AC wire for the four pilot lights separated from the parallel shielded reverb cabling, and that reduced the noise level somewhat and also eliminated a strange oscillation problem I had when the volume was turned up past a certain level. I had a hunch on that one, based on some past experiences.

Rev/Dist & Harm Div footswitches are confirmed to be working. I've been trying to find a picture of the footswitches/pedals that would have been available for this thing but that seems to be 10x as rare as this amp is. I think it would have had LED indicators on it but I am not 100% sure. I may make some sort of footswitch arrangement for it down the road.

Anyway, happy to have finally made some progress on this thing. It had multiple issues to figure out! I was about to give up on it today. 

[g1]

So despite the notion that any old FET should work, this circuit is very picky about what FET it will accept.

Not sure where that was said.  'Any old' would still require it be n-channel jfet with same pinout.
2N5952 has a different pinout than the PN4093 or J113, so it will not work if installed with the same orientation.