DIYstompboxes.com

Hello good people of the audio-electronic kingdom. 

Background: This is my first-time post here.  My problem child is the AionFX version of the Sunn Beta (Lead version).  It is the 9th pedal that I have built (including another Sunn Beta clone by PCB Guitar Mania).  I am currently at a beginner high-intermediate low level of understanding of pedal electronics. The intent behind the project is that the Beta pedal will be one of two boards running as a modular, pedal-format preamp running into a 100w Class D power amp in emulation of the original Sunn Beta.

The build was done with the aionfx pcb, as well as a mix of components that I had left over from other projects, Aion's Mouser BOM, and orders from Stompboxparts.com.  To my knowledge, I built it exactly to stock specs without substitutions, and tried to order high-quality components when I could choose them.

Here is the documentation with schematic: https://aionfx.com/app/files/docs/beta_preamp_documentation.pdf
I performed the recommended jumpering of IC2 pins 9 and 11 to the ground terminal of the mids pot.  Off-board wiring done with 24AWG solidcore wire (I'm trying to be more tidy), according to AionFX diagrams.

My gutshots, PCB unmoored from the enclosure for probing (please pardon my messy wiring): https://photos.app.goo.gl/JvC4RkjVkpcjhq3fA

The problem: I have run into something odd that I am at a loss to understand, explain, or find the source of, and I bring this to you because 1) I would like my pedal to run correctly throughout the range of possible settings, obviously, and 2) because I would like to understand why it's happening. 

The guitar has signal and effect as expected.  But, with the "level" pot at unity gain (~9:00) and the "drive" at about 60%, the last 10-20% of travel in the treble pot yields a high-pitched whine/squeal that becomes more pronounced when the drive pot is turned up.  Cranking the level pot brings the beginning of the noise back.  With level at 100%, the noise starts at ~40-50% of travel in the treble pot.  At lower "Drive" pot levels, gradually increasing the treble pot at the onset of the noise results in sounds or modulations that remind me of changes in radio frequency.

I searched this forum, contacted aionfx, facebook pedal building forums, and reddit for any additional info, the only thing I found is a user on facebook who just built the same version board (albeit jumpered to ground at pin 7 of the CD4069) and reports the same issue.

Troubleshooting per the forum guide:

With the audio probe, I could detect the whine/squeal noise at audible levels (but not nearly as loud as the level on output) beginning at R24, a 22k resistor, which seems to be correctly soldered and of the correct value.  The noise was also present in the bass pot, treble pot, of course, and in R29, but not in R23 (unless it is very very faint), or in c15 or or C19.  Noise was absent in all RC4558 signal out pins (1 and 7) except for IC 6, in which it is extremely present.  I also couldn't detect anything out of the CD4069. 

Let's get to the numbers.  Measured with a Mestek DM100C multimeter, pedal on and connected to 9V OneSpot wall wart measuring at 9.426 v DC at the input jack and at entry to the board:

Q1 RC4558P
1: -0.352
2: 0
3: 0
4: -8.667
5: 0
6: 0
7: -0.16
8: 9.135

Q2 CD4069UBE
1: 3.626
2: 3.631
3: 3.610
4: 3.558
5: 3.602
6: 3.753
7: 0.001
8: 7.841
9: 0 (jumpered to 11 and to ground)
10:  7.84
11: 0 (jumpered to 9 and to ground)
12: 3.624
13: 3.617
14: 7.843

IC3 RC4558P
1: 0
2: 0
3: 0
4: -8.66
5: 0
6: 0
7: -0.018
8: 9.127

IC4 RC4558P
1: 0
2: 0
3: 0
4: -8.665
5: 0
6: 0.002
7: -0.015
8: 9.130

IC5 LT1054CP
1: 9.129
2: 4.923
3: 0.001
4: -3.957
5: -8.663
6: 2.571
7: 1.431
8: 9.127

IC6 RC4558P
1: 0.001
2: 0
3: 0
4: -8.665
5: 0
6: 0.006
7: 0.618
8: 9.126

Z1 1N4742A
A: 9.131
K: 0

I'll be happy to give any additional details regarding any more of the weirdness of this phenomenon.

hi, welcome,

I am probably just as knowledgable in electronics as you are. But I have 1 or 2 ideas you can try :

From instructions, p.7 :

"Thanks to Brad from Nerd Knuckle Effects for diagnosing the issue and running noise tests to determine
the optimal way to disable the CMOS gates. At max volume & gain with all tone controls at noon, the
noise was measured at -36.1dBV when gates were tied to ground and -34.1dBV when tied to VB, so we
opted for ground in version 1.1."

I guess your noise is the one mentioned above, probably coming from IC2.

I would try an other CD4069 in IC2, , the one from your Guitar Mania build for exemple. First without the v.1 fix, and with the fix. It is said that the chip will get hot, but won't overheat without the fix, let's see if the noise remains...

If it doesn't change anything, I would try to connect gates to VB, just in case : C35 + side, instead of the mid pot. On schematics, IC2 Pin 14 seems connected to VB too, but I wouldn't touch it, to avoid more unexpected issues from IC2.

Hey, thank you much!  Does FR=le beau pays?

OK, will try your suggestion, I have an extra CD4069, and at least I can see if it's a different noise.  I forgot to say, I checked CD4069 and couldn't find the whine either.

Report: ohhhh, no, no, no.  Without the fix, that was just a festival of craptastic fizz.  And still had the squeal to boot.

Just so I don't do dumb stuff, can you elaborate on what you mean by "C35 [capacitor 35] + side"?

Just as a general update of things I tried, I replaced R24 on the off chance it was defective, and it didn't do a thing.

Quote from: comradehoser on October 08, 2022, 08:04:29 PM
Hey, thank you much!  Does FR=le beau pays?

yes, by the way, have you ever tasted natural white wine from Anjou ? good stuff, "natural" means no chemicals. They're using what they call "noble rot"... that's why it's so good...

Quote from: comradehoser on October 08, 2022, 08:04:29 PM
Just so I don't do dumb stuff, can you elaborate on what you mean by "C35 [capacitor 35] + side"?

Instructions say you can connect the CMOS chip either on ground, or to VB. If I understood correctly, Aion suggests to choose ground because they noticed less noise on their build, but maybe VB will work better on yours...
Electrolytic capacitors are polarized, positive side has a + sign on the pcb. I suggest to try connecting IC2 pins 9 and 11 to C35 square pad :

(https://i.postimg.cc/S2kMZ6wt/Screenshot-2022-10-09-at-03-58-15-Beta-Preamp-Aion-FX-beta-preamp-documentation-pdf.png) (https://postimg.cc/S2kMZ6wt)

That way these pins will be connected to VB instead of ground.

But i am assuming the noise comes from IC2,  and maybe this is wrong.

From your pictures, I'd say your soldering skills are very good, but you never know...maybe you have a bad solder joint somewhere, sometimes they look good but they aren't.

Did you go on the hunt for shorts ? And try to reflow all suspicious soldering pads ? the pads connecting the main board to the footswitch are looking a bit suspicious to me, for exemple, but that's probably not the issue. I would reflow every pads, just to be sure, and clean the tiny spaces between pads on every IC's sockets, going quickly with the iron to avoid overheating the board.

Maybe a long shot but I see you have some of those yellow caps.

I do too and someone on this forum told me once they were "microphonics" or something, meaning they are noisy, not to be put on the signal path.
But i am not sure if yours are the same. Can you show their specifications ? like where you got them and what kind of caps are they ?

Edit : remember to take every heat-sensitive parts off the board if you use your soldering iron near them (only ICs in your case).

Edit 2 :

Quote from: comradehoser on October 08, 2022, 08:04:29 PM
ohhhh, no, no, no.  Without the fix, that was just a festival of craptastic fizz.  And still had the squeal to boot.

Well, from what I can understand p.7, the v.1.0 version is supposed to work without any fix, the fix only prevents the chip from getting a bit hot, and that's all.

Non, je ne peux pas dire que j'ai goute le vin Angevin--mais c'est aussi vrai que je suis plutot biere que vin.  Par contre, les petits vins et alcools esoteriques, j'adore explorer--tel le vin jaune de Franche-Comte. (desole, pas d'accents).  Pour le vin de pourriture noble, le Sauternes est assez spectaculaire, aussi.  Vous venez d'Anjou, donc?  Ma famille est du cote au nord de Cognac, plutot la Charente Limousine, si vous connaissez le pays la-bas (bravo le pineau).

OK, back to English for the electronical benefit of all.

Thanks for the idea on soldering to VB.  I will check it out, as it's a very easy thing to test anyway.

Yes, I definitely reflowed the board and I try to be very meticulous about things as I'm soldering and I will check for possible shorts and poor joints after every component.  The only other thing left is my off-board wiring--this doesn't sound at all like ground-related noises I am familiar with, but who knows.  I will try to post a recording, if I can.

I'm not sure if the noise is from IC2, as I remember checking the outs on the CD 4069, and couldn't hear it.  I just don't understand why it would start at R24 as it doesn't really seem to be a particularly complicated place where things are coming in to mess up the signal or amplify lurking noise frequencies.  Something coming from the charge pump, maybe?

The yellow caps are mostly 100nf MLCC or smaller pf value caps obtained from Mouser and specified in Aion's layout and BOM.  I don't know if they are particularly microphonic and have never heard/observed that, but what do I know?

I don't know about the precise reasons for the jumpering fix, but I think they were experiencing noise issues, and IC2 with all pins in socket was pretty horrible sounding--maybe it was a bad IC? but I think it's more probable that it was the problem they were trying to fix.  It also didn't heat up noticeably, so maybe that's a clue, I don't know. 

also welcome.

most all IC's will upset when any/some/all their pins are connected to some voltage outside the limits defined by their supply pin conections. in the case of the 4069, ground and VB. it doesn't really matter what the value of VB is, within the chip max values, obvs. and THE RULE of cmos ic's is ALWAYS always terminate unused input pins to a logic level. your logic levels here are 0/low/ground and 1/high/V+ supply, in this case VB. either of those must be safe to connect to, by definition. so, really, don't be traipsing that terminating wire all over the board, take it to the only 2 safe pins on the board, either pin 7 or pin 14 of the 4069.

because you have 4 of the invertors in that package biased to their linear regions in order to do amplifyings, the IC will draw more current than if it was doing ordinary logic 1 and 0's, so it can be expected to get warm even when working correctly. measure the voltage across R34 [from end to end, polarity not important] and apply Ohm's law for the current drawn.

it might be an interesting exercise to audio probe all your IC supply pins, because you are expecting clean DC on all of them, so any noise might be a useful pointer to your problem.  check very carefully the two diodes at pin 7 IC6, make sure they are pointing correct ways. as for that "LEVEL" pot configuration, I've not seen it before, dunno quite how it works. but it looks suspiciously close to your R24.

Roger, wilco on the three things and will report back.

I suspected the charge pump because the whine, but I subbed it out with the Guitar Mania pedal's known functional LT1054, and I have the same problem.  Also subbed all of the RC4558s and the CD4069 to no effect. 

Report:  soldered terminating wire to pin 7 per aion recommendation.

voltage across R34 was 1.177.

Audio probed all of the RC4558 voltage supply pins (8 and 4).  Quiet.  Verified that no whine present in CD4069.  The only place where I could find a similar noise in amplitude and similar frequency was pin 7 of the charge pump, LT1054.  I can't tell where it goes in the schematic.  I assumed it was a dead terminal.

In unscientific fashion (because I did a bunch of stuff and didn't test in between), I also twisted the ground and 9V supply wires, reflowed all of my grounds and any solder that looked remotely funky.  When I had the whole circuit out of the enclosure for audio probing, I thought I had something reasonable: no squealing except at everything-dimed levels and when the guitar was muted. 

But when I remounted in the enclosure the result was.... the same.  So maybe to do with offboard wiring or some interaction with the enclosure?

The pedal is very temperamental.  Sometimes the squeal is much louder; at one point, I had nothing but hiss.  Obviously it's due to poking around and moving stuff, but there is something I am not grasping about this because I can't make sense of what I'm observing.

Quote from: comradehoser on October 10, 2022, 01:06:44 AM
When I had the whole circuit out of the enclosure for audio probing, I thought I had something reasonable: no squealing except at everything-dimed levels and when the guitar was muted. 

But when I remounted in the enclosure the result was.... the same.  So maybe to do with offboard wiring or some interaction with the enclosure?

I can't really see on your pics...is there some isolating material on the pots to prevent them from touching the pcb and the jacks ?

If not, some plastic tape or even a piece of cardboard will do the trick.

If that's not the issue, can you post a pic with everything in the enclosure ?

Edit : looks like you made a mistake drilling the holes for the jacks, I guess you forgot about the pots when you did the first ones  (;D).

Are we sure there's enough room in there for the jacks, without touching the back of the enclosure  ?

Yes, the pots have dustcovers on them.

You are exactly correct on the drilling.  I was like "oh, I'll drill the jack holes offset so that the 9v cable has more room."  I actually did it to 4 pedal enclosures  :icon_evil:

The jacks sit slightly proud, but I rotated them so that the grounds, if anything, will make contact.

Have electrical tape on the back of the enclosure.  Pedal squeals with or without it on.

Here are pics (on and squealing): https://photos.app.goo.gl/CMDqNndKRgyCQ44H9

If it is of any importance, I'm using a little Orange Crush 12L as my test amp.

[edit: decided to change the test parameters:]
Squeal issues:
Single coil different/less than humbucker/two single coils selected

Orange 12L Crush (solid state): Present

Silverface Twin Reverb (tube): Present

Sunn Sonaro Bass head (tube) and 2x15 cab: Present

Nux MG-30 multi-effects pedal with cab-only IR and with/without preamp: You can detect it as background/residual noise, but out-and-out squeal/whine NOT present.  However, with the squeal muted a bit, it is a noisy AF pedal.  Lots of ground noise that is affected by touching the enclosure.

Could it be the presence of another preamp that is the issue?  Unfortunately, none of my IRL amps have a bypass effects loop to test the idea.

Quote from: comradehoser on October 10, 2022, 12:17:42 PM
Lots of ground noise that is affected by touching the enclosure.
(...)
Could it be the presence of another preamp that is the issue? 

I doubt it 's related to an other preamp, you can test it by connecting headphones to your build, instead of an amp.

The fact that the enclosure is sensitive like this, points toward a ground issue, or a short somewhere.

You can do a continuity test, with your DMM, on all your grounds locations, and check if everything is ok.

Probably nothing, but on your second pic, looks like one side of CD4069 isn't correctly inserted in the socket.

I guess you already reflowed all the pots pads, and checked carefully the R24 aera for shorts.

(https://i.postimg.cc/Yh40DNYS/Screenshot-2022-10-10-at-20-54-59-Beta-Preamp-Aion-FX-beta-preamp-documentation-pdf.png) (https://postimg.cc/Yh40DNYS)
As your audio-probe test indicates something happening at R24, I would give a closer attention to this location, check all the grounds with a continuity test, reflow all pads. Still on continuity mode, check all connections  between components according to the schematics, etc.

On the other hand you noticed that the circuit is "temperamental". Once everything is inside and screwed to the enclosure, if you shake it a bit you can hear some variations in the sounds or in the squealing ?

This would indicate a loose wire or some unstable connection from a bad solder joint, could even be some piece of hair or some dust somewhere on the board. You can clean the pcb with a dry toothbrush.

An other test could be to feed the circuit with a battery, and not with your usual power supply. To check if there is a bad interaction there.

Salut l'ami!

Yup, forgot to say that I did check every ground point in the circuit and in the wiring.  Everything checked out. Is it possibly the fact that I'm using solidcore vs. stranded wire?  Just shook it about while in squeal mode.  No audible changes.

Can you tell me what you see for CD4069?  It's in there pretty much as far as it can go.

I did indeed reflow every solder point, including pots.  The only thing I didn't resolder were the in and out wires.

That's a good idea on the battery.  Will do. 

The temperamentality is not readily explainable to me.  I just don't get why sometimes it works better than other times.  It's been a bit random and all over the place because I am not getting the something that is causing this.  Poking in the charge pump with the audio probe and inadvertently bridging pins reliably shut the pedal down and it would pop back up after a while. But other than that, I don't know.

I feel very much like your signature quotation, there.

Quote from: comradehoser on October 10, 2022, 06:09:30 PM
I feel very much like your signature quotation, there.

Indeed ? i thought it was a bit too much. I ll keep it for a while then.

Quote from: comradehoser on October 10, 2022, 06:09:30 PM
Can you tell me what you see for CD4069?  It's in there pretty much as far as it can go.

(https://i.postimg.cc/TLDHVpNk/IMG-20221010-121442060.jpg) (https://postimg.cc/TLDHVpNk)
Zooming in on the CD4069, I thought... Probably just an optic illusion...sorry.

On Aion's Spectron, I had one weak trace, the current couldn't pass between 2 components, and I had to solder a jumper to get a connection. Then it worked.

My point is to check that the parts are really connected as they should, according to schems. Especially in the R24 aera. For exemple, R23 should ring with R24, BASS lug 3 should ring with R29, and R29 with TREBLE lug 3, and so on...

I highly recommend building the Spectron, by the way.

LT1054 may be a suspect, but I wouldn't touch it with the DMM, unless you know exacly what you are doing. I'd be more confortable taking it off and checking the socket's connections, power supply unplugged, looking for shorts or bad connections.

Edit : You can try your charge pump in your guitar mania build to make sure it's working.

Alright, I will check continuity.  But it seems like this isn't a matter of go/no go continuity/no continuity, but something in the degree of connection or what the parts are doing too much or too little of.   

I actually did the opposite and swapped in the 1054 from the GuitarMania pedal.  As usual, no effect.

Maybe I should just chalk it down to it being a relatively high gain pedal and use it in the parameters it's comfortable playing within.

Still, it would be nice to have a 100% functional pedal and understand what is going on.

I will try to post audio if I can.

An audio sample might help more experienced forumites to find the issue.

There's also the possibility that you made a mistake populating the board.

In such cases, downloading a free software can help you read the resistors markings :

http://www.atlence.com/resistor/index-fr.html

Edit:

Quote from: comradehoser on October 10, 2022, 01:06:44 AM
When I had the whole circuit out of the enclosure for audio probing, I thought I had something reasonable: no squealing except at everything-dimed levels and when the guitar was muted. 

But when I remounted in the enclosure the result was.... the same.  So maybe to do with offboard wiring or some interaction with the enclosure?

If there aren't any noises anymore out of the box, it should be a very good sign that the circuit is fine. Then it's surely a simple mistake when you install the circuit inside it...

For exemple, in my experience, pots can behave oddly when there is too much tension after being screwed onto the enclosure. Better to solder them last, only once they are already screwed inside the box, thus avoiding unnecessary tension on their lugs and on the board.

May I suggest to slowly fit the circuit in the box while listening to it, in order to hear when the squealing appears ?

here's something - pull IC6 from its socket, power on and audio probe pin 6 for squeel. vary the level pot, too.

duck_arse: will do.  Do you mean to power and probe pin 6 the chip independently of the circuit, or probe the pin 6 socket?

eh la bas ma: I forgot to say I did check all of the component markings, including the resistors--everything checks out so far.  I am wondering how you determine the direction of reading other than having the brown tolerance band at right?

on the fitting and pots--that is an interesting idea and I will try it as well as seeing if I can reproduce the lower noise out of the box.  I have to add that in my audio probing I actually clipped the grounds to the enclosure.  Maybe it is the pot?

I will try to record and post audio later tonight if I can figure out how to do so.

What's going on with the 4069? I see a jumper from the +power pin 14 to pins 2 and 4. Those pins 2 and 4 are outputs... are they not? Wiring outputs to any supply is not a good idea, it should only be done for inputs.
Ah, but... is the 4069 in backwards? It's just that I see what looks like the pin1 ID notch at the top of the picture but I think in the build doc it should be at the bottom.

Aaahh, anotherjim--well, unfortunately if it were only so simple!

It's just poor photography on my part.  Here it is in the middle of disassembly--what you are seeing as the orientation notch is just a molding dimple, unfortunately: https://photos.app.goo.gl/GW6aGUQ5jPUPLP2t9

Further developments:

Resolving to be more scientific, I took the board out and audio probed the signal in the squeal range as I unbolted each pot and jack. When the input jack rested on the enclosure (grounding itself) and I touched the tip flange of the jack, the squeal all but substantially disappeared.  Touching the output jack, squeal; touching both jacks, lower squeal, but squeal. 

And then a while later, the squeal happened regardless of what I touched, and was everywhere--all ICs.  I think this is called gaslighting or witchcraft or possession. The pedal is just trying to drive me crazy.

Partly a grounding issue, maybe?  I am actually thinking the ground noise from touching the tip input partially cancelled the squeal, as it tends to be displaced by signal:

The squeal is present and strongest when the guitar is muted, or with no input at all, i.e. no cable plugged into input at all.

And for duck_arse, I pulled IC6 and at that point, no squeal in pin 6--or anywhere else, actually, so maybe this is gain-related(?)

I also tried replacing IC6 with another RC4558 and a TL072--no change.  As soon as I populate IC6, we are back in squealandia.

Well, so it does sound like it's a stability problem with some stray feedback going on from output to input.
Listing some things...
Sometimes it helps to use screened hookup wire to the in and out jacks or at least just for the input jack.

I don't see a schematic for the switch PCB. I think Aion get it right and do use the middle contacts of the 3PDT for the LED indication but if they haven't it leaves input and output on directly adjacent contacts and that makes a feedback path via capacitive coupling.

I'm not sure I like the output stage, I think it might suffer from a phase shift with that Level control feedback taken from the final output. Difficult to do with a PCB build, but what happens if you take the Level pot out? Also, the feedback resistor R32 has no accompanying feedback capacitor (notice many of the opamps in the scheme do). Try 100pF across that R32.

Ok, will a lay a 100pf cap across R32 to test, will determine the signal routing in the 3pdt, and in the last resort, will take out the level pot and report back.

For the shielded wire, would wrapping the input wire I have in aluminium foil be sufficient to improvise a quick shielding test?

comradehoser - pulling IC6 was to prove IC6 was the problem. if the rest of the circuit is squel free with nothing in that spot, the fingers start pointing to that last section. couple that with the weird connections around the stage, and we get a strong candidate for the cause.

AJ - have you ever seen that method of "level" control before? can it work like that? what's the idea, any idea?

and, as for resistor ID, yes exactly, you know the tolerance band is brown, so you start there. then you practise on thousands and thousands of E24 E12 E12 E12 E12 E12 series values till you know the legal colour combinations. most people find 12k 1% the most troublesome value to read.


[edit :] right on time! https://www.diystompboxes.com/smfforum/index.php?topic=129725.msg1254679#msg1254679

[edit edit :] I always manage to put E24 when I mean E12. learn the E12 sequence, check the E24's when they appear, just hope they don't.

That Level control we may have seen before just looks different and it's an attenuator really. Only good for an inverting amp. The gain reduction goes down to Rf/Rin just the same I think. But, the feedback control can come from either before or after the output resistor R33 and always before C20. I'd pick before R33 because as it is, the protection from R33 and the diodes could be missed by the path to the -input via the pot.

As for screening, for sure you can wrap a wire. A screen only needs grounding in one place - at the jack will do. If you find some screened cable (some steal from redundant/cheap hi-fi stereo phono leads, the kind that is twinned for stereo) cut the screen back at the non-grounded end and insulate.

Report from the squeal zone:

Shielding of input and output wires with copper foil grounded to side of enclosure: no change

82 pf capacitor (closest to 100pf I had)  "across" R32, which I assume means lead:lead contact between the cap and resistor: slightly lower frequency of squeal.  I also tried putting one end of the cap on pin 6 and then 7 of IC6 with the other end on first one end of the resistor and then the other.  That seemed to be the only direct line to R32 in the schematic.  That resulted in the signal disappearing.

Removal of the "Level" 100kA potentiometer: Squeal.  If anything, it was more all over the circuit.  I could make it squeal with all pots.  Once the signal reached a certain threshold in drive, bass, mid, and treble or combination of them, it freaked out.  More squealy when the pots weren't in the enclosure.  Squeal when I touched the input jack tip flange this time.

I guess I just have to ask if this is the point at which you just scrap a pedal.  If it's not, how do you determine it?  I feel bad about keeping on asking people's attention on this problem, which you all have been very generous with and for which I'm sincerely grateful.   

I suppose I just want to understand and have a pedal which is dependable through all of its range if I'm going to be using it in front of an amp and cab, and when it's not squealing it sounds great!  But again, I'm not seeing the underlying pattern at all other than removal of IC6 which I'm guessing is lowering the signal so that it can't attain that squeal threshold.

I don't know what it is about the Beta pedal--my last GuitarMania build was also difficult.

Quote from: comradehoser on October 12, 2022, 10:20:21 PM
82 pf capacitor (closest to 100pf I had)  "across" R32, which I assume means lead:lead contact between the cap and resistor: slightly lower frequency of squeal.  I also tried putting one end of the cap on pin 6 and then 7 of IC6 with the other end on first one end of the resistor and then the other.  That seemed to be the only direct line to R32 in the schematic.  That resulted in the signal disappearing.

Whatever combination between cap, pin 6, pin 7 and R32 leg(s) shouldn't disappear the signal.. :icon_wink:

P.S.
Double check connections between pin 6, R32 and Level pot lug 1 & 3..
(it seems to me like there is an amount of positive feedback either around IC6B or from OUT to IN..)

Will do.

Should I reinstall the level pot? Or just measure to the port?

Rene

Quote from: comradehoser on October 12, 2022, 10:20:21 PM
I guess I just have to ask if this is the point at which you just scrap a pedal.  If it's not, how do you determine it?  I feel bad about keeping on asking people's attention on this problem, which you all have been very generous with and for which I'm sincerely grateful.   

don't walk away, Rene. if you have a breadboard, you could at least test the circuit with a new config for the IC6 section lashed in. if that doesn't squeel, you could maybe build a little daughter board and hang that on - or, at that point, turf the build.

Don't short pins 6 & 7 together in this case as that will be the same effect at the Level pot turned fully down. We want the trial cap across pins 6 & 7 which is the same as putting it across the ends of R32. The near end of R32 probably goes to pin6 so the other end of the cap to pin7 should be the same thing assuming the cap leads are too short to bridge R32.

A wild outside chance, those turned pin IC sockets have been known to fail to contact the inserted pin. If an opamps +input pin "floats" out of contact, it can pick up stray noise so make sure those pins (3 & 5 for most of them) are connected to GND where shown on the schematic.

Quote...don't walk away, Rene.

See what he did there?

I don't want to Bragg, but yes I did see that.

I did test all grounds, but will retest the IC6 section again.

The 82pf cap has ridiculously long leads, so bridging on top of the resistor leads is not a problem.  Will try again as well as bridging pins 6&7.

I am using the snappable pin sockets from Stompboxparts.com. I suppose I can check for continuity (?)  They are in there pretty well.  I suppose I could put in a bit of contact cleaner.

What should I be looking for in feedback at ins and outs?

Thank you for the words of reassurance, y'all, I guess I am just getting a little bit down about this because I can't see the headway.

Aion did state that there have been a few cases of this, but that they are rare--so this would suggest user error or component problems... right?

It may be an optical illusion but is that 4069 180 degrees out... as in, backwards?

All comments below based on observations without the level pot installed

Antonis: i have straight continuity between pin 7 and the right lead of R32, pin 6 and the left lead of R32 and pad 1 of the level pot ( rightmost pad read from the component side). Nothing straight through to pad 3 as it has to go through r33 and c20

Guitarhacknoise: it's a molding dimple that looks like the orientation notch because I'm a bad photographer

Another Jim: cap bridging at R32 and pin 6&7 changes the pitch of squeal, but other than that, not much.  Ic Pins and socket pins have continuity.

After checking continuity, plugged pedal in for cap bridge check, and got nothing but hiss.  Pulled IC6, got signal; put it back in, ye olde squeal. 

Maybe there is a solder bridge under the socket pins that i can't see?  I am at the point where I am thinking to desolder and resolder the whole IC6 section

A dumb, almost scattergun, bughunting technique I use is to pull the chip (if I can) and with constant reference to the schematic, continuity test (DMM beeper) each IC/socket pin in turn to adjacent parts. The object is to see if it connects to other parts (including the copper traces) that it should connect to and does not connect to anything nearby that it should not connect to. A proper PCB with solder-resist coating over the traces hinders this process, so have to find part of the trace with a solder pad. Obviously, mark off on a paper copy of the scheme as you go. This has a good chance of finding anything hidden from the eye.

Quote from: comradehoser on October 13, 2022, 12:29:54 PM
I don't want to Bragg, but yes I did see that.

I suppose it's the same old song, but I was thinking of a different version.

Quote from: duck_arse on October 14, 2022, 10:33:47 AM

Quote from: comradehoser on October 13, 2022, 12:29:54 PM
I don't want to Bragg, but yes I did see that.

I suppose it's the same old song, but I was thinking of a different version.

...same old song. With a different meaning since you been gone...

DMM'ed all around IC6 and everything checked out, far as I could tell. 

As for top songs you mention, out of left banke is fine if you like that kind of thing.

Sorry, ladies and germs, that's the best I can do. 

I haven't got the nerve to reply to that, comrade, but I've got something on my mind - a question.

(https://i.postimg.cc/VNCbNF9C/beta-beater.jpg)

given that:
- D2 and D3 are connected to the output of an opamp
- the opamp output can't swing to a diode drop from the supply rails
- the diodes are DC isolated from whatever follows by C20

what is the purpose of those diodes? please, correct any of the givens above that are wrong, but it looks like design feature gone wrong to me.

Those diodes actually should be there when there is a connection to the outside world. They should be on inputs too. Normally they do nothing unless some stupid alien high voltage gets to the jack tip when they should clamp/discharge it to the power supply. Ideally, there is a cap and a resistor between diodes and the world to absorb some of the pain.

I should add that the fact that adding the little feedback cap on the final stage changes the feedback frequency only a little suggests (to me) that the feedback problem is a little more general. It's as though the 0v ground plane somewhere is floating (other weirdness from just touching jack bodies should not happen).

I wonder if the ground plane to some points in the circuit isn't connected, perhaps some vias between the sides of the PCB are missing? It could take a while to track something like that down and there's no guarantee that there is a problem with that to find!
You do something like try an audio probe around all circuit points that should go hard to 0v while it's squealing and if a ground point isn't really ground, the probe will pick up some audio.

stupid aliens.

Hahaha---

Yeah, I had to work really hard to make those last puns.  I had no idea the 4 tops had sung walk away, Renee, much less that the left banke wrote it.

duck_arse--yeah, I've seen the protector diodes in several other circuits that I have built (I would have to look to see which particular ones).

Thank you very much to you and anotherjim for helping me to think through this methodically, though.  If I am going to have pain, I may as well get knowledge, and I have learned a lot about troubleshooting and circuitry.

Instinctively, it has always seemed to me like it could be a grounding problem, or maybe some sort of bleeding of DC into the audio line since the problem occurs most when the signal is at its most wide-open and removed from ground--coupled with the information that seems to fill the space left when the audio signal is removed and recedes somewhat when it is present.  I will definitely try the audio probe of ground points.

Quote from: comradehoser on October 16, 2022, 10:14:59 AM...much less that the left banke wrote it....

Trivia: the guitarist for the first "Left Banke" touring company later founded Spinal Tap (who will re-unite, if they live so long).

Michael McKean, the actor?

I never knew he had an actual music career besides the one he invented in Spinal Tap.

Okay.  I changed the testing setup.  Moved the board out of the box with jacks connected to the outside of the enclosure for ground; audio probed all grounds as per anotherjim's suggestion, with all pots 100% on full, level pot still out.

Anotherjim: maybe your surmise about the grounds might have something to it.  I detected a more-or-less faint presence of the whine at the ground-facing leads of c21 (negative) and c22 (positive).  The rest of the ground points when audio probed, produced a "bump" and then quiet as the signal was shunted to ground.

For fun, I probed the v+ and v- coming off the RC4558s, they and pins 1 and 8 of IC5 had whine.  Ports 7 and 14 of IC2 were silent.

However, it is very difficult to say anything definitively, as the probe picked up the audio signal ambiently in the space around and in  the circuit without contact. (but also if I was touching components with the probe in my other hand).

I tried jumpered the C21 and 22 pins to ground and besides nice sparks trying to find the right pins, no change playing into the amp (no audio probe).  Subbed in a new RC4558 in IC1, no change.

While the board was in bypass mode/effects off and going to the amp (out of audio probe), though, I made it whine by touching the sides of the board around the middle (R1-5 on left, and r20-22 on the right).  It didn't happen if I held it further up or down.

Other puzzles while probing: again, the presence of the squeal seems to change from previous probes--maybe because it's out of box and because instead of settling at threshold squeal (to save my hearing) I just blasted everything on full.  Detected in all of the ICs, and also right at the input jack and wire.  The signal seemed somewhat effected at the input with the probe, maybe there's some kind of bleedback there or just picking up the ambient signal? 

In any case, I just finished building the bass version of the circuit with the proper routing for 4069 (v.1.1).  I'm waiting on some 100nf caps to come in and then I will fire it up to see if we have equivalent squeal.  If so, I'll know that it's just the circuit.

I wanted to thank everybody who contributed to help me troubleshoot this.  Duck_arse and anotherjim in particular, thanks for your advice.  I actually learned a lot from you all, and it changed my practice for the better, even though I still have a lot of questions.

Here is what I think is the final analysis: it's just a freaky circuit. 

Built the bass, "correct" version 1.1, same squeal with drive maxed and the last 10% of the treble pot.

Then, I went back to my pcb guitarmania version of the beta with the same conditions, listened hard, and.. it was real close to outright squeal, but stayed in the background.  I had subbed TL072s for all of the RC4558s, and so, hey, I subbed all the 4558s back. And... same conditions, same exact squeal.  So, three different builds, three different component lists, two different values, two different PCBs, it all points to one conclusion to me: the original circuit (or original circuit mapping) being funky.  And it did (and does) behave pretty strangely when putting it through its testing paces.

BTW, as a preamp, the TL072 in the beta sounds really amazing, particularly in the clean and light overdrive range and puts out A LOT of signal.  But the overdrive is not as guttural as the RC4558s, which really nail the tone of the original overdrive sound.

My best layman's guess is that something in the TL072s has more capacity than the 4558s, and that the 4558s in the circuit are just maxed out and can't cope with the signal and gain wide open. So maybe the overtopping gets thrown back in the mix somehow and creates the feedback?

And, I did 2 more Aion builds (Binary fuzz and elysium delay), and they worked flawlessly off the first try.

Thank you folks!

so what are you going to do with the three failed - reject or rework?

4558 is known for better clipping sound than 072, and some of those stages do clip. Maybe try subbing only IC6 with 072 and leaving the rest as 4558, since that is where you hear the noise.

There is a very funny feedback arrangement on that last op amp, with the return route coming from after the output cap... just weird. I would think that alone would make it sensitive to what ever it is plugged into.

idy--that is a good idea.  I'll try it out.

duck_arse--They are very functional and actually have pretty great tone (if noisy) other than the squeal-producing parameters. I will probably just keep them and be mindful of levels to avoid the squeal-producing combos, while trying things here and there.  Seems like the squeal is unavoidable if I want that Beta overdrive sound. I imagine they will react the same in line with a 100w power class D power amp,  but who knows.

I was hoping someone knowing would comment on that output stage, the only place you can trace the oscillation. Isn't that an odd way to implement an output level?

(https://i.postimg.cc/4mRCdV75/Screen-Shot-2022-10-30-at-10-23-07-AM.png) (https://postimg.cc/4mRCdV75)

If you look further back there is more weirdness in the tone section, or maybe just oddly drawn.

(https://i.postimg.cc/D8nCHg4t/Screen-Shot-2022-10-30-at-10-26-13-AM.png) (https://postimg.cc/D8nCHg4t)

both bass and treble are directly attached to that level pot...

I'd get rid of D2 & D3, place a 47 - 100 pF across LEVEL pot and never deal with it afterwards.. :icon_wink:

idy: I believe duck_arse had a similar observation and anotherjim commented on it--it's something like a surge protector, apparently.

Antonis--an interesting idea. 

All that should be happy though, since the output stage level pot sends negative feedback to the tone control common connections. Should be stable? But the mids inverts so does IC4A's contribution imply some positive feedback or at least phase shifts that degrade the negative feedback?

Quote from: anotherjim on October 31, 2022, 06:04:22 AM
All that should be happy though, since the output stage level pot sends negative feedback to the tone control common connections. Should be stable? But the mids inverts so does IC4A's contribution imply some positive feedback or at least phase shifts that degrade the negative feedback?

if only we had someone with a squeeling build willing to try a different last stage, and proove one way, or another.

Hahaha.... well, I can try.  Y'all tell me what to do and I can give it a shot. Preferably with a minimum of desoldering. 

I already put the volume pot back in BTW.  I figured after 3 boards it was just the circuit.  Talked to Aion and apparently he traced it right from the factory schematic.  I don't know where PCBGuitarMania got theirs--I think maybe here.

The interesting thing would be if anyone has an OG Sunn Beta to see if that squeals at the given parameters.  It would be interesting to A/B the schematics. as well.  Given that I am a 63% color-by-numbers noob not given to schematic decoding--like detecting what is weird around IC6 is totally beyond me--it could be interesting. 

I do know that Aion's TS-50b build was part-for-part exactly the same as the amp's original schematic, with changes due to obsolescence (the transistors) or redundancy (footswitch circuit).

try as antonis said above. I dunno, does that stage need gain-as-volume or should it just buffer the tone stage and proper volume control?

I can do that!