Aion FX Sunn Beta Lead build whine/squeal in the treble pot

[comradehoser]

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.

[PRR]

...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).

[comradehoser]

Michael McKean, the actor?

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

[comradehoser]

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.

[comradehoser]

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!

[duck_arse]

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

[idy]

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.

[comradehoser]

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.

[idy]

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?



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



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

[antonis]

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

[comradehoser]

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. 

[anotherjim]

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?

[duck_arse]

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.

[comradehoser]

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).

[duck_arse]

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?

[comradehoser]

I can do that!