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

[comradehoser]

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?

[duck_arse]

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.

[anotherjim]

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.

[anotherjim]

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.

[comradehoser]

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.

[antonis]

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..

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..)

[comradehoser]

Will do.

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

Rene

[duck_arse]

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.

[anotherjim]

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.

...don't walk away, Rene.

See what he did there?

[comradehoser]

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?

[guitarhacknoise]

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

[comradehoser]

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

[anotherjim]

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.

[duck_arse]

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.

[anotherjim]

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...

[comradehoser]

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. 

[duck_arse]

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



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.

[anotherjim]

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.

[anotherjim]

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.

[duck_arse]

stupid aliens.