Buzzaround charge pump whine... when guitar volume on 0

[idy]

A student and I made a pair of guitar pcb Fuzzy Bees, Ge Buzzaround clones with charge pump. He liked one I had made years ago. After the usual nonsense picking good transistor sets we notice they whine...when the guitar is turned down. I never notice that because I don't use my volume control like that, but he dials down all the time. All three pedals whine..
They whine wether I use 1044scpa or 1076scpa or LT1054 (with the leg 1 lifted.)
They only whine when guitar volume approaches 0. Like 1 and 2 it starts, gets loud on zero.

This reminds me of the recent "pnp fuzz noise when guitar on 0" thread but is a little different. Schematic coming up.... I tried touching 100n across the in and out caps, no help.

[idy]

Sorry, I had to learn how "record" and "capture" work on screen shot!

[antonis]

A little help for my friend..

[ElectricDruid]

The three chips you mention aren't the same, so results with one might not be reflected with the others.

I can't find a datasheet right now for the 1076scpa you mention, but here's the other two:

https://eu.mouser.com/datasheet/2/268/21348a-29235.pdf
https://www.ti.com/lit/ds/symlink/lt1054.pdf

The significant thing for me about these is that they use different methods to increase the oscillator frequency. The 1044 uses the "boost" pin 1 connected to +V (as you've shown in the schematic), but the  LT1054 uses a cap C2 connected back from pin 7 to pin 2 (see figure 12 in the datasheet).

Furthermore, there's some good hints about PCB layout in the LT1054 datasheet (Section 10.1 Layout Guidelines) which are intended to help keep noise away from sensitive lines.

[MikeA]

Re: Tom's comments on location, C8 will radiate a ~10 kHz square wave that can be picked up by nearby wires or components (within 2 inches/50 mm or so) even when you boost the switching frequency.   I usually put the charge pump and associated caps on a separate board far enough away from low level signals to mitigate.  If space is tight, or if C8 is on the same board as your audio path, you can roll up a cylinder of aluminum tape, slip it over C8 and ground it, and that will shield it.  C8 should also be low ESR.  Some charge pump brands want C9 to be equal value to C8 (10 uF), some want it 10x the value (100 uF), depends on the manufacturer, that's usually in the data sheet.   Another possibility is the whine is propagating down the -9V bus, if so, filtering the bus for a 10 kHz or so signal should help that.

[idy]

Thanks, good starts. I tried some transistors that had the base attached to the metal can...mucho whine, little antennae...

I can't find a datasheet right now for the 1076scpa

because I am being stupid. 7660scpa do exist. sorry.

Will give these suggestions a shot.

[idy]

Yes, adding a 51pf cap pin 2-7 of the 1054 helped a lot. Super people.

I see I should have used 5-20pf. Oops. 100pf helped a bit, 51 was better... I think we have this.

I may not need to make a "hat" for C8 after all.

Will do more experiments later.