Celsius Preamp - Whining noise with switch on High

[airbud]

Hey everyone, I mentioned this in response to a build report but wanted to get some more in depth help here if possible.

The preamp sounds like it should, but on higher settings with the switch set to High there's a "whine" that dominates the sound. I switched out 3 different each of charge pumps and opamps, but the issue persists.

Here's the frequency response running my Tele through the Celsius into my SSL mixer.



I appreciate any help you guys can offer, thanks!

 

[chris.knudson]

Got any photos? Is the whine in the 10-12K range, or is it in the 100hz to 1khz - sorry, I'm not exactly sure what I am looking at.

 

[jwin615]

It's octaves of 4.5k.(edit)
1)What charge pump are you using?
1B)did you use different charge pump part numbers or 3 of the same?
2)Did you route the wiring away from the circuit board?

 

[jwin615]

Also, may want to try it on its on with a dedicated supply, no other pedals in the chain. G-chourus-amp
There's a weird thing that happens with charge pumps that aren't isolated. Just to rule that out.

 

[airbud]

It's octaves of 4.5k.(edit)
1)What charge pump are you using?
1B)did you use different charge pump part numbers or 3 of the same?
2)Did you route the wiring away from the circuit board?

yes I figured it was octaves but I had no idea where it would come from.

1a. TC1044SCPA from Mouser
1a. three of the same part number, from the same order.
2. i'll attach a photo shortly, i don't recall it being any different than any other pedal i've built.

 

[jwin615]

yes I figured it was octaves but I had no idea where it would come from.

1a. TC1044SCPA from Mouser
1a. three of the same part number, from the same order.
2. i'll attach a photo shortly, i don't recall it being any different than any other pedal i've built.

TC1044SCPA switches at 10k.
Unpowered and with the charge pump removed, make sure you have continuity between pin 1 and pin 8 of the charge pump socket. IDK if that was the issue with the first revision or not. @Robert could say for sure.

I would route your IO wiring around the edge, maybe try shielded wire, grounding one end.
Also, try it solo. The charge pump could be heterodyning with another. Boosted switching freq (pin 1 to V+(pin 8)) is 45k. 10x your noise. Good chance that's the root. Correlation/causation etc but that's a funny coincidence if nothing else and worth entertaining.

 

[Robert]

Unpowered and with the charge pump removed, make sure you have continuity between pin 1 and pin 8 of the charge pump socket. IDK if that was the issue with the first revision or not. @Robert could say for sure.

The first revision had a high noise floor but not charge pump noise.

Pin 1 and 8 should be connected on the PCB. Any project here where they aren't already connected would have solder jumpers to connect them.

 

[airbud]

I have continuity between 1 and 8 while the pedal is unconnected and unpowered.

I attached a pic of the guts, I don't think the wires are too close but it's definitely not shielded wire.

@jwin615 what do you mean by trying it solo and heterodyning? New term to me.

I really appreciate all the help, I didn't understand the charge pump part of the circuit but it's appearing to be a really important concept for pedal building. All my circuits knowledge is from college almost 20 years ago and the last few months of research, so it's great to have a chance to learn more.

 

[Robert]

@jwin615 what do you mean by trying it solo and heterodyning? New term to me. I didn't understand the charge pump part of the circuit but it's appearing to be a really important concept for pedal building.

If you remember the old voltage multiplier circuits involving a lattice of diodes and capacitors that's basically what it is. They depend on AC to function so the charge pump acts an an oscillator to generate the pulses needed to charge the capacitors.

The problem is that sometimes that oscillation bleeds into the audio circuit (or interferes with other pedals or power supplies that use a similar setup).

The solution is to run the oscillator at a frequency outside of the audio range, which is what jumping pins 1 and 8 achieves. Only specific charge pump ICs support that feature.

What is the full part number of the IC you used?

Nevermind, I see it in your picture. (Typing from a phone, I hate phones)

 

[jwin615]

What Robert said
That's the point in trying it solo. Just making sure that it's not interfering with another pump in the chain.
May be worth trying to move the input wire around, closer and further from the cp ic and see if the level of the noise changes. Maybe try stuffing closer to the top.

 

[airbud]

Ah ok, the switching frequency makes sense now. No luck soloing it or moving the wires around.

 

[airbud]

Here's what I got using an LT1054P. Sill a significant whine. Slightly higher pitch and a bit quieter.

Going to examine other charge pump options with 20kHz+ switching frequency.

Is 20kHz the right threshold, or is it higher because of some sampling thing or something?

 

[airbud]

after doing some more research, i found this where the guy solved it with a different power supply. this worked, one of mine is apparently really crappy, and the other two are relatively quiet (one very quiet).

i was about to buy more charge pumps so this is a relief!