LED oscillation noise bleed through in DIY chorus pedal

[tonebreaker]

I just finished a PedalPCB Sea Horse-based clone build. The layout wires the LED effects with the rate together, which was rather brilliant but somehow presents an oscillating noise bleed-through from the LED blinks.

Here's a video.
https://streamable.com/hhm6c3

Here's a gut shot.
https://imgur.com/hiGZ3WY

- I use a 5mm white LED. The noise was gone when I removed the LED altogether. Tried a 3mm (red) with the same problem.
- Power leads are solid core twisted, unshielded.
- The pcb is a friend's custom print based on the Pedal PCB Sea Horse

What could be the cause of this, can it be remedied?

[antonis]

Try a higher value current limiting resistor..

[DrAlx]

+1 on what Antonis said.  The schematic shows only 1k in series with the resistor.  Try a 10k.

I could not find a PCB layout, so it is not clear how the black wire from the LED (which goes back to the board) ends up having its return current routed through the PCB.  It could be a poor layout, i.e. the return current from that LED might be sharing return paths with the signal current.
I would try the following.

1) Don't run the black wire from the LED back to the PCB.  Instead, try run it to ground point on the DC connector, so the return current from the LED totally bypasses the ground tracks on the PCB.  See if the clicks get quieter.

2) Regardless of whether that helps, add some more resistance in series with that black wire (10k say).  Saves you having to desolder and replace the existing 1k resistor on the PCB.

[tonebreaker]

I tried the following with same result on the bleed through:

- Replace 1K to 10,15,20K
- LED Ground to 3PDT instead of board
- Run the LED off board through the 3PDT and power jack

When I run the LED off board, there is no noise, but the effects were bypass too. So perhaps there is a ground short somewhere? Bad layout?

The pcb has two ground lugs so I'm gonna try moving to the other one and see if it changes things.

[antonis]

Something weird happens there:

[DrAlx]

I tried the following with same result on the bleed through:

- Replace 1K to 10,15,20K
- LED Ground to 3PDT instead of board
- Run the LED off board through the 3PDT and power jack

When I run the LED off board, there is no noise, but the effects were bypass too. So perhaps there is a ground short somewhere? Bad layout?

The pcb has two ground lugs so I'm gonna try moving to the other one and see if it changes things.

Sounds like an error in the switch wiring.
Looking at your 3PDT pins in the picture

A B C
D E F
G A G

It seems like (I am guessing here, so confirm)

A = Signal In (Input Jack)     B = Signal Out (PCB)              C = (presumably LED related link of some sort)
D = Signal In (PCB)              E = Signal Out (Output Jack)   F = Signal Gnd (Output Jack)

So with effect off:
    D-G  E-A  F-G  ==> Signal In (Jack) - Signal Out (Jack)
                                  and    Grounded Signal In (PCB)
With effect on:
    A-D  B-E  C-F  ==> Signal In (Jack) -Signal In (PCB)
                                  Signal Out (Jack) -Signal Out (PCB)
                                  Signal Gnd (Output jack) to  C (whatever that is)

You should wire:

     F to the Ground on the DC jack, NOT the signal in/out ground as you have in the picture.
     C to the black wire from the LED.

No signal ground wire should be going to the 3PDT switch.  It should go direct to the PCB.

[DrAlx]

If you break the connection between the two G points on the switch, then you can wire the left G point on the switch to the signal ground on the jack, and leave the right G point unused.
You still need a direct connection from the signal ground jack to the PCB though.

Use pins C and F purely for routing the black LED return to the DC jack.

[amptramp]

With this sort of problem, you could always drive the LED with a differential amplifier.  A constant current source is your friend when trying to quiet a thumping circuit down.

[tonebreaker]

Sounds like an error in the switch wiring.

Thanks, will recheck and rearrange based on your suggestion. I do have been getting stomp switch noise on the amp, this could be the solution.

With this sort of problem, you could always drive the LED with a differential amplifier.  A constant current source is your friend when trying to quiet a thumping circuit down.

Can you expand on this? What component and how do I wire the LED and the power in the box?

[antonis]

With this sort of problem, you could always drive the LED with a differential amplifier.  A constant current source is your friend when trying to quiet a thumping circuit down.

Can you expand on this? What component and how do I wire the LED and the power in the box?

A simple CE amp with fixed Base bias (hence fixed Emitter voltage) and an Emitter resistor of value V_Emitter/I_LED..(LED should be wired between power supply and Collector)
Bias configuration could be Diodes, Zeners, BJTs/FETs or anything else providing stiff Base bias..
https://arduinodiy.wordpress.com/2015/12/12/driving-an-ir-led-constant-current-source-or-not/

or something like below:


R1 value is set for about 6mA (theoretically) constant current when T1 is ON..
You can tweak R2 value according to op-amp output taste..
(taking into account both Collector current/h_FE and diodes working current..)

The actuall goal is the significant reduction of driving current by "indirectly" triggering the LED..
(like Ron suggested above..)

[ElectricDruid]

Use a low-current/high efficiency LED to limit any thump in the first place. Use the biggest current limiting resistor and the lowest brightness you can to limit it further. Finally, make sure it's wired up correctly. Like this, I've never had this problem. More complicated solutions have never been necessary.

HTH