RG, ZV, HELP!! Bypass/LED popping

[JisforJustin]

OK. So I am building an original fuzz effect and cannot seem to cure the bypass LED popping problem. I have tried to combine RG's suggestion (from http://www.diystompboxes.com/smfforum/index.php?topic=44109.0):

"Split the LED resistor in half and put [a] cap there"

And Zach's suggestion (same thread):

"Use at least a 15k to 22k resistor and an ultra-bright LED"

What I tried was to use two 10k resistors with a 47uf cap, to ground, in between, and a super bright LED. I even tried adding a second stage of 10k resistor and 47uf cap. This seems to help slightly with the popping, but it is still quite prevalent. I also tried various LEDs and they all did the same thing. I am 99% sure it is coming from the LED, because when I cut the lead to the LED the popping goes away and the effect sounds great. Does anyone have any other suggestions to try?

Justin

[JisforJustin]

Anyone?

[R.G.]

Not knowing either the schematic or the physical layout of your circuit, all I can do is give you general principals.

There are only three ways things can pop.
(1) there is a resistive coupling of either a transient click or a sudden new DC level. But resistors are sneaky. Copper wire is a resistor, just a very low one. It makes you think that it's a short circuit, and its not. Sudden changes on either ground or power supply level caused by the LED being switched on and off can cause the ground or power supply to move, and your circuit can amplify that. The cure for that is to not let the LED current come through the same wires that the power to the rest of the circuit does, or to slow the LED current change ... way... down... somehow.
(2) Capacitive coupling. Any two conductors separated by an insulator is a capcitor. The higher the input impedance of the circuit, the more it picks up capacitive changes. FETs and FET input opamps are worst. Bipolars can be almost as bad. The cure for capacitance coupling is shielding and distance.
(3) magnetic field coupling: A sudden change in current in a loop of conductors causes a radiated magnetic field change. Any other loop of conductors that overlays any of that M-field (which is infinite in extent... 8-)) will pick up at least a little of it. You cure M-field coupling with making the loop smaller, the transient slower ( it couples more poorly to the impedance of free space that way) or by shielding. Shielding is less effective for M-field than for E-field (capacitive).