Tone Machine octave switch pop - any way to reduce it?

[aion]

I'm building a Tone Machine with a footswitch for the octave - a simple SPST to connect or disconnect the other half of the signal that forms the octave. I'm getting a pretty good amount of switch pop when I do so. It's the same with a toggle switch (and I even compared it to an original TM, the switch pop is there too) but I am wondering if this is an issue that can be remedied. Since it's a footswitch it's much more apparent, since it's being used as less of a tone setting and more of a mode.

I thought of putting a high-value 1M resistor across the two lugs of the switch so that it's never fully out-of-circuit. Would that essentially mute it to the point that it wouldn't be a factor?

I've used that trick when, for instance, switching out R-C filters to ground, like in tone controls and that sort of thing where there is already a series resistance. But this is a bit of a different application so I'm wondering whether the same theory would apply.

[duck_arse]

what if you add another "R10" to D3, so it is same potential ??

[aion]

what if you add another "R10" to D3, so it is same potential ??

Isn't it still changing the DC potential since the two 100k's are now in parallel? Or would the switching-in of R11 cancel it out, being the other half of the divider? The DC voltage would be slightly different than stock, but closer to center, right?

[duck_arse]

mmm, ahhh, you got me. maybe use 1M to V+ on D3. otherwise - it needs a better brain than mine.

[Mark Hammer]

I've never tried it myself, but I wonder if, instead of connecting/disconnecting D3 to produce/cancel the octaving effect, one left D2 and D3 in place, and used a footswitch to simply bridge R11.  After all, if there aren't equal-amplitude complementary signals arriving at the junction of D2/D3, there's no audible frequency-doubling.

Fundamentally (no pun intended) the objective is to remove one of the complementary signals.  Lifting a diode connection IS one way to do that, but it isn't the only way.

[aion]

I've never tried it myself, but I wonder if, instead of connecting/disconnecting D3 to produce/cancel the octaving effect, one left D2 and D3 in place, and used a footswitch to simply bridge R11.  After all, if there aren't equal-amplitude complementary signals arriving at the junction of D2/D3, there's no audible frequency-doubling.

Fundamentally (no pun intended) the objective is to remove one of the complementary signals.  Lifting a diode connection IS one way to do that, but it isn't the only way.

That seems like it would impact the bias point as well, since shorting the 100k resistor to ground would also change the ratio in the voltage divider.

But, I'm looking back at duck_arse's suggestion and I think it may work. If I'm looking at it right (redrawing just the part between the capacitors since it's an isolated DC voltage) then it looks something like this:



If you measure at TEST (which corresponds to the audio path) then you would get a different DC voltage when SW1 is closed vs. open since D3 & R3 are in parallel with D2 & R2.

But, by adding another 100k resistor from VA on the other side of the switch, it looks like this:



So the effective resistance changes when the switch is closed (R1 and R4 are now in parallel, as are D2-R2 and D3-R3) - but the ratio stays the same and the sides are symmetrical, so the DC bias point should be the same as well.

The DC will be different than the stock voltage in octave-on mode, but it's the same as it is in octave-off mode - which is closer to the centerpoint and so it's more "right".

Am I looking at this right?

[PRR]

> Am I looking at this right?

I know I'm not, tonight. But I couldn't even grasp how it would be implemented, so I crayoned-up how I think it would go together. Does seem to solve the 6V/4.5V DC shift. Does give new/different diode currents. Gain changes somewhat due to change of loading. Too much for my head tonight; maybe someone else will say yay or nay.

[aion]

> Am I looking at this right?

I know I'm not, tonight. But I couldn't even grasp how it would be implemented, so I crayoned-up how I think it would go together. Does seem to solve the 6V/4.5V DC shift. Does give new/different diode currents. Gain changes somewhat due to change of loading. Too much for my head tonight; maybe someone else will say yay or nay.

Yeah, your  drawing is how I was seeing it - it just helped me to be able to see them rearranged as voltage dividers since it was a DC bias issue moreso than its interactions with the circuit before & after. But that's helpful to know about the gains changing. I suppose they already do, though, in the stock configuration?

[PRR]

What is the not-Octave bypass sound? Looks like it is just Clean?

[Mark Hammer]

It is a rare thing to elicit "clean" tone from a Tone Machine.  Even when one lifts the diode pair further downstream.

[pinkjimiphoton]

i built a ftm with a second switch to choose octave or normal fuzz, seems to work great, no popping. i used simple 4.7m resistors to ground as i recall and it eliminated it. i used this particular layout:


https://effectpedalkits.com/shop/foxx-tone-machine-fuzz-pcb/

i think. been a while.

mine will clean up like a fuzzface almost when i turn it down, low hfe transistors again, the octave still screams, too.

good luck!

this is what i did, just used a dpdt with one side to change indicator led's
but its a really simple switching.

[R.G.]

One way to prevent pop is to slow it down - use a JFET switch that closes ...slooooowly... so that the transient is outside the human audio response range. As a practical matter, 30-100mS is slow enough.

For this circuit I'd use two JFET switches, one always "closed" and the other openable, but with the control signal on its gate being slowed down with a resistor/capacitor so it fades in over maybe a tenth to a twentieth of a second. The always-on one is to preserve symmetry when both are active (i.e. the octave is active). You might get away with a resistor that approximates the off/on JFET's on resistance.

[Mark Hammer]

That's a good idea.  And the nice thing is that it can still be done with a SPST footswitch.  Here's how the earlier Small Clones did it.

[aion]

Thank you guys! One of these ought to do the trick. I'll start with adding the resistor to V+ on the other side of the switch, and then if that doesn't work I'll try the JFET route.