Noise from momentary havoc switch

[mark2]

There's a tiny, barely audible clicking when I connect the switch here (its poles labeled VAST1 and VAST2). It's inconsistent in sound, duration, and volume.  i.e. seems purely mechanical.

Here's a clip of it going into a reverb to accentuate it. It's a crude phone recording so I didn't isolate the mechanical sound in the room from the sound in the audio line, but you should be able to hear the difference from the reverb.

The top connection just goes to a pot pin on the FV-1. i.e. the switch momentarily maxes its value.

Is there a straightforward way to filter that out? Or is it inevitable (or going to require lots of experimentation)?

[idy]

simple thing to try is large (1M?) R across the pole. so it's "almost" cutting the connection. Cheap and easy... may not help.

So that is a digital control pin to 3.3v. Digital guys may know. I've used FV-1 a bit, but usually grounding a pot wiper with a foot switch, turning off the reverb level. Different maybe. But that was quiet enough.

[ElectricDruid]

I can't really tell you without seeing where that wire on the RHS goes!

Like IDY said, it looks like a digital input, so that means all sorts of things are possible, but we'd need to know more of the circuit to know for sure. It may be that you're coming up against limitations in the firmware? But I can't even see a processor, so that's total speculation at this point!

[idy]

you didn't say what kind of parameter is being switched. I think that makes a difference. The pots aren't real "organic feeling" on FV-1, that I have done... not designed (?) for taking the knob and twisting in real time like an old echoplex, sometimes artifacts. But I'm not an engineer.

For me it was reducing a (mix) value to 0, not jumping to a preset close to 3.3v.

[idy]

Also I/we can't hear it. Soundcloud says no track...

[mark2]

Sorry for the bad link. Here's one with fixed permissions.

I can't really tell you without seeing where that wire on the RHS goes!

Sorry if it wasn't clear, but I mentioned it went to a pot input pin for an FV-1. It controls the reverb level and size.

simple thing to try is large (1M?) R across the pole.

Interesting thought, but no that didn't make a difference.

Maybe related and helpful I'll likely switch that CEILING control to a voltage divider instead.

Worth mentioning: I also tried wire-to-wire connections, eliminating the springs (and hopefully any bouncing), and it still happens.

So I probably either need to find a way to fix it in code (unlikely), or by ramping up the voltage rather than jumping straight to 3v3. Any thoughts on the latter? If it seems feasible, anyone know offhand a way to accomplish it?

[idy]

Now that I can hear your file, I would say: impossible to judge the relative level of that click against the level of sound you would be putting through there. That hiss sounds pretty loud and the click is just above it. The important thing is "signal to noise." With no signal, no ratio.

I can see that my "big R across the switch idea" was futile because there already is an R that you are adding a parallel to.

Simplest way to ramp up voltage slowly is a cap. Takes time to fill up. "Time constant." It might go from "VAST2" to ground. It would charge up to the level set by the pot, then "fill" up to the other level, maybe. The cap is another cheap and easy thing to try. Not sure if you would need to empty it when not in use...(a resistor from + to ground.)

I also see that you made your normal pot a voltage divider (terminal 1 to ground) but made your trimmer to set your "havoc" level a variable resistor. Maybe it should also be a voltage divider. I know I had trouble with "grounding the wiper" which interfered with the 3.3v. Not your problem...


These things can be frustrating. Microphonics. Tapping to seek them out... Switch, PCB, enclosure.

[mark2]

I tried filtering via code, thinking the FV-1 didn't like the voltage spikes (it doesn't, more on that below)... but no luck.

So I thought maybe it was sinking too much current in spikes and that was disrupting something else (a long shot, but hey..) so I put a traditional RC filter on the VAST1 side. i.e. a cap to supply the initial voltage spike. No luck.

Simplest way to ramp up voltage slowly is a cap. Takes time to fill up. "Time constant." It might go from "VAST2" to ground.

I ran a 1k resistor from VAST2 to a 10uF cap and, though I still need to do a lot more testing, this seems to have cured it.

The FV-1 must not like those voltage spikes, and my filtering code must not have been adequate.

Thank you!

[idy]

So relieved. Lucky call.

[mark2]

Interesting (to me anyhow), I need a resistor before feeding the 3v3 into the switch. I assume what's going on is that even though the FV-1 doesn't sink much current, the cap certainly does. So without the resistor the problem is actually worse.