Author Topic: Killswitch via optocoupler  (Read 521 times)


Killswitch via optocoupler
« on: June 14, 2018, 08:51:03 AM »
Hi, everyone!

I'm installing a killswitch on a guitar. Everything is fine with the standard wiring (killswitch connects signal to ground), besides the classic POP on switch (noticeable when no distorsion is engaged).
I've read about connecting a resistor to limit the POP, but I didn't like the solution (also, the POP was still noticeable).
I've read this interesting thread:
Worth a try!

I've prepared this schematic:

This is how it should work

Circuit OFF:
- Signal directly connected to KILL
- VCC_GND floating (batteries disconnected)
KILL goes to GND, SIGNAL goes to GND (audible POP).
Long story short, just a normal kill switch.

Circuit ON:
- SIGNAL Connects to SIGNAL_IN; will be processed by circuit
- VCC_GND connected to circuit GND
KILL goes to GND, U1 LED turns ON, SIGNAL_IN (connected to SIGNAL) goes to GND "slowly" (hopefully, no POP)

I've tested it, but the result is not as expected...

First, there is still a POP. From the datasheet (the unit is actually a 4N28, not a 4N35), I've read that the switch will occure in microseconds. Probably a not-so-slow transition? (so there's still a POP)

Second, when the killswitch is pressed, there is still sound from the guitar. The volume is quite low, and with an added ... very fast tremolo. A quite awful sound. Any idea why this is happening?

The main differences with the "schematic" proposed in the thread are:
- 3V (actually 2.4V) VCC instead of 9V. From 4N28 datasheet this can be enough. Or is it?
- killswitch connect/disconnect the optocoupler ground instead of VCC. Doubt it will make any difference...
- 2k2 resistors replaced with 330 (using way lower VCC, and optoucoupler maximum rating for forward LED current is 10mA, so)
- added a pullup resistor on the KILL signal. Overkill, I know... Anyway, same behavior without it

As you can see from the schematic, there is also a test LED, to check that SIGNAL_IN actually goes to ground (LED will turn on). And it does.
As stated in the note, I've removed the LED from the circuit once I've connected the guitar signal.


Re: Killswitch via optocoupler
« Reply #1 on: June 14, 2018, 04:06:20 PM »
I wish I had time to respond to this but I'm sorry right now is a busy time.
As soon as I get the chance I'll chip in.
I recently discovered the limitations of using a 4n25 with some tremolo circuits.
I ain't no muff builder boi.
Smoke me a Kipper. I'll be back for breakfast.

Grey Paper.


Re: Killswitch via optocoupler
« Reply #2 on: June 14, 2018, 04:19:36 PM »
I'll be glad to listen!


Re: Killswitch via optocoupler
« Reply #3 on: June 14, 2018, 04:57:19 PM »
Maybe that coupler is just too fast?  Do you have another typel at hand? Like a Vactrol VTL5C3?
Tapflo filter, Gator, Magnus Modulus +,Meathead, 4049er,Great Destroyer,Scrambler+, para EQ, Azabache, two-loop mix/blend, Slow Gear, Phase Royal, Escobedo PWM, Uglyface, Jawari,Corruptor,Tri-Vibe,Battery Warmers


Re: Killswitch via optocoupler
« Reply #4 on: June 14, 2018, 07:49:03 PM »
It is quite difficult to mute signals with a bipolar transistor (that's what's inside that opto) quietly. The problem lies in both the speed of the control signal and the offset voltages native to bipolars. That's why bipolars are not used for shunt switching nearly everything.

What you could do is to try to slow down that control signal transition so that it gggrrraaaddduuuaaalllyyy turns the transistor on. The blip will still be there, but will be so slow that other things in the circuit will filter it out so you don't hear it.

Another thing to try is to use one of the LED to dual-MOSFET "solid state relays". There is a part that's popular that I can't quite come up with the number of right now - something like "TD222G" or similar. This has two MOSFETs turned on with an LED. The MOSFET is much more like a variable resistor than a bipolar is for small signals, and there are two so that the unavoidable problems with the body diode is avoided mostly.

You'll still need to switch it slowly. Consider: if your signal was a pure sine wave (it isn't, but imagine it so) and you switched it on at a zero crossing, no funny business would happen, other than in oddities the human ear adds in. But if you switched it on at a peak, the signal after the switch would look like a vertical leap to the biggest value of the sine peak. Your ear will definitely hear this as a "tick" or some other flavor of grak.  If you ramp switch on and off slowly, the grak from instant switching becomes unnoticeable.

Quick IQ Test: If anyone in a governmental position suspected that YOU had top-secret information on YOUR computer, how many minutes would you remain outside a jail cell?


Re: Killswitch via optocoupler
« Reply #5 on: June 15, 2018, 01:35:25 AM »
Thanks for your answers!

blackieNYC, regarding the "too fast optocoupler", this can be why there is still a POP. Awfully, I don't have a slower one to test at the moment.

To R.G., slowing down the control signal can be an idea. Maybe connecting the killswitch to VCC and adding a big-ass capacitor (in the proper place) can lead to some results. In other words, what I'll be doing is to turn up the LED slower.
Regarding the statement "You'll still need to switch it slowly"... well... I know. That's exactly why I'm trying to switch from the abrupt signal-to-ground killswitch (that generates the "vertical leap") to an active switching  :D But it looks like a switch in microseconds in the audio realm is just like a zero-seconds switch... Didn't thought about this when looking for the optocoupler; just bought (almost) the first one I've found.

Your answers address the problem "the POP is still there". No ideas regarding the "signal not actually muted" problem?


Re: Killswitch via optocoupler
« Reply #6 on: June 15, 2018, 10:24:28 AM »
In re: signal not actually muted problem
A muting circuit is not really reducing the signal through it to zero. It's really a "volume control" that reduces the signal to some degree. There's always some signal left, you just have to arrange things so the amount left is low enough for you to either not hear it or be happy disregarding it.

Volume controls are at their heart a two-resistor voltage divider. The top resistor is the sum of the signal source's internal impedance (Rsource) and the actual resistance after the signal source (Rtop). The bottom resistor (Rbot) is the resistor to ground. The amount of attenuation/muting you get is given by the voltage divider equation:
Vout = Vin * Rbot/(Rsource + Rtop +Rbot). The take away from this is that the guitar pickups may be pretty low impedance in the bass end of things, but not at treble where the inductances come into play with a vengeance. Also, the Rbot of the opto may not be close enough to zero to really mute things.

You could do a quick test of this theory of things by sticking a 10K to 100K resistor in series ahead of the collector pin of the opto. This would raise the series resistance ahead of Rbot by a fixed amount, and make whatever value Rbot is when muted more effective. If you are still not happy with the amount of attenuation, you could stack two 10K/opto sections, one after the other. The second section would then mute the output of the first muting section even more.

I don't have a good theory about the fluttering/oscillation you're getting. May be a side effect of the opto having an exposed base, may be something to do with the LED power and or switching.

Quick IQ Test: If anyone in a governmental position suspected that YOU had top-secret information on YOUR computer, how many minutes would you remain outside a jail cell?


Re: Killswitch via optocoupler
« Reply #7 on: June 21, 2018, 02:53:52 AM »
Sorry for the delay. Quite busy.
Regarding the not-completly-muted-signal, note that I'm not talking about a barely audible remaining signal, but about a quite loud one. To give an idea, when I press the switch the volume goes 40%, with the "tremolo" peaks to a 70%.

At the moment I have very little time to experiment further. I'll get back to the topic when I can work on the project again.
At that time maybe I'll have some different optocoupler to test.