Momentary switching from Latching switch?

[Ant music]

Hi guys, I know this might seem weird but please hear me out....

I'm doing a mod on a friends DIME from Hell Wah pedal. We are trying to find a way to have the ON/OFF switch on the Wah pedal also control on /off states for effects on his Axe FX. The problem here is that the ON/OFF swtich on the wah pedal is a latching switch but the AxeFX needs to have a momentary switch. I'm wondering if there is anyway we could put something together that would allow a latching switch to also trigger a momentary behaviour for an external device?

I'm open to all ideas and suggestions. Love to have your help. Thanks guys.

[R.G.]

There are ways. They will in general sense that the switch in the wah has changed state, and then make a pulse with circuitry. The simplest way is to replace the wah switch with another switch that adds another pole. So if the switch in the wah is a SPDT, change to a DPDT. If the switch in the wah is a DPDT, change to a 3PDT.

The extra pole lets you cleanly sense what position the wah switch is in, and generate the pulse. Using an extra pole added to the switch to trigger a pulse generator. I would use a CMOS Schmitt trigger gate and an XOR gate with some Rs and Cs to turn on a small MOSFET or bipolar transistor. The circuit can be set up so the Rs and Cs determine the width of the resulting pulse. There are other ways, but this is pretty straightforward.

[Ant music]

Thanks mate. I do have a 3PDT switch for this mod. I ordered one because I was pretty sure I'd need one for this. I will look into it further. I'm very much a novice with things like this so I'm kinda stumbling my way through it. Any further info or details is very much appreciated.

If anyone else has any suggestions I'd be super happy to hear them.

[R.G.]

Here's a quick and dirty schematic.

I couldn't find any easy XOR chips that will run on 9V, so you'll have to make 4.5 to 6V for the chip to run on. A zener diode supply or a 78L05 would work fine.
An XOR (exclusive OR) gate as used has two inputs. The output of the gate is high when the inputs are different. If they're both high or both low, the output is low.
You set the switch up so one throw goes to 5V, the other to ground. The common pole goes to one input of an XOR gate. The switch also drives -a 100K resistor and a 100nF cap to ground. The voltage on the cap ties to the second input to the XOR. When the switch changes, one input of the XOR sees the change instantly, but the other input is slowed down by the R-C network. So the inputs are different for some amount of time determined by the R-C delay. At each change, the XOR puts out one positive pulse for a while.
I showed a second XOR gate. You get four of these in the same 74HC86 chip, so it's free. 
The other XOR has its inputs tied one to +5 and the other to the output of the first XOR. So it goes high when the first gate's output is different from high. In effect, this makes the second XOR gate be an inverter, and so it makes a negative-going pulse at the same time the first XOR does a positive going pulse. You get to pick which polarity of pulse to use.
I showed the pulse(s) driving a 2N7000 mosfet. This can pull down on a resistor or relay coil or whatever is needed to drive your remote momentary inputs. Some effects pedals using momentary inputs can work directly from a transistor drain (or collector - you could also use a bipolar NPN there) pulling them to ground. Some can't. I ran into this with the "Time Bandit" pedal that looked for a click track or tap tempo input, and not only figured out what the timing was, also had an output for tap-tempo-ing other pedals along with a click track. Some (... odd, I think...) pedals had to have a floating momentary contact, not tied to ground. And some were very persnickety about exactly how long the momentary contact was made.

Ask questions if this is confusing.

[merlinb]

If it is a double pole switch as in RG's diagram, then here is another classic solution. As the switch arm travels between one pole and the other, before it makes contact, you get a pulse.

[R.G.]

That does work with some switches. It's heavily switch dependent.

If your switch is a break-before-make type, it does work because there is some travel time when the pole is between the throws. How long the pulse is depends on the travel time, and may change as the switch wears. Some switches may be designed to have nearly zero open time, or even make-before-break, where the open time can be zero. The archaic Carlin stop switches were designed for switching AC line voltages and were definitely break-before-make and worked beautifully this way. Newer types, especially if they're mostly sold for lighter duty switching may be chancy. The XOR pulse-on-change circuit works reliably even with make-before-break as it looks only at the electrical voltage changing from what the RC delay "remembers".

Both approaches are open to bouncing problems. In a make-before-break switch, the pole will bounce for a few microseconds or milliseconds as it hits the new "make" position. This may or may not be a problem with this application. In the XOR circuit, a second RC network on the other XOR input can be used to eliminate bouncing, and a longer delay for the other input gives the pulse output. Using an XOR with Schmidt trigger inputs like the 74HC86 lets you manipulate the delays over a very wide range.

Frankly, a US$0.50 microcontroller in an 8-pin DIP and programmed in Great Cow Basic is a cost competitive and more generic and flexible solution. It can be made to debounce and generate any kind of pulse at all, or even multiple independent ones, flash LEDs, etc. I just don't recommend uCs here much because only a few of the participants can do the programming.