Multi-Guitar Selector Design Help Needed

[Satch12879]

Hey everybody,

A buddy of mine commissioned me to build him a switch box so that he could keep his multiple stage guitars plugged into one location and push a stomp to select which one gets sent to his amp.  It's pretty much like an A/B/C in reverse - multiple ins, one out.

So I started thinking about it and I was initially going to make it totally passive which I nixed in favor of either a Fetzer Valve-type preamp or a unity-gain buffer on each input (still haven't decided as I need my friend to demo the circuits before I implement them).  I decided to also add a dedicated tuning output and a master mute switch.  Fine, I can do all that.

The main switching between guitars was the main issue remaining.  Originally he asked if I could do something for 3 guitars but in making the sucker passive, I knew that in using either a DPDT or a 3PDT I'd only be able to switch 2 signals.  I told him so and the reasoning and he agreed.

I haven't built it yet as I'm still laying out in my (very limited) free time, but the switching problem remains.  I really want to be able to give him the 3 (or even 4) selectable inputs that he originally wanted.

I've been reading up a little on electronic switching (I know heresey around here but I'm not so swayed by all the mojo-talk) - Anderton's got a CMOS based electronic switch in his classic book and there's a schematic with another CMOS and a Flip Flop over at J.Philpott's site.  Neither will do what I need it to as they are simply electronic true-bypass schemes.  Then I was thinking about using relays which combines the best of both electronic control and mechanical signal transfer.  But I don't even know where to begin with relays.

Basically what its gotta do is only allow one signal to pass and mute or switch out the other two, based on which selector stomp is pushed.  If you push A, B and C are taken out.  If you push C, A and B are out.  You get the idea.  And, you guessed it, LEDs would be nice to have, too.

I was just wondering if anyone could point me in a good direction, if it's possible, feasible, or if I should stick to the 2-way passive setup and be done with it.

Thanks a million in advance.

[Mark Hammer]

There's a great little project from Polyphony magazine at my site (http://hammer.ampage.org) titled the "Quad Sequential Switch".  

It is essentially a little sequencer that goes from stage/step to stage/step via pressing of a momentary switch.  That steps a 4017 counter through its paces, with each stage output turning on a different CMOS switch (in the form of one section of a 4066 quad switch).  The project as shown only involves four steps before returning to step one, but it could easily be expanded to as many as you want.

Imagine you have 6 guitars all plugged into the rear skirt of some common block.  Facing you are seven LEDs (one for "No guitars, thank you) and a momentary footswitch.  You're using guitar #2 but want to use guitar #5, so you step on the switch once, twice, three times, and watch the lit LED change as it moves across the control block.  When it gets to guitar #5, you stop.

Does that sound convenient enough?

[Johnny Guitar]

Well, I'm still very much a newbie, but if you are going to have a buffer involved, why don't you think about having a switch for each input (ie, guitar) and just grounding out the inputs which aren't used. Of course, with a switch for each input you could have more than one input on and you could also have all inputs off with something like this. This could be a problem.

You could do the same thing with CMOS or relays. With CMOS you could probably easily eliminate the multiple inputs on/ no inputs on problem (with something like a 4051 for 8 guitars) while only grounding the inputs and never running the guitar signal through the IC.

Let us know how this goes.

J

[Satch12879]

JG, that was the intent - to have individual selector switches for each input; you step on one at any time and that's the input that goes to the output while the others are taken out or grounded.  I see what you are saying but I want to be able to prevent having two guitars active at one time and the master mute will take care of shutting the whole works down.

Mark, I'm going to look at that, I'm just wondering if there's a way to substitute multiple switches for the single "scroll through" switch that that design uses.

I'm looking at Lehle's 3@1 and that seems to be just what I'm after...

Any more ideas would be more than welcome :mrgreen:

[Mark Hammer]

I want to be able to prevent having two guitars active at one time and the master mute will take care of shutting the whole works down...... I'm just wondering if there's a way to substitute multiple switches for the single "scroll through" switch that that design uses.

The sequential switch does exactly that (only one on at a time).  Unfortunately there is no simple way to use the circuit as shown and have direct random access to one of N choices.

There ARE switch banks that will permit selection of A to defeat selection of B, but those switches in mechanical form tend to be clustered tightly and not amenable to foot use.  The alternative, which I'm certain RG Keen is preparing a note for right now, is a PIC based control centre, where a bank of switches is scanned by the microcontroller and whatever switch you've most recently stepped on actuates that signal source and defeats all others - a digital priority system.

[Johnny Guitar]

The alternative, which I'm certain RG Keen is preparing a note for right now, is a PIC based control centre, where a bank of switches is scanned by the microcontroller and whatever switch you've most recently stepped on actuates that signal source and defeats all others - a digital priority system.

I *really* look forward to this! especially if all the "exotic" hardware (ie, PCBs, PICs, etc.) are to be sold somewhere.

J

[Satch12879]

The sequential switch does exactly that (only one on at a time).  Unfortunately there is no simple way to use the circuit as shown and have direct random access to one of N choices.

There ARE switch banks that will permit selection of A to defeat selection of B, but those switches in mechanical form tend to be clustered tightly and not amenable to foot use.  The alternative, which I'm certain RG Keen is preparing a note for right now, is a PIC based control centre, where a bank of switches is scanned by the microcontroller and whatever switch you've most recently stepped on actuates that signal source and defeats all others - a digital priority system.

Ah... yep, what I thought...

[Paul Perry (Frostwave)]

If anyone can do CMOS logic, it is easy to have 3 momentary foot switches as input, and a multiplexed output selector.
I can't personally do the logic, but I just KNOW it is a 'three chipper'.

For the true bypass guys, the outputs could power 3 relays with the outputs joined.

[Gilles C]

It would be easy to do with only a PIC and 3 transistors or FETs and a few resistors and caps for the logic.

For the switching, you just have to decide if you want to use relays with a wall-wart (because of the higher current involved) or self latching relays, or a kind of mixer which would have the unwanted inputs shorted to ground when in the OFF position.

[R.G.]

Guys, this is easier than you're making it look.

See:http://geofex.com/Article_Folders/fxswitchr/fxswitchr.htm, figure 7, for the logic of a two-chip CMOS radio button system for up to 8 momentary switches. You **could** do it with just the octal latch, but I like belt and suspenders, hence the schmitt trigger inverters on the input switches. The circuit works well indeed if you use the specified parts, in particular the 74C373, not 74HCT- or 74HC- types.

The LED stuff on the outputs is nice for seeing what is happening.

Once you get a one-of-N switch selector going, all you have to do is make the logic signal run whatever signal switch you like. JFET switching works well, especially on pure-guitar-level signals, but you can also use a bipolar or MOSFET device to drive the coil of a relay. Ordinary CMOS switches like the CD4016 or CD4066 are good too, especially since the thing will likely be used between songs, not during one, so any small clicks won't be intrusive.

A PIC is indeed a single chip solution, and I have done the code for and built a footswitcher based on one. But ordinary CMOS works, too.