Momentary mini-toggle that Latches?

[ashcat_lt]

So I'm planning to hack a couple of Korg NanoKontrols into a pedalboard form factor.  It has a bunch of SPST (normally open) momentary switches which can be configured via software to do any number of different things.  Some of these functions I intend to put on stomp switches, and momentary stomp switches are pretty easy to find. 

For other functions, I'd really prefer to have mini-toggles because they feel nice and give a definite reading of their current state.  All of the momentary toggles I've ever seen, though, spring back to their original position, rather than indicating which way they were last flipped, and that defeats the purpose.

So the question is simple: Does anybody know where I might find, or how I might even search for a "latching momentary" switch at a reasonable price point.  Failing that, any good suggestions on a make-before-break DPDT toggle which I could use to fake it?

Thanks for any advice.

[pappasmurfsharem]

I'm confused.

That would defeat the purpose of momentary wouldn't it? Why not just use a normal latching switch?

If you want indication you would use a latching relay and an LED to define whats state it's currently in.

[mth5044]

I see what you're trying to do, but I don't know why a manufacturer would make such a thing or even if they could! Since these switches are all mechanical, it would go way beyond the physical act of pushing the switch to change the contacts.

If you are dead set on this idea, you could use a uController or CMOS switching or such to send a pulse to a bank of switches in IC form to momentarily connect them.

[Govmnt_Lacky]

Momentary and latching are Apples and Oranges. Not really gonna find one with both.....

I am also confused on your needs. On one hand, it sounds like you need a momentary switch with the ability to indicate which "mode" it is in. If this is the case, have you looked at momentary illuminated push-button switches? Push it once, the button light comes on and the momentary switch is activated... puch again, the light goes off and the momentary switch is activated.

The other idea you propose... DPDT make-before-break... switch usually only applies to latching switches.

You might want to explain your requirements a bit better. 

[R.G.]

At a modestly metaphysical level, switches have two properties: connecting or disconnecting circuits and having or not having a memory of what they did last.

Momentary switches, by definition, have no memory. The memory of alternate action switches is mechanical: the locking mechanism "remembers" for the switch what it was switched to last.

To make a momentary have a memory, you ...could... introduce an electronic memory to help the switch. That is, have the switch do what it does, but tap off that "what it does" into an electronic flipflop or some such, and have that in turn light up an indicator, like an LED above the switch. It's some work, but one CD4013 would do two switches.

Just a thought.

[ashcat_lt]

So, I've got a 5VDC voltage on one wire, and ground on another.  When I briefly short these together, the magic box sends some sort of MIDI message.  I'm not sure that it would hurt anything to leave that signal grounded (assuming internal current limiting...), but I'm not sure that it will send another message when I un-short the signal.  I need it to short again when I flip the thing back.  Yes, it would be exceeding easy to hook up an indicator LED (the thing already does this, I'd just have to move the led to the new panel), but it's kinda more messing around than necessary if the switch just tells me which way I flipped it last.  Fancy light-up push-button switches 1) also more messing around than it's worth B) probably more expensive than necessary III) not a toggle, which is what I want!

My idea was that a latching make-before-break DPDT would work if I just put, say, 5V on upper left, 0V on lower right, and tie the wipers together.  It will short in the process of switching, but be open at all other times.  Starting to look like this is the way I'm going to have to go, assuming I can find one that specifies make-before-break.

[karbomusic]

For other functions, I'd really prefer to have mini-toggles because they feel nice and give a definite reading of their current state.  All of the momentary toggles I've ever seen, though, spring back to their original position, rather than indicating which way they were last flipped, and that defeats the purpose.

Wouldn't this still have the problem of the position of the switch only being loosely tied to the actual state? It might be fine but seems there could be cases where the actual state might get reversed compared to the switch state.

[ashcat_lt]

For other functions, I'd really prefer to have mini-toggles because they feel nice and give a definite reading of their current state.  All of the momentary toggles I've ever seen, though, spring back to their original position, rather than indicating which way they were last flipped, and that defeats the purpose.

Wouldn't this still have the problem of the position of the switch only being loosely tied to the actual state? It might be fine but seems there could be cases where the actual state might get reversed compared to the switch state.

Yes, particularly at startup when the parameters all go back to default and the switches are still where I put them last time I played with it.  This will also be true to an extent of all of the knob positions.  It's kind of the same as is.  The buttons don't know if they should be lit or not, but it always figures it out once I flip the thing.  Now that I think of it, I cant tell you why it works, but I've played it a number of times, and it just works.  Something about "soft takeover" maybe?  At worst I'll have to "zero out" all of the switches before plugging it into the computer, but I'm pretty sure that if I just switch it once I'll know where I am...

I perused a few switches at mouser, but neither catalog nor data sheets seem to want to answer the break-before-make question.  I think I have some I can test when I get a good chance.  This project is coming together rather slowly.  Partly because I'm afraid to start soldering things to this double-sided SMD board, and partly because my live machine just doesn't have enough processor to run all of the effects that this thing is going to control.  It works beautifully on the studio machine, but the laptop just ain't cutting it.

[merlinb]

My idea was that a latching make-before-break DPDT would work if I just put, say, 5V on upper left, 0V on lower right, and tie the wipers together.

There's an easier way. Ignoring the problem of things getting out of sync at startup or whatever, there is a classic solution for making a momentary switch using an ordinary (break before make) double-throw latching switch, like so:

You have some sort of pull-up or pull-down arrangement so the circuit switches states as the wiper crosses the gap between the two latches.

[ashcat_lt]

My idea was that a latching make-before-break DPDT would work if I just put, say, 5V on upper left, 0V on lower right, and tie the wipers together.

There's an easier way. Ignoring the problem of things getting out of sync at startup or whatever, there is a classic solution for making a momentary switch using an ordinary (break before make) double-throw latching switch, like so:

You have some sort of pull-up or pull-down arrangement so the circuit switches states as the wiper crosses the gap between the two latches.

Tell me more, please.  What sort of "pull down arrangement?"  How does this work?

[duck_arse]

if you connect ^^ the contacts to ground and the wiper 100k to V+, the wiper is pulled-up while not in contact with, but low when contacting. each toggle will go low-high-low.

[ashcat_lt]

if you connect ^^ the contacts to ground and the wiper 100k to V+, the wiper is pulled-up while not in contact with, but low when contacting. each toggle will go low-high-low.

That would seem to be opposite of what it wants, but I guess I've never tried it.  It's basically like holding the button down, then letting ago and pushing (and holding) again to activate.  I can see where that would work.  It's probably already current limited, so I might not even need the resistor. It could use just an SPDT, as long as it's break-before-make.  Still need to know which I'm getting before I order...

Edit - Thanks, though!  This gives me more confidence that I can make it work one way or the other.

While we're here:  I have seen a couple of tutorials and examples of people hacking these simple, cheap little controllers.  They always show them just tacking the new pots onto the lugs of the existing ones.  I can't for the life of me figure out how that works except on accident.  From what I could tell, they are actually wired as voltage dividers.  Maybe if the old pots were permanently set exactly in the middle, but...  If anybody wants to speculate on that, I'm interested to listen.  For safety's sake I'm just gonna snip the things out.

[merlinb]

I knew I should have just drawn a picture to begin with 

[ashcat_lt]

I knew I should have just drawn a picture to begin with 

Oh.  Duh!  Thanks again.

[PeterPan]

You said that in one direction the switch connects to +5, and in the other it connects to ground, and... you'd like it to be momentary, event though it holds its position. This being the case, there is a slim chance (but one you owe it to yourself to try) that a regular SPDT switch (that holds its position) may work if you connect a capacitor between the center position and the circuitry it goes to. There are variations of this, of course, because without the load conditions being known (what the +5 or GND connection "drives", its hard to know how large a value of capacitor will work, or whether additional load resistance (between GND and the capacitor side opposite the switch) will be needed. And yeah, there is the consideration of using some diodes in this rupe-goldberg idea to prevent short voltage excursions in excess of VCC or negative spikes below GND to adversely affect the sensing circuit. But the bottom line is that it just might work! 

[ashcat_lt]

You said that in one direction the switch connects to +5, and in the other it connects to ground, and... you'd like it to be momentary, event though it holds its position. This being the case, there is a slim chance (but one you owe it to yourself to try) that a regular SPDT switch (that holds its position) may work if you connect a capacitor between the center position and the circuitry it goes to. There are variations of this, of course, because without the load conditions being known (what the +5 or GND connection "drives", its hard to know how large a value of capacitor will work, or whether additional load resistance (between GND and the capacitor side opposite the switch) will be needed. And yeah, there is the consideration of using some diodes in this rupe-goldberg idea to prevent short voltage excursions in excess of VCC or negative spikes below GND to adversely affect the sensing circuit. But the bottom line is that it just might work! 

Thanks for this.  I think I've got it from the couple replies before yours, but I'll keep it in mind.  Just want to make sure that everybody understands that this is essentially a rehouse of a commercial product (Korg NanoKontrol).  I know for a fact that each switch has both a diode and resistor already dedicated to it - I'd imagine they're there for exactly the reasons you've stated.

At this point I'm pretty sure that I know how to make it work one way or another.  The question at this point becomes "which method will the switches I buy support".  Could be missing something, though, so definitely interested in further discussion.

Anybody got insight re: parallel pots?