Urgent Help Needed! ABC Pedal

[Jaicen_solo]

Ok,
I've been asked by one of the bands i'm working with to create a custom pedal to switch between three sets of inputs to a single output.
Now this would be simple if not for the fact that only one input is needed at any one time, therefore, when switching inputs, the other two need to be muted. Does this make sense??
The diagram below (just worked out how to post them ) shows what i'm after. It looks so simple, but I can't get my head round it at all.
I've seen the remote bypass articles on Geofex using the CMOS logic looks like it could work, in conjuction with a set of relays. However, how do I make the 'on' state for each relay mutually exclusive?? Also, I'd ideally like to have non-latching footswitches to keep costs down.
Can anybody point me in the right direction here, as I need to get this project off the ground as quickly as possible.



edit: This is the page I'm most interested in. If I can find a way of selecting just one of the inputs at a time, i'd be laughing.
http://www.geofex.com/Article_Folders/cd4053/cd4053.htm

[niftydog]

the easiest way would be to use two switches, but that may not be the best way coming from the users perspective.

How do they want it to switch? Seems to me there's two options;

- three footswitches, one for each input
OR
- one footswitch that toggles through each input (perhaps with some LED indication of which input is active)

either way, you're going to need momentary footswitches.

I'll think about it some more, but please let us know which switching arrangement they would prefer.

[Aurin]

Don't they make connect before disconnect type switches? Might work for what you need... If you can find them heh =) ... Lame post with little help, I know, but... well, maybe it's a start.

[Bagge]

I just happened to find this thread. Don't know if it will help.

http://www.diystompboxes.com/smfforum/index.php?topic=28035.0

Otherwise you would need inspiration from Lehle's routers.http://www.lehle.com/frameset.php?country=us&lang=en&A_Link=products

[The Tone God]

Doing it with relays and discreets would be tricky. You could use a 4052 and maybe a 2-bit BCD counter simply.

Andrew

[Mark Hammer]

Take a peek at the quad sequential switch article on my site (http://hammer.ampage.org).  It should be somewhere around page 7-9.  This is an old POLYPHONY article on a CMOS counter-based switch that lets you use a single momentary and step through each of several stations in round robin fashion.  The unit shows 4 steps but it is reduced to 3 or expandable up to 10.

[R.G.]

OK, take a look at http://www.geofex.com/Article_Folders/fxswitchr/fxswitchr.htm, down at figure 7. This shows how to connect up a 74C393 octal latch so it does a one-of-eight selector function when driven by footswitches. When you press a switch, the output for that output and only that output is made active. It also does one-of-two up through one-of-eight by just not connecting switches to the inputs you don't use. Then hook a relay to each output.

The relay connects one input to a common output when it's energized, and grounds the input when not energized. I believe that does what you're describing.

[Jaicen_solo]

Thanks for the input guys. I already have a technique for stepping through multiple steps using a decade counter, which might be useful but it's not really what i'm after.
Basically, i'd like to use three momentary switches to control some sort of CMOS logic which when engaged, will cancel whichever relay is currently active, and open the selected one. For this, i'll need a device that has a low output that goes high when selected.
However, if possible it occured to me that it may be useful to have input A high (relay on) as default, since I intend to use non-latching relays. In the event of a power failure, at least one input can then be used. I'm not averse to using IC's, but I don't want them in the signal path. There's no reason that shouldn't be passive, a ground loop issue is never going to arise with this design and the relays I plan to use have a 120ohm resistance, which is less than a 20' cable. I should mention this is only ever going to be connected to the input of a Vox Tonelab, don't suppose it's important.

R.G, you posted while I was typing! I'll check that out now, looks promising! :> Oh, by the way, you got namechecked in Guitar Buyer & Bass (again) in the uk, nice work!

[niftydog]

the relays I plan to use have a 120ohm resistance

That'll be the coil, not the contacts... I hope!

[Jaicen_solo]

R.G, you are a legend! I think i've absorbed most of what you were trying to get across. I've cobbled together a schem showing one input, could you cast your eye over it to make sure i'm understanding you correctly?
From my suppliers, I can get a number of different 74C373's, such as 74HC373 and various other suffixes such as 74HC373N etc.. Will they all work in this circuit?? I don't really have a lot of CMOS experience!
It seems Nelson was right re: the relays. Contact resistance is a paltry 150 mOhms, coil resistance is 180ohms at 5v, that equates to 2.7mA of on current, is that correct? (Ohms law  ). I plan to run it on just 6v, to match the IC's, so I guess it could be run on batteries with such a low current draw. I doubt the IC's and LED's will take it above 20mA. I'll probably run via a power supply just to be on the safe side.
Anyway, the relays are here: http://www.maplin.co.uk/Module.aspx?ModuleNo=37494&TabID=2&WorldID=&doy=1m11

I could use SPDT relays, but the price is the same. I was planning to use the other half to drive the LED's, but I guess that's not necessary with your design! Thanks for the help, if the schem's right I'll try and make up a layout tomorrow evening.

EDIT: I just checked the data sheet for the 74HC373, it seems that should work. I also noted that it uses an oe pin, could this be used to mute all the outputs by taking pin 1 high??

[R.G.]

I can get a number of different 74C373's, such as 74HC373 and various other suffixes such as 74HC373N etc.. Will they all work in this circuit??

Unfortunately, no. Only 74C373's will work reliably in this exact circuit. The suffix is usually package specific. You want the DIP package and the "74C". 74HC and 74HCT parts will  not work without fiddling with RC timing delays on the inputs and clock. The 74C type will.

coil resistance is 180ohms at 5v, that equates to 2.7mA of on current, is that correct?

No. 5V/180R = 27.8ma, not 2.7. You missed a decimal place there. The LED will need between 1ma and 20ma. The actual CMOS logic chips use so little current that you can assume zero. You'll need a power supply because of the relays, most likely.

I also noted that it uses an oe pin, could this be used to mute all the outputs by taking pin 1 high??

Possibly. But you said you wanted one input to be active when there's no power. You can do that by wiring the relay pone and throws backwards and then using an inverter on the one channel so that the relay for that channel is on unless actively turned off. However, that would be backwards from your "one channel defaults on" objective. You can do one but not the other with the simple circuit I've shown.

[Jaicen_solo]

Thanks R.G, I guess I was not paying proper attention. It was 2am or something stupid here when I posted that (re: current draw).
I found a MM74C373, will that prefix work?? If not I guess i'll have to look for something else, having a hard time finding that particular logic chip for some reason.
I think that the passive on idea is a non starter actually. It's not really necessary, can always swap some plugs over I guess.
Is there some way I can alternate the ouput channel, to swith any input between two outputs (tuner/amp)? I guess I could use a dpdt?

[R.G.]

Hey, we all miss decimal points at times. I know I do.
The MM74C373 will be just fine. It's the "C" in 74C that makes it work. The 74HC and 74HCT are different logic families.

That particular circuit relies on a quirk of the timing specs of the 74C373 chip to work. But using that quirk works remarkably well. Use something like 74C14 or CD40106B for the schmitt trigger inverters. Those are more readily available.

Is there some way I can alternate the ouput channel, to swith any input between two outputs (tuner/amp)? I guess I could use a dpdt?

Well, you only have six choices there - three input channels to two output channels. One 74C373 gives you the instant exclusive selection of up to eight choices. Looks to me like you could easily use one 74C373 and then have the outputs feed three relays to select an input channel and one output relay to select the output channel. Use diodes to avoid interference between the outputs of the '373. Pretty simple, actually.

[amz-fx]

Decimals? We don't need no stinking decimals!

http://www.faqs.org/docs/javap/c9/s1.html

oops, maybe we do....

-Jack

[Jaicen_solo]

Thanks R.G. I have some of 40106's lying around from my previous PWM synth pedal so I'll use one of those.
I think with regards to the output channel, it'll just be easier to switch it with a latching stomp switch, simply because one stompswitch costs less than a relay, momentary switch, led and a mosfet.
Whilst searching for octal latches to use in this application, I found this: http://www.maplin.co.uk/Module.aspx?ModuleNo=2395&doy=2m11D
If I was to use all eight of the 74c373 logic switches, I guess it would be more economic to use one of those to drive the relays, but how would I add an indicator led?? Can I drive an LED on the high output of the darlington array??

[R.G.]

The 2801 and 2803 make great relay drivers.

A question - why are you using 5V for a power supply? The 74C family CMOS and CD4xxx family can run on any voltage from 3 to 18. 12V is a little easier for relays, as the currents are lower. If there's another reason for 5V supplies, great, it works there. But you do have the option to use 12V. Not really a suggestion, just a design option you have. I would really prefer running things like this on 9V, but although relay manufacturers do offer 9V coils, they are usually special order only. 5V, 12V and 24V seem to be stock items.
The 74HC and 74HCT families do not work above 5V. In fact, they die.

one stompswitch costs less than a relay, momentary switch, led and a mosfet.

That may well be. It really depends on the user interface. Will the user be doing a set-and-forget on the output switch, or does he need to switch in the middle of playing? Set-and-forget is best for the separate switch, because you don't want to accidentally hit that on the fly. Switch in the middle is best suited to one-of-N. By the way, you don't need a MOSFET to switch the relay. An ordinary US$0.05 NPN transistor works fine. And the LED does not have to be driven from the 373. You can attach the LED/resistor across the relay coil so that when the relay coil is driven, so is the LED.

Driving LEDs can be done from the top or bottom. You can tie the LED/resistor to ground and turn on voltage from the anode side. You can also tie the LED/resistor to +V and pull the cathode side to ground. The LED doesn't care where the electrons come from as long as they're going the right direction.

[Jaicen_solo]

The 74HC and 74HCT families do not work above 5V. In fact, they die.

That was why I was planning on using 5v, I had mistakenly looked up the wrong datasheet. When I found the MM suffix chip I checked the sheet and noticed the 18v psu. The relays i'm using are 12v, so I'll probably go for that. Easy enough, but can I just put 9v in and expect it to work?? Current draw shouldn't be a problem as long as it doesn't exceed 100mA.
I had looked at using a 2N390. I couldn't see why it wouldn't work, so I'll probably use one of those. They can supply up to 100mA so that's more than enough current to drive a resistor. If I put the LED in parallel with the relay, would it still actuate?? I'm not sure if that's what you mean, but I guess if it has a limiting resistor of the same resistance as the coil it should be fine??

With regards to the output switch, It's just so it can switch to a tuner, muting the amp input at the same time. Easy as that really.
I'm going to download eagle for this PCB job, I don't think I can make this layout look professional enough without help

[R.G.]

The relays i'm using are 12v, so I'll probably go for that. Easy enough, but can I just put 9v in and expect it to work?? Current draw shouldn't be a problem as long as it doesn't exceed 100mA.

That won't work reliably. Relays are rated at a certain voltage because they wind enough wire so the resistance of the wire limits the current to the right value. A 12V relay has enough coil resistance to let the right amount of current flow so the ampere-turns of magnetomotive force are correct to pull in the magnetic armature. At 9V, the coil resistance doesn't let enough current through. Some 12V relays will pull in at 9V barely. But I would not trust it for something I was going to use on stage.

I had looked at using a 2N390. I couldn't see why it wouldn't work, so I'll probably use one of those.

I'm guessing you meant 2N3904, and if so, yes, it will work fine.

If I put the LED in parallel with the relay, would it still actuate?? I'm not sure if that's what you mean, but I guess if it has a limiting resistor of the same resistance as the coil it should be fine??

Sorry, I was not clear. You need to put the LED PLUS RESISTOR in parallel with the relay. The relay coil has either 0V across it when it's not activated or 12V across it when it's activated. You can put a series combination of an LED and a limiting resistor across the relay coil. When the relay is not activated, the LED/resistor has no voltage, so no current flows and no light comes out. When the relay is activated, the LED+resistor has 12V across it and the resistor limits the LED current to whatever you decide you want it to be.

By the way, you also want a diode in parallel with the relay coil, with anode to the low voltage side, cathode to the high voltage side. This lets the diode protect the NPN transistor from the inductive spike of the relay coil opening.

[PharaohAmps]

Possibly. But you said you wanted one input to be active when there's no power. You can do that by wiring the relay pone and throws  backwards from your "one channel defaults on" objective. You can do one but not the other with the simple circuit I've shown.

Could you put in a "dead man" relay, that would send signal to the switch matrix when energized, and to a dedicated output when not?  It would call for an extra relay, but it should work.  If you're building from scratch there's really no reason not to have it do everything you want.

Matt Farrow

[Jaicen_solo]

I hadn't thought of that. I guess I'll speak to the guitarist and see if he really wants that feature. If it was me, I'd have it included, but then I don't worry about cost issues. Also, i'm not making money from this, i'm doing it as a favour at cost price. This makes me less inclined to do a 'kitchen sink' job on it

Edit: I just took a look at this websit and found exactly what I want, this is what I want to build.
http://www.lehle.com/frameset.php?country=us&lang=en&A_Link=products