XLR A/B Switcher

[Sharpegaz]

I am looking at creating an XLR A/B switcher so i can send a single vocal mic to two outputs, one a standard clean signal and one to a vocoder.
I was looking to build it with a 3pdt switch, and with the output section i want it so that only one output is active at one time, so am looking to make sure that i short out pins two and three on the unused output so there is no hum or noise.
I would also be looking to make sure that the click when switching between the two was as less audible as possible.
It would not need to have any indicators IE Led etc
I was basically hoping for some ideas and schematics that would help me be able to make this up although there are lots of designs for standard Jack A/B switchers there are none i could find for an XLR one

Any Help would be great cherz

[MrStab]

hi Sharpegaz, welcome to the forum.

i can't think of a way you'd ground both signals on both channels with just a 3PDT switch, unless you use active circuitry, in which case you could probably just operate everything with a SPST switch. if you want it to be entirely passive (in so far as signal alteration), you could use relays. i'm not entirely sure on any possible loading effects from using transistors in their off-state on an unbuffered signal, but using FET or NPN switching is probably something you'll have to look into anyway.

popping should be minimal if any prior circuits aren't causing problems, but i have heard that 3PDT footswitches cause a lot of mechanical "clunk" due to the large surface area of the contacts. i haven't noticed much difference but i mostly dabble in DPDTs so can only judge based on the one i have.

just my thoughts. i haven't worked too much with XLR switching personally.

cheers!

[bluebunny]

My immediate thought would be to use CMOS switches, and use only an SPST switch like Grant suggests.  Read about CMOS Switching at GEOfex for ideas.

[Sharpegaz]

Thank you both for the information greatly appreciated. My initial thoughts for the grounds was just to send them to the casing of the project box.
With the CMOS switches is there a diagram anywhere on how i could run it with the XLR. I have made a few of my own guitar pedals of schematic but nothing with XLR from scratch.
Cherz

[MrStab]

^ that's pretty much how you do grounding, but i got the impression you wanted to ground out the unused channel upon switching. if that's not the case, then you could easily use a DPDT switch (just use one set of poles per signal from the cable, and connect all grounds... though some folks prefer more complex XLR grounding options, e.g with a transformer to isolate).

i can't think of any XLR-specific switching schematics, but if you get a CMOS switch working (such as in the GEOfex link Marc posted), you just need to figure out which wires carry signal and do what you want with them instead of putting them in a typical guitar switching arrangement

[Sharpegaz]

Yeah good point i was aiming to short the unused channel to minimise noise and hum. The issue i have been finding is schematic i did have a look over what was suggested but wanted to keep it as simple as possible with minimal inserts along the signal chain i was kind of looking at maximum a couple of resistors to stop the popping on the switch. Although i would be happy to compromise and say have channel a with  minimal insert but channel b with what was necessary to make a clean circuit as the second chanel is less cruicial but necessary. Thankyou for all the help and suggestions as far

[MrStab]

you MAY be able to ground at least the hot OR cold side of each unused channel using a 3PDT with this method:



but it's late and my brain works even less than usual at this time, so someone might wanna double-check that. also, i dunno if only grounding the positive OR the negative would be worth it (or have any side-effects).  the "X" shaped wires overlap, they're not joined together.
otherwise, just ignore the black wires & use a DPDT.

so long as any active circuits that go through this switcher are designed properly (such as having a pulldown resistor), the only pop you'll have to worry about is mechanical (usually not an issue, but inevitable). i've heard that 3PDTs are noisier than DPDTs, because they have a wider surface area to switch.

let me know if i've just confused you further lol, i tend to do that.

[MrStab]

this is probably overly-complicated and imperfect, and you might wanna keep DC off the switch and use PNPs instead, but here's one idea to shunt all unused signals to ground using NPN transistors:

you wire up the rest of the switch as you normally would, but the last set of poles "activates" those transistors with the current and the signal is sent to ground. there are probably more components needed to make it more ideal, but that's the rough idea.

Note: make sure the 9V switch is wired the right way, so it doesn't ground out the channel you wanna hear!

[Sharpegaz]

Mr stab you have just confirmed one of the diagrams i drew up but in a more logical sense that would activate the ground properly before i had pin two from input xlr going to Middle and then the outs top and bottom with a connection in with them linking to central pin and then i had the same cross over which meant i used all pins with no way to ground off the switch so that idea was what i was thinking but in a more promising manner. Hopefully the main side will stay clean and the secondary will have a low noise floor. In my initial thoughts i was thinking do i have to short both pins two and pins three for this to work.
Thanks again for all the help and advice

[MrStab]

the only side-effects i can theoretically see are, as you say, you may lose the benefit of balanced cancellation on the unused channel - but that'd probably only be an issue if it bleeds through for some reason (high-gain distortion etc.). happy to help! good luck!