switch pop question

[tempus]

Hey all;

A few years ago I built an FX switching system based on the one at the Geofex site (http://www.geofex.com/Article_Folders/fxswitchr/fxswitchr.htm. I made some modifications to include an A/B bank switch. Also, I have the unit set up so that it can select any combination of (pre-programmed) effects with one stomp. So, I can go from, say, bypass, to clean channel with delay and chorus with only one switch. It works beautifully (thanks RG for posting this at your site) except....

I've never been able to get the switching pop free. I've used relays, CMOS switches, BS170's with the control voltages ramped up, you name it, but I always get a pop when I switch. To make matters worse, sometimes it's a loud pop, other times it's barely noticeable, and it often varies even on the same switching operation. I'm starting to wonder if the problem is in the switching circuitry. Is it possible that the switching is somehow working its way into the audio path? The box containing the circuitry is shielded, but should the wires connecting the pedals be shielded also?

I'd be grateful for any help or ideas that any one may have. I've read some of RG's articles claiming to get NO clicks or pops of any kind. Since I've yet to experience this, I wonder if there's something fundamentally wrong with my design or layout or power supply or...

I can post a schematic of the relays board, but I'm having trouble getting that to work.  How are you guys getting pictures in your posts?


Thanks

[GibsonGM]

Usually there is some kind of DC potential difference involved in switch popping, barring any defects in the switch itself.  You're going from a level of higher or lower potential than the circuit you switch into, and it's heard as a 'pop'.    Search the forum for "switch pop" and you should find a bunch of posts that have some fixes for this - it's about keeping both ends of the switch at the same potential.  Pulldown resistors etc.  I'm sure with some looking, you'll find the answer!!   

To post pics, you have to host the pic somewhere, like Photobucket. Generally a smaller kb size pic is best (100 kb or so).  Then copy the link location at Photobucket, click the button when posting that says "insert image", and put the URL to your pic in between the (img)(img).    Saves space on the forum if the pic isn't hosted here....

[tempus]

Gibson thanks for the reply. I'll search the forum again but I've been trying to fix this problem for a couple years now (no exaggeration) so I'm pretty well versed in the potential problems of switching. I've looked closely at the DC issue. All my inputs and outputs from each pedal measure within a 10th of a millivolt of 0vDC. I was wondering if there might be something specific about my setup that's causing this (seemingly) unsolveable problem.

Thanks

[jaysg]

I can't speak to the design you're dealing with, however, if you have ground referenced LEDs as indicators, that can be a problem.

[tempus]

Thanks for the reply jaysq.

I thought of that too, and disconnected all the LEDs. This made no difference.

[GibsonGM]

And it ain't the switches themselves, eh?  Really weird!  You should go to the man himself on this one, I think (R.G.)!  lol.   It's gotta be something simple, as you know, but it can really mess with your head trying to track that 1 simple issue down, huh?

[tempus]

Thanks for your reply Gibson.

What do you mean by 'the switches themsleves?' Is it possible that the switch itself can induce a pop into the audio signal path? I'm using momentary contact switches (those little round kind that are on VCRs and TVs and stuff) to ground an input on a 4049 inverter. This sends a logic high to a latch, which opens or closes a relay.

Thanks

[jaysg]

I'm not really up on modern relays, but for switching of heavy current AC there are ones called, zero-crossing detection relays.   They don't switch until the sine wave is at ground.  I don't know if these are available foraudio applications, but that's probably the issue.   There's a discontinuity, hence pop, whenever you turn off a signal that isn't right at ground.

[tempus]

I'm becoming certain that it must be something like that jay. Today (don't know why I didn't try this before) I unplugged and disconnected from the DC power supply all the effects pedals and connected the input jacks to the output jacks so the signal could flow through. When I stomped on different buttons (that would've normally switched fx in and out of the loop), I got the same amount of switch pop as before.

As for zero crossing relays, I have tried CMOS switches as well, which gave me the same result. I hope that bit of info about disconnecting the pedals helps some others who've been trying to hunt down the source of switch pops. I'll have to concentrate my efforts on the switching circuitry itself now. Any opinions on the shielded wiring subject? Is it possible that I'm picking up some switch glitch because the wires connecting the jacks to the relays aren't shielded?

Thanks

[GibsonGM]

I'd say this is one to send directly to R.G.!  Lots of things seem possible in the setup (although the idea is intriguing, and a good one too).  By "in the switches themselves" i was referring to a bouncing contact or something broken...since you're using like tactile switches, I'd go with Tempus' suggestion about the relays switching at different potentials.

If the jacks are grounded (or floating, as long as you're consistent), you shouldn't have trouble with pickup, I would think.  Are you using star-type grounding?  Not that it would cause pops to not use it, but it could make the whole unit quieter as I'm sure you know. 

I'd try taking a bunch of DC measurements at the switch terminals, try to see if it 'pops', and then goes to 0VDC.  That will tell you something!    Pulldowns might help if that's the case....do you have a schem you could put it?  I for 1 would like to see the setup! 

[LHC]

Hey, you might be interested in my "declicker":

http://www.bluemonk.de/~zi/public/declicker-pulldown-vactrol.png

When the input and the output are switched on, the two "pulldown"-resitors are only 2k small  --> the click is reduced. In less than a second the capacitor is loaded and the resistance increases to more than 20M --> no negative influence on the sound after the switching.

If you need an additional status led, just put the 9V on the pole of the third switch and a LED + resistor on the free throw.

yours
LHC

[tempus]

Here's a schematic of the relays switching section of my pedalboard:

.

Thanks for the declicker link too. I'll have to look at that a little closer when I have more time.

Oh by the way (I forgot to mention this earlier) the pops are HUGE. Sometimes they're just little clicks, which I can live with, but sometimes they are really loud, almost like unplugging my guitar from the amp.

Thanks

[R.G.]

Hey, you might be interested in my "declicker":

I've used an H11F1 LED-PhotoFET module for momentary declicking. I can mute the audio line for a millisecond or two. Same idea.

[LHC]

Hey, you might be interested in my "declicker":

I've used an H11F1 LED-PhotoFET module for momentary declicking. I can mute the audio line for a millisecond or two. Same idea.

Not exactly. Your Photo-FET with his resistance and capacitance is in the signal path. An you would need two of them (input, output).
My "declicker" with dynamic pulldowns is just in parallel with the signal and minimizes potential-differences. Whith more than 20M it is nearly invisible, when the signal is on.

Greets