Switch popping - usual fixes not working - fixed

[Willybomb]

Gudday all.

You might remember my Magnum Opus build.  You might not.  It can be found here:  http://www.madbeanpedals.com/forum/index.php?topic=30309.0

Anyway.  I'm having problems with switch popping noise, but I'm not getting it in the usual on/off switch - I'm getting it on both the other two footswitches, which basically just swap between 3 sets of pot and soft clipping options.

Both are 4pdt switches and have common anode tri colour LEDs clipped to 2 colours (red/green, red/blue), and I've put 47uf caps to ground off both 9v feeds at the switches following their 8k CLRs.  As one side of the LED is always on at these switches, I don't seem to need to worry about another CLR after the cap.  The On/Off LED is also a clipped tri colour, but it is set to red only, and I blew this one when I didn't use the additional CLR after the cap.  After replacing the LED I used another 2k following the cap and all is good there (but I don't recall popping on that one anyway).

So... what can I do about this?  I'm open to the idea that it's mechanical switch noise on these Tayda 4pdt and I may have to replace them...  It's bugging me quite a bit as this is a killer sounding build...

Willy.

[jubal81]

Any chance we could see a schematic? I've used resistors across switch contacts to stabilize DC levels, but I'm not exactly sure what the situation is in your build.

[Willybomb]

I'm using this, minus the 390r, using a 47uf.
http://www.muzique.com/lab/led.htm
Second diagram.

On a switch pole, the red would go to lug 1, earth to lug 2, red to lug 3.

[EBK]

I'm going to wildly speculate here (because I still don't have a complete picture of what you are doing).

Try redoing the ground connections to the offending sub-circuits.  Specifically, run dedicated wires directly from your DC jack, without connecting to other things on the way.  Break corresponding ground traces/connections on the PCB as necessary to do this.

[jubal81]

I don't think its the indicator lights that are popping. I think it's the clipping circuit changes and a DC mismatch or shift. To combat this on a build, I used 470K resistors across the lugs so the DC is never impacted by flipping the switch. If you post a schema, I could indicated where.

[Willybomb]

Ok, here's the schematics/drawn layouts of what's going on.  I hope they're making sense.

The first pic shows the connections between the two 4pdt switches and I've included the caps that are currently attached.  They're actually on the 9v cable though, before the LED (ignore my polarity marks, I've probably got it wrong) as they're tri colour LEDs. Needless to say, the caps are going from the 9v directly to each lug, not from lug to ground.  The caps are going from the LED 9v feed, directly to ground on the 4pdts.

The second pic shows the gain pot wiring.  The 4pdts are switching between pots AND soft clippers

The 3rd pic show the volume pot switching.  This connects up to Parasit's bandaxall vero layout, replacing the trimmer with a volume pot, switching between 3 different pots....  I've included that layout too.

I haven't included the on/off switch as that's not popping...

[Cybercow]

If you're switching components in and out of a live circuit, there is going to be a "pop" because a stompswitch is not a "make-before-break" style of switch. The split second that a pair of components are switched in a live circuit, there is nothing connected in that short span of time. But in that split second, the potentials change and when a switched component is suddenly re-engaged, that slight voltage differential is released as an audible "pop" because the audio signal path is momentarily broken and then suddenly remade with different value(s). A stompswitch is a "break-before-make type of switch.

Swapping clipping diodes typically doesn't "pop" because the signal path is not "broken" during the switching process. Switching pots with a stompswitch in a live circuit momentarily breaks the audio path, a voltage differential then accumulates in that split second, and then when the connection is remade, that differential has to go somewhere; and it does it so rapidly that it injects a "spike" into the audio signal path.

Bypass switching typical doesn't (or at lest shouldn't) when there is a high-value resistor (typically 1M Ω) to ground that shunts the momentary differential.

A "make-before-break" type of switch avoids breaking the audio signal path.

Here are a couple of links that explain 'bypass switch pop' (however, I think your situation is different):

http://www.muzique.com/lab/pop.htm

https://www.mrblackpedals.com/blogs/straight-jive/6629778-what-really-causes-switch-pop

http://www.geofex.com/Article_Folders/box_pop.htm

Hope this helps.

[Willybomb]

Thanks, is read those before but Jubal's comments about resistors across the switch lugs and your comments on the the make or break connections might be the trick.

I don't get pop on the bypass, just the channel switching.

[Cybercow]

WillyBomb - I recognize your issue was not directly bypass-related, but thought to toss those articles out there. Switching components in a live circuit's audio path almost always calls for "make-before-break" consideration. You might want to consider a relay switching configuration to solve for such popping. Some relays come in 'MBB' and 'BBM' options. Just adding high value resistors across the lugs of the switches are not likely to solve the issue in this situation.

[Willybomb]

Yeah, I tried the resistors (had some 510k, 470k lying around) across the lugs to no avail.

[Willybomb]

Ok, so I've fixed this by setting the Baxandall vero trimmer to a set value (might be 500k, can't remember), and set the switching up to switch out the actual volume pots (instead of that trimmer).

The pops are gone, and there's no resistors or capacitors going to ground or anything like that on the switches.

[Cybercow]

Glad to find that you've solved it.