Popping sound when pedal is engaged

[Lino22]

So what pops when the pedal is disengaged? How is that road constrained?

[antonis]

So what pops when the pedal is disengaged?

I'm not aware of such a case..
Disengaged pedal means both IN & OUT are "on air" (floating) so no involvement in signal path..

P.S.
Older design pedals used a DPDT switch, grounding the effect at some point and passing its input to output (e.g. see Marshall Guv'nor original pedal..)
In such a case, pedal's input cap wasn't "isolated" from signal path..

[Lino22]

So why do we put a pulldown on the input?

[antonis]

So why do we put a pulldown on the input?

In case of input NOT grounded in bypass mode..!!!
http://www.muzique.com/lab/pop.htm


P.S.
I always wire 3PDT switch for IN and OUT grounded (in bypass mode), despite the possible existence of OUT pull-down resistor in the form of Volume pot ..

[Lino22]

But how does the input cap discharges to the output? It looks like i had a wrong idea ....
I understand we can either use a switch grounding or a pulldown.

[antonis]

But how does the input cap discharges to the output?

Sorry Lino but you lost me..

Input cap discharges via any path to lower than its potential..
Honestly, I can't get your query..

[bluelagoon]

okay, thanks Antony, for the your gracious further education. Need to hear these things a few times over sometimes before they sink in.
Think I have a better grasp on the pull down resistor concept, now.
Please excuse my blathering ignorance of previous posts.
Cheers.

[Rob Strand]

So why do we put a pulldown on the input?

Suppose you don't have a pull down resistor and the the pedal is left in bypass for a while.  The input cap eventually discharges due to leakage.   When it discharges you have 0V across it.

Now suppose you switch the pedal to effects mode.   The effect side of the input cap is at potential 4.5V (say for an opamp) and when switched the input side of the cap is connected to the input source.   The input source has a DC path to ground so connecting the input side of the input cap to the input source is a good approximation to connecting it to ground.  When you connect the discharged input cap to ground it temporarily pulls the 4.5V via bias down to zero volts.  The effect side of the input cap then starts to charge toward 4.5V via the bias resistor.   From the circuits point of view that process is like a -4.5V pulse on the input.  So it creates a very large thump.   Once the cap charges the DC on each side of the cap reaches an equilibrium and the AC input signal is reflected as an AC signal on the effect side of the input cap - just like the input cap is not there.

But how does the input cap discharges to the output? It looks like i had a wrong idea ....
I understand we can either use a switch grounding or a pulldown.

The input cap itself doesn't directly create a signal on the output.   The previously mentioned pulse occurs at the input.  The pulse then passes through the effect.  A loud thump at the input generally results in a large thump at the output by passing through the effect circuit.

If the input capacitance has low leakage and/or the PCB has low leakage once you switch the pedal once the capacitor charges up.  Then any further switch will cause no thump or very little thump as the cap more or less remains charged in the time frame you are playing around pressing the switch.   

When you add an input resistor to ground the thump occurs at power up.  From that point on the input side of the input cap is at 0V so when you switch between effect and bypass there is no pulse.

With a single supply pedal you will always get a thump on the input and the output at power up.  The input and output caps have to charge up (as well as a all the other caps in the pedal with DC across them).    If you don't have an input or output RESISTOR then the respective cap will charge up when you switch to effects mode and then you will hear a thump.

[Lino22]

Rob, thank you for saving my sanity. I understand now. 
Antonis, i am sorry if a confused you and thank you for your endless patience. There was a missing piece in the puzzle and i had to fill it in.

[antonis]

Antonis, i am sorry if a confused you and thank you for your endless patience. There was a missing piece in the puzzle and i had to fill it in.

No problem, Lino..

You lost me 'cause I wasn't sure if we were talking about pull-down resistors necessity despite in-out caps grounded (via 3PDT switch) when effect bypassed..

[bluelagoon]

Great explanation Rob, Thank you, I think its sinking in.

[Baran Ismen]

So, putting a cap serial to signal lane or directing to gnd makes a big difference here I see.

[antonis]

So, putting a cap serial to signal lane or directing to gnd makes a big difference here I see.

It makes THE difference..!!!

In former case it doesn't affect the signal at all where in later case it mutes the signal completely..

(of course, there are always adjacent resistances forming HPF with series cap and LPF with grounded cap..)

edit: Sorry, you're talking about poppin issues..

[DrAlx]

I have the M-vave Mini Universe pedal and the exact same problem that Baran described in the original post on this thread.
I'll call this a "start up" problem, where a loud switch noise feeds into the effect and reverberates.
This "start-up" problem continues for a while whenever the effect is turned off then on again using the footswitch, but after some time (say a minute or so), this problem goes away.
However, there continues to be switch-pop noise using the footswitch, but it doesn't seem to feed into the reverb anymore like it does at start-up.

I opened the pedal up.  As far as I can tell its just a 2 layer board, and I managed to trace a little of the circuit out near the footswitch.

First thing to point out is that neither of the LED pins is connected to the footswitch.
In fact I removed the LED completely and found it does not affects either the "start up" problem, nor the remaining switch-pop problem. It is just as noisy as when the LED is in place.
First pic is a close-up of the top of the board near the switch after I removed the LED.



Next pic is my sketch of what I could make out for those tracks going to the footswitch.  Note the sketch is from the component side of the board, not the side where the footswitch is soldered.




I didn't make much progress with the lingering switch-pop problem, but have managed to solve the "start-up" problem.  It's simply a case of making sure that the input from the jack is permanently connected to the effect input.  i.e. solder two pins on the footswitch together as follows.



Doing this means you don't have true bypass anymore, but I didn't notice any "tone suck", even when my guitar (with passive pickups) was directly plugged into the pedal with nothing else in the signal chain.

[DrAlx]

EDIT:  I think I have solved both the "start up" problem and the residual switch-pop now.
I just added 100k between the "effect out" pin and ground in addition to the other hack I described in my previous post. In other words, I have this now



Output impedance of the effect without the 100k was likely of the order of 100 ohm minimum from R17 (I don't know the value of C48).  I think 100k is probably OK value to use and reckon it wont change the output impedance by much.


BTW, the other suggestion that was shared in a video further up the thread and tried out by Baran (i.e.  soldering 220k between "effect out" pin and ground, and also 220k between "effect in" pin and ground) will cure switch pop, but not immediately.  In other words, that suggestion does not solve the "start up" problem with switch noise going into the reverb.   That is probably why Baran experimented with lowering those values in order to solve the "start up" problem, but that is not a good approach to use.  You end up having to use too low a resistor value on the input side so it can start to suck tone.

[Baran Ismen]

EDIT:  I think I have solved both the "start up" problem and the residual switch-pop now.
I just added 100k between the "effect out" pin and ground in addition to the other hack I described in my previous post. In other words, I have this now



Output impedance of the effect without the 100k was likely of the order of 100 ohm minimum from R17 (I don't know the value of C48).  I think 100k is probably OK value to use and reckon it wont change the output impedance by much.


BTW, the other suggestion that was shared in a video further up the thread and tried out by Baran (i.e.  soldering 220k between "effect out" pin and ground, and also 220k between "effect in" pin and ground) will cure switch pop, but not immediately.  In other words, that suggestion does not solve the "start up" problem with switch noise going into the reverb.   That is probably why Baran experimented with lowering those values in order to solve the "start up" problem, but that is not a good approach to use.  You end up having to use too low a resistor value on the input side so it can start to suck tone.

Hey Alex,

It's so surprising for me to see someone else also trying to find a solution for this, and thanks so much for your contribution to this matter, mate!

I've ended up with 22K of resistors between in - gnd and out - gnd, but could not find any solution yet for delayed engage of effect (a couple of milliseconds, maybe 10-15ms). As you mentioned also, the resistor trick (at least in my scenario) does not solve entirely, but just helps with the popping to decrease its amplitude. And yes, removing the LED does not help with anything at all, so weird.

However, I had found out that shorting middle and middle-up legs (as far as I remember) of 3DPT yields with an instant switch into the effect, but LED stays on permanently. Also when done so, there's no need for anti-pop resistors at all, as the transition between effect and bypass is simply flawless and also the reverb decay/trail does not get affected in the background.

See, when you dis- and reengage the pedal with a moderate amount of reverb effect going on, the decay or trail effect resets itself. But when done like this (shorting middle and up-mid), solves this also. I think there's something quite wrong with the circuit itself as it's designed to ground the input or output (or both) when the pedal is disengaged. Right now actually, it's not really a true bypass one to me, as the effect does not run in the background like a phaser or chorus pedal.

So you say that, you solved these both popping and delayed engage issues or just the popping that bleeds into the effect? A bit confused here because from your words I understand that you're actually mentioning the same thing as the "start-up" problem and switch pop, as to me they both are results of a common problem.

Also, I suggest you make the trials for the popping issue by removing the DC power, waiting a couple of seconds, and replugging, so that the caps are drained and the pedal resets itself electronically As you also mentioned, the longer the pedal stays powered, the lesser the popping effect becomes regardless of the resistor mod.

[DrAlx]

I've ended up with 22K of resistors between in - gnd and out - gnd, but could not find any solution yet for delayed engage of effect (a couple of milliseconds, maybe 10-15ms).

I never noticed a problem with "delayed engage". But reverb is a delay based effect where it takes time for signal to go from input to output, so isn't this normal?  I tried this test:
Put mix knob to 100% wet so I can hear only the effect, and used Room reverb with minimum decay.  I pluck guitar string, and I don't hear the wet effect till a short time later (let's say 15ms).  If I do the same thing with LoFi reverb I hardly notice a delay at all.  The effect only "starts listening" to the input when its engaged, not before.  I've had another reverb work like that.  It's not necessarily bad, but may not be what you want.

However, I had found out that shorting middle and middle-up legs (as far as I remember) of 3DPT yields with an instant switch into the effect, but LED stays on permanently. Also when done so, there's no need for anti-pop resistors at all, as the transition between effect and bypass is simply flawless and also the reverb decay/trail does not get affected in the background.
See, when you dis- and reengage the pedal with a moderate amount of reverb effect going on, the decay or trail effect resets itself. But when done like this (shorting middle and up-mid), solves this also.

Yes, if you short out the middle and bottom (not top) leg of the 3PDT you can get rid of the "delayed engage".  But I found that this on its own did not solve the problem with "start-up" noise.

I think there's something quite wrong with the circuit itself as it's designed to ground the input or output (or both) when the pedal is disengaged.

Er, no it doesn't.  Maybe you were trying to measure resistances between legs of the 3PDT but forgot to put jacks into the sockets?  Without a jack, a socket will mechanically ground it's signal pin.

Right now actually, it's not really a true bypass one to me, as the effect does not run in the background like a phaser or chorus pedal.

That's not what is meant by true bypass as I understand it, or at least how its marketed.  True bypass means that you have input connected directly to output, with no other circuitry connected to the signal line and potentially "sucking tone".  If the effect is always on in the background as you describe in the phaser/chorus example, then the effect is always connected to the signal line, and not true bypass.

So you say that, you solved these both popping and delayed engage issues or just the popping that bleeds into the effect? A bit confused here because from your words I understand that you're actually mentioning the same thing as the "start-up" problem and switch pop, as to me they both are results of a common problem.

What I call the "startup problem" (i.e. switch-pop going into the reverb) and the "switch pop" problem itself are 2 different problems as far as I can tell, though there is some relation between them.
If you do the first hack I describe of shorting out the two pins on the bottom left of the 3PDT but use no bleed resistors at all, then the start-up problem is gone (i.e. you don't have to wait for a minute or more from pedal power up for the switch-pop to stop bleeding into the reverb.  It's gone in much less than a second after power up).  However, switch-pop continues to be there.

Also, I suggest you make the trials for the popping issue by removing the DC power, waiting a couple of seconds, and replugging, so that the caps are drained and the pedal resets itself electronically As you also mentioned, the longer the pedal stays powered, the lesser the popping effect becomes regardless of the resistor mod.

Yes I have been doing all tests like that.  I went back to the pedal just now to test some more.  Listening carefully with headphones, even after I have done by my 2 hacks of shorting the input pins, and adding 100k bleed on the output, there  is still some switch noise but it is very low.  Adding bleed input resistor (I tried 22k like you) does not reduce it further, but it does filter things at the input (i.e. I hear reduced hiss).  That's why I say 22k sucks-tone.

I tried shorting the middle and bottom pins so the effect (and LED) is always on in the background.
That did reduce the remaining switch pop a little further, but I would not call it completely gone even in that case.  But then I am testing with high gain and headphones.  For practical purposes I think the mod I have is fine.

[Baran Ismen]

I never noticed a problem with "delayed engage". But reverb is a delay based effect where it takes time for signal to go from input to output, so isn't this normal?  I tried this test:

Not entirely. If you watch this video of mine, you'll understand better. I focused on this matter after 11th minute. There's really a gap without any sound then the process of DSP begins, and in this very short gap, there's no sound coming out at all.

https://www.youtube.com/watch?v=RiucQMcivC0

And yes, you're right, middle & bottom mid lug, not the top, misremembered.

[DrAlx]

Hi Baran,

I used Audacity to see if I could also see a gap in the output that you showed in the video clip.
I tested with using these settings...

Decay: Maximum
Mix: 100% wet
Param 1 : 50%
Param 2 : Maximum
Param 3 : Minimum

Using those setting with the Room reverb, I too see a small gap in the output when the effect is first engaged.  But with the Cloud reverb I don't see a gap.

Each reverb algorithm produces its output by mixing a set of delayed versions of the input signal.
Maybe the Cloud reverb (configured with above parameters) contains a very short delay component so you get something at the output very quickly, while the Room algorithm's shortest delay component may be much larger, hence the gap.

The only way to avoid that gap is to (as you say) short the centre pin of the DPDT to the bottom pin so that the effect is always on in the background.
But you should probably short the two inputs pins together as well in that case (as I do in my mod), so the effect is not only always on in the background, but also always listening to the input.

[Baran Ismen]

Hi Baran,

I used Audacity to see if I could also see a gap in the output that you showed in the video clip.
I tested with using these settings...

Decay: Maximum
Mix: 100% wet
Param 1 : 50%
Param 2 : Maximum
Param 3 : Minimum

Using those setting with the Room reverb, I too see a small gap in the output when the effect is first engaged.  But with the Cloud reverb I don't see a gap.

Each reverb algorithm produces its output by mixing a set of delayed versions of the input signal.
Maybe the Cloud reverb (configured with above parameters) contains a very short delay component so you get something at the output very quickly, while the Room algorithm's shortest delay component may be much larger, hence the gap.

The only way to avoid that gap is to (as you say) short the centre pin of the DPDT to the bottom pin so that the effect is always on in the background.
But you should probably short the two inputs pins together as well in that case (as I do in my mod), so the effect is not only always on in the background, but also always listening to the input.

I did the mod as you suggested Last night, and its all good, thanks for your help, mate!

I also seek ways to fix the always on Led when 2 middle lugs are shorted. If i remove the Led from there, i ll need to find some other place to solder so that it Will be on when switch is on, as in current config theres no way to do that.