Popping Pedals 401...Graduate Level

[Philippe]

OK...so you've got your 1M5 to 2M2 pulldown resistors properly mounted to the input & output caps + circuit ground & the pedal still pops obnoxiously.  Next (& after measuring the input/out jacks for any cap leakage/voltage) you disconnect the LED indicator to uncover whether a voltage surge is the culprit when activating the circuit...if the pop is now gone, the time-honored Jack Orman/AMZ solution usually does the trick (i.e. a delay from 9V to the LED via an inline 10uf to 22uf electrolytic cap (negative to ground) & 390 ohm resistor.

But if the effect still pops, what next?  Encountered this ongoing irritation the other afternoon while completing a DIY envelope filter (aka auto-wah)...turns out the dual opamp TL022 chip was the cause of the problem. After running down to RS & picking up a TL082 (the only one they had that was somewhat similar) the problem was eliminated. Glad that the IC was socketed.

*curious* as to what an opamp has to do with popping? Granted there are a lot of lousy/noisey chips out there (the 741 immediately comes to mind) but so far, I haven't the foggiest as to what differentiates a TL022 from a TL082 other than the model number.

[Toney]

 Phillipe, I have just been through that with a Rangemaster I built.... a Rangemaster! Nothing to it right?
Eliminated the well known possibilities but that prick was loudly popping no matter what I did. No dc on the ins and outs plus a grounded input. I added in and out cap resistors despite the switch input grounding, in case there was a DC change WHILST switching IE break before make. I then proceeded to remove all the "extras" LED, Cap switch but no dice.
Basically it just drove me nuts.

My best guesses:
Mechanical noise/cross talk IN the switch.
Leaky DC at the amp input -Vox AC30
Poltergeist activity...

[Quackzed]

i just installed a 'kill switch' in a guitar that has a treble bleed cap/resistor btween lug 1 and 2 of the volume control. the spst switch is between the volumes lug 2 and ground, pops loudly!!!
tried a 1M pulldown resistor on the switch , pop! no dice!
so i simmed it with and without the treble bleed cap, in falstads... with the treble bleed cap, some big pops and spikes, without ,much better...
with the bleed cap the bias seems to shift when the signal gets shorted to ground, so when i release the 'kill switch' the signal can be more positive than the center line and create a spike above the signal level... it then settles to the center line, but with repeated kill presses there are some nasty spikes!

[R.G.]

There is really only one way to have silent switching. That's to do the muting part of the signal that's being switched out and the enabling part of the signal that's switched in by fading out/in over a few milliseconds.

Well, OK, two ways. The other way is be be very lucky and switch both coming and going signals at a zero crossing. Even then, the human ear will insert a discontinuity.

Any time you suddenly change a signal, the ear perceives a "tick" of varying degrees. Metal switches switch **instantly** from conducting to not and vice versa, and that's hard to keep quiet. Especially when they bounce.

In addition to switching transients, you can make it worse by changing the DC level of the switch instantly, as with leaking caps, different DC levels and so on.

Also some more subtle ways. Audio circuits vary in how susceptible they are to noise on their power supply (or supplies) and on ground. In the audio biz, power supply rejection ratio (PSRR) is a term you hear quoted a lot; that's the amount that any noise on the power supply is attenuated before getting into the output. Good audio stuff has PSRRs of 40-100 db, bigger being better. Simple single-ended transistor circuits may have PSRRs of 3db. Worse, they may *amplify* power supply noise if it gets into their bias chains. That's why bias strings are decoupled with caps - to try to keep power supply noise out of what is, after all, another signal to be amplified.

Ground noise is also pernicious. Many circuits have 0db rejection of ground noise. Diffamps tend to be rated in terms of common mode noise rejection ratios (CMR). Really good differential or transformer coupled circuits have CMRs of 60-120db (!).

Power supply noise and ground noise is what the "LED tick" is doing. The LED current is suddenly changing either the power supply or ground voltage at the amplifier's sensitive points by suddenly changing. It's usually a wiring or wire routing problem.

Any sudden change in current in a current loop also broadcasts itself as either a magnetic pulse to be picked up by another current loop, or a radio pulse. This is why wires are twisted - it decreases the area of the current loop, which makes the loop less able to transmit and also less able to pick up other pulses.

[Quackzed]

thanks r.g. -your sig is particullarly apropriate for this subject-
with my killswitch i'm definately getting dc shift from the treble bleed cap as well as your garden variety 'tick'. -also some mechanical click from slightly microphonic pu's..

didn't mean to derail the tread, just wanted to get treble bleed caps on the list of suspects for pop!

if there was a simple mechanical solution... like a pushbutton potentiometer... unpressed =full resistance,pressed = 0 resistance ,and in between off and on
a 1/4" of resistive strip or somesuch that goes from one to the other...

[R.G.]

It's really tough to get quiet switching with a mechanical switch only. Sometimes it's fine, but the more you amplify it, the more the imperfections show up. If you can get power to it, the premium solution is to use either a slowed-down JFET shunting the signal or an LDR, although the LDR is maybe too slow.

Is there any way to get rid of the DC shift? Where's it coming from? Generally AC-only things don't show much click if there's not a diode in there somewhere.

[Toney]

If you can get power to it, the premium solution is to use either a slowed-down JFET shunting the signal

Ok I am going with this.
I am thinking of trying  a "hack solution" as the pedal is already fitted with a 3pdt and I have no momentary switch or suitable CMOS devices. So I'll try to use the 3pdt to soft switch a couple of JFETS ala this:



Please excuse me RG for "lifting" the image from your original PDF which can be found here:
http://www.geofex.com/Article_Folders/bosstech.pdf

Will try 2n5485 and SK30A as the JFETS.
Can you or anyone suggest suitable value for the Rslow and Cslow?

[Toney]

 OK I have put this together from the above schem snippet which should  slow-switch either the input or output as required in the "Clinton bypass" method.
Rslow and Cslow set the delay.



From my reading the most suitable JFETs are 2n5952, 2n5485, 2SK30A, 2SK117 but I guess anything within reason will work for testing.
Will work on a dual switch now.

This is a bit of post installation hack, so I will try mechanically switching the input via a 3pdt to the soft switch which hopefully will slowly ramp up and avoid the pop.

EDIT:
Updated image to include a 1m at the cap on the JFET's drain.
N.B. There is no cap at the source of the JFET based on the assumption that the first thing it will "see" is a cap at the effect's input. Need to remove any cap drain resistors.

[Toney]

Wiring would be as so:
I guess you could flip it around to switch the output and do the input mechanically if that helps.





The JFET board introduces a delay to the switching (hopefully) losing the "pop" - works for Boss...

[Toney]

I suppose its just as good using a Dpdt.

[kaycee]

Thanks for sharing, following this with interest as I have an op amp based distortion behaving exactly like yours

[merlinb]

turns out the dual opamp TL022 chip was the cause of the problem.

*curious* as to what an opamp has to do with popping?

It would be interesting to see the exact schematic you were working with. The opamp certainly cannot be the ultimate source of the problem- some secondary effect must have occurred while you were swapping devices.

[Philippe]

turns out the dual opamp TL022 chip was the cause of the problem.

*curious* as to what an opamp has to do with popping?

It would be interesting to see the exact schematic you were working with. The opamp certainly cannot be the ultimate source of the problem- some secondary effect must have occurred while you were swapping devices.

You are absolutely correct merlin. It must have been the weed speaking along with with some wishful thinking on my part as the effect still has a slight pop (albeit it somewhat less than previously). I upped the electrolytic delay cap from 10uf up to 47uf & the unit is only 'pop/click-free' with the LED fully disconnected.

Maybe it's one of the reasons (along with economics & additional labor) that some of the earlier mass-produced 'pre-clone' effects didn't have them...afterall, LEDs have been around for over 40 years now.

[Toney]

 Well if the LED isolated removes the problem that's great , time to use the millennium.

  LED dc shifts are upsetting things it may well be that the switch just doesn't "like" have 9v or ground dumped through it.
You could try another set of contacts on the switch (once you'd continuity tested all the contacts) but the millennium is the obvious call here.

[Toney]

 With the soft switch Vero above you can use it either on the input or the output to introduce a micro delay to the switching. I cant edit the post anymore but I realized this morning that flipping the JFET is all that required to convert it for output use without adding a cap (once you remove any pulldown resistors from the output cap). Actually I think it makes more sense to use it that way as it will choke off the circuit output and slowly switch it to the output jack, giving time for the mechanical switching of the circuit input and any dc shifts/dumping to settle before releasing to the output jack.
I wonder if it is possible to expand on this and do a hybrid of soft and mechanical switching for both input and output? That's to say; mechanically switch both the ins and out to two JFETS that are switched off, then slow on, using "normal" 3pdt wiring. The advantage is keeping true bypass in the off position without anything hanging off the line and affecting impedance. If I can use a 3pdt to present both input and output to the JFET caps without any noise it should be fine. The input and output will only "see" a cap with pulldown resistor and a closed JFET channels during the mechanical switching.

I am going to build that little board today as I simply cannot get rid of the tick with my Rangemaster despite it having no DC on the ins and outs - all components orientated correctly/voltages normal and good quality output cap that definitely not leaking. I have already changed out the switch and am using the millennium method. This little biatch will just not play ball. Its OK with my solid state testing amp but definitely not OK with my valve amp at playing levels.
Perhaps its the cheap-assed $2.99  3pdts I am using? The continuity is a consistent 1.2 Ohms with each contact. It must be a cruddy surface or time difference with the throws. I don't know and not knowing is irritating the hell out of me.
I will build the soft switch and try it out today.

[kaycee]

Others mileage may vary. I find the MBP very good for most usages, Fuzz Faces, various treble boosters etc. I do find occassionally that there are builds that it doesn't work so well on (popping on switching), sometimes changing the type of LED seems to help, and I find that the 2N7000 is the most reliable item to use.

However with high gain circuits I find that I get bleed through on bypass, and sometimes engaged with gain dimed, squealing from a feedback loop I assume? I put this down to not being able to ground the circuit input on bypass.

[Toney]

 If you are getting popping with adding the millennium then the diode  is probably excessively leaky, or if it a  Millennium C, the resistor value at the control is wrong. I learned this the hard way by rushing and not concentrating because its such a tiny build and annoyance at LED thumping made me want to quickly deal with it.
It can be measured from the control wire. It should be less than something like 4mV with a 1M resistor to ground at the effect output in the bypass position.
 More than than 10mV can charge the output cap and actually be the cause of popping, not the cure.
"bleed through on bypass"
I don't think the Millennium can cause that. Maybe excessive ground noise with a high gain circuit as you you said. You could short the input to ground and test.

[Toney]

 OK I built and tested the Soft Switch.
I think that hanging it off the output is the better method as when its off, it chokes off any noise within the signal path and then it turns on and releases slowly. I tested it with delay caps up to 1uF (C1) and it really slows the transition. Far too much but I was mucking around with it. Boss uses 4.7nF so that a good starting point. I also tried mechanically switching to the Soft Switch using "normal" 3pdt wiring and that worked fine for me. I suppose there are some conditions, such as unsolvable ground noise issues or switch contact noise, where it might be better to go with the Clinton Bypass type wiring but reversed with the Soft Switch on the output.
The board is tiny should be able to fit in most builds. This one is from the earlier version- little different from the final but same size.



Here is an updated Vero with regular 3pdt wiring. I have added another cap so it can be used in any position. Also I was wrong about needing to flip the JFET for use at the input. Further reading taught me a JFET channel conducts happily in either direction so the switch can be used at either input or output as is.




Alternative wiring using reversed Clinton Bypass method. It might be better in some cases.




To test it I set up a boost pedal with a voltage divider dropping ~0.9v after the output cap. It popped horribly going to the amp with just a 3pdt but with the slow transition of the soft switch, sweet silence.