Stomp 3pdt´s and dpdt´s..I´ve just had enough!!

[dschwartz]

Hi all..
I´m frustrated with 3pdt´s and dpdt´s failing over and over. I don´t know how my customers use them, but 90% of repair jobs on my sold pedals are related to stomp switches going bad, doing strange contacts or not working at all..
I even changed to alpha dpdt´s and they keep coming back to my shop..
I don´t know if my customers are allways bigfooted gorillas that like to jump over the switch or simply stomp switches are not made to last..

I´m seriously thinking on using simple pulse switches with a flip flop and a 4016..

What do you think?? has someone had the same troubles ?

[R.G.]

Welcome to the club. A major part of all of our warranty repairs at Visual Sound used to be broken stomp switches.

We spent a largish amount of money to change over to actuators and momentary switches as a result. So far no one has broken an actuator or the ALPs sealed dome switches we use.

I'm convinced that mechanical switches are what you do when you can't do anything else. Reliability? What reliability?

[aron]

Wow, I haven't had any of them fail on me. The old X wing DPDT ones all went bad a while ago. I replaced with 3PDT and never had problems.

[dschwartz]

indeed RG..reliability my A$#$..the bad thing is since switches are not reliable, customers think YOU are not reliable. I care about quality on my builds, so i´m going to go the extra mile and use actuators and electronic switching..it will be more work, more parts, but cheaper and very reliable..

[aron]

Daniel, let us know how it goes.

[Derringer]

I have a brand new X-wing that is intermittent

my usual routine to "warm it up" before a gig is to toggle it about 4 - 6 times ...

[doitle]

I've been working on a way to do true bypass with just a DPDT and a DPDT relay. It overall gives you 3PDT but you might be able to get a more robust DPDT and certainly a relay would be pretty hearty. Using one with a 9V Coil it should work fine with the power supplies already present in pedals.

I wonder what are some of the more exotic switching methods that could be used though... You could make the player wear a magnet on the bottom of their shoe/foot and put a reed switch up near the top of the case and use it to toggle a relay...

You could do capacitive touch sensing and make a foil plate on the top of the pedal as an antennae or something. You could use a sonar sensor mounted straight up so you just have to swipe over the top of the pedal to activate it. You'd need a little micro controller in there likely though for that to work... That actually sounds kinda cool though I might try to make that one.

[dschwartz]

my plan is to copy the flip flop part of boss pedals, and a 4016/66 for switching (fets are hard to get and expensive in chile). all this in a little "utility" board on the side of the box..
I am working on a pcb and so far is 2.4cm X 3cm..(i´m on the cmos part now)

[earthtonesaudio]

You guys that have experienced switch failures, did you ever investigate the cause?  Was it just normal use, damage during soldering, or something else?

If it was only heat damage, I suppose one could use sockets...

[dschwartz]

well, at fisrt i use to do autopsies to the switches..Most common fails were:
1- inside parts dont make good contact
2.- dirt or the switch losing lubrication
3.- over soldering (specially on x-wings)
4.- oscilating switches (at least the oscilation stopped when the switch was changed)
5.- Most common: mysterious fails that only happen to the user but works fine at my place (temperature/humidity/stupid user?).

[R.G.]

I've been working on a way to do true bypass with just a DPDT and a DPDT relay. It overall gives you 3PDT but you might be able to get a more robust DPDT and certainly a relay would be pretty hearty. Using one with a 9V Coil it should work fine with the power supplies already present in pedals.

Why? Adding more mechanical parts usually decreases reliability. And you can do true bypass with only a DPDT and the Millenium Bypass (see GEOFEX). If you're willing to spring for a relay, you can do true bypass plus indicator with only an SPST, no DPDT needed.

I wonder what are some of the more exotic switching methods that could be used though... You could make the player wear a magnet on the bottom of their shoe/foot and put a reed switch up near the top of the case and use it to toggle a relay...

You could do capacitive touch sensing and make a foil plate on the top of the pedal as an antennae or something. You could use a sonar sensor mounted straight up so you just have to swipe over the top of the pedal to activate it. You'd need a little micro controller in there likely though for that to work... That actually sounds kinda cool though I might try to make that one.

All of these have been used.

You guys that have experienced switch failures, did you ever investigate the cause?  Was it just normal use, damage during soldering, or something else?
If it was only heat damage, I suppose one could use sockets...

I do failure analysis on all our failures anytime the service guy notices a trend. I've had all kinds of issues pop up, from no internal lubrication, to contamination (ever see a beer-soaked switch?  ) to the actuator sticking in the bushing, to contacts no making on audio but working fine for higher voltages. We've also had "flat topping" where an overenthusiastic player broke the top button (and all the control knobs) off the pedal. That's one reason for the cast-in up-sweep on the V2 line of pedals. You can still break the knobs, but you have to work a lot harder.

And switching failures stopped when we went to actuators.

[Mark Hammer]

Clearly RG and Bob Weil have had far more cumulative experience with stomp switches than I have, but I'm no switch virgin either.

My own take on DPDT/3PDT switch failure is that, in addition to poorly-planned physical installation, it is both excessive soldering heat, and the goop/grease inside the switch that are to blame.  I've written about this frequently, but I'll do it one more time.

First, installation.  The plunger on the switch MUST be allowed to behave like a piston.  It has to be depressed straight down, with no lateral pressure, or else the switch structure is stressed.  Too often, people install the switch seated too high up (i.e., too much switch sticks out of the chassis), and located at a position on the chassis, which invites people to press it at angles other than 90 degrees.  To my mind, that shortens the lifespan of the switches, although to be fair, not as quickly as the other cause.

Now, the other cause.  The solder lugs on the outside are molded to the chassis, and have an inner exposed contact area.  There are rocker contacts inside the switch that pivot on the exposed part of the centre contact.  Press the switch and the rocker tilts the other way, making contact between the centre and the other set of contacts.  Press again, and the rocker tilts the other way.  During assembly, a small dab of grease is applied to the middle pivot point where the rocker contact sits.  This serves to keep the rocker from becoming unseated during assembly, but also serves to damp vibrations after assembly to minimize switch bounce.

When heat is applied through soldering, the heat is not only on the outside of the switch, but is conducted to the inside, via the solder lug.  If the lug you are soldering is in electrical contact with the rocker, it is also in thermal contact.  If too much heat is applied for too long, the heat is conducted from the outside along the rocker contact to where the dab of grease is.  If warm enough, the grease turns from thick to liquid and flows along the rocker contact, coating it.  The grease is non-conductive.  If it coats the rocker contact, it impedes continuity.

To my mind, this is a principal source of switch failure.  My rationale is that if commercial clients experienced the rate of failure that us hobbyists report, switch companies would be out of business in weeks.  So I'm assuming the switches are decent, but we use them improperly.  Call me naive.

There are a few things you can do.  One is to simply open the errant switch by prying the tabs, cleaning the crap off the rocker contacts, and re-assembling the switch.  I've done this successfully many times without problem.  It was during such a repair that I realized the nature of the grease problem.  If you run a service bench, though, that's way too much work, and I recommend it only for those who have time.

The other thing you can do is use careful assembly techniques when wiring up the switch, to make sure the grease never liquifies:
1) Make sure the switch lugs are clean, fluxed, and tinned before soldering connecting wires.  This will make the soldering of wires much quicker, letting you get in and out with the soldering iron.
2) Wait a bit between applying heat to each solder lug on the switch.
3) Make sure the wires you connect up to the switch are also tinned and ready.
4) Hook a bit of the exposed wire through the lug and bend it with needle nose.
5) Use your meter to identify which outside lugs are not presently in electrical contact with the centre/common.
6) Solder those lugs/wires.
7) Wait a few seconds for things to cool down.
8 ) Push the switch to remove contact with the other set of outside lugs, and solder them.
9) Wait a bit.
10) Quickly, solder each of the centre lugs, being sure to wait a bit between.  This will assure that all heat is dissipated from the inside.

RG et al are probably right to collectively say "Screw it.  I don't need that meticulous an assembly process if I'm going to make thousands of pedals."  The rest of us, though, turn to stompswitches because they are convenient for one-offs and small runs.  When you think of how many pedals companies like Dunlop and Fulltone sell with stompswitches, clearly it IS possible to have very few failures if you do things right.  Our collective challenge here has been to find out the optimum way to use them, in the absence of any help from the manufacturers.

[Toney]

I have seen many DIY jobs with chronic over-soldering. The stomp switches take longer to get a solder flow, especially with a wire attached as the large lugs act as heat sinks during the process. The X wings seem the most vulnerable to heat damage.
It's simple but the solution is to tin the lugs and the wire and use a hotter/larger iron for a shorter time, if available.
With the X wings, I solder the lugs , then using my handy-dandy lug poker (bamboo skewer) I clear the solder, tin the wire and then a quick burst and it's done.

[Solidhex]

Awesome info dudes

  I have a friend who was going through switches like crazy. Like Mark mentions he had a habit of stepping on his switches with his heel on the floor. Pushing sideways on the switch not directly down on it. Best thing to do is to mount it with only as much of the switch sticking out as needed. That and not placing it too close to the outer edges of the pedal.

[dschwartz]

ok, guys. After seeing tonegod´s article "wicked switches" ( http://www.geocities.com/thetonegod/switches/switches.html ), i decided to ditch the boss-type flip-flop (too many parts) to a 4069 flip flop (1 chip, 2 R, 1 C, nice!). I simulated it on live wire and seems to work fine.

I´m changing tonegod´s switching, shunting fx input when bypassed, but the layout for the 4066 connections is a nightmare, since i have to apply bias voltage to some inputs..looks like i´ll have to use some wire jumpers...i hate those.

questions about 4066 (hey, its my first time with them) :
- Is there no difference on which of the sides of the switch i apply Vref? or i have to apply vref to both sides to avoid pops?
- I´m shunting the fx In to Vref directly.. is that ok? my guess is if i shunt to ground it will pop.
- am i complicating too much, and just apply vref to the input and forget about the rest?

also, im thinking of adding a buffer in front of all that..a fet buffer maybe, but i´m worried about headroom..should i?

[R.G.]

If you are using a CMOS flipflop then you likely have logic signals of both polarities available. Why not use the Boss signal passing trick with JFETs, so you can put the JFETs out where the signals switch instead of routing them back into a single switching chip? Makes for simpler routing of signals.

[dschwartz]

If you are using a CMOS flipflop then you likely have logic signals of both polarities available. Why not use the Boss signal passing trick with JFETs, so you can put the JFETs out where the signals switch instead of routing them back into a single switching chip? Makes for simpler routing of signals.

That´s an option, but i´m thinking of having a kind of "utility" board..so i can use the switching board on any pedal, whithout changing the layout
.. Also, i don´t want to re-do the layouts of my previous pedals.. I´ll use the fet trick from now on, but for previous design..argh..too much work for a couple of more wires..

[R.G.]

If you are using a CMOS flipflop then you likely have logic signals of both polarities available. Why not use the Boss signal passing trick with JFETs, so you can put the JFETs out where the signals switch instead of routing them back into a single switching chip? Makes for simpler routing of signals.

That´s an option, but i´m thinking of having a kind of "utility" board..so i can use the switching board on any pedal, whithout changing the layout
.. Also, i don´t want to re-do the layouts of my previous pedals.. I´ll use the fet trick from now on, but for previous design..argh..too much work for a couple of more wires..

CD4053

[R O Tiree]

As R.G so succinctly said, there, "CD4053". It has three SPDT switches inside, just like you're used to. Hook up the control signal to all 3 and they'll all change at the same time, just like you're used to.

[dschwartz]

RG, why cd4053?, i know is a 3 spdt sitches in a chip, but what´s the difference?

easier layout?. i´ll give it a try!!!
anyway, my questions about popping still apply to 4053, aren´t they?