Mechanical Design Considerations for Pedals

[MrStab]

FreeCAD, GrabCAD and manufacturers' own 3D models.

[amptramp]

I do like the idea to open a medium-production pedal "and see how they do things". You can learn a lot by looking. And for this purpose an older or damaged pedal is as good as any.

Everyone should do this! But we should never assume that how they do things is good - because the vast majority of times that is mainly concerned with cost reduction - quality is secondary. Why they do things that way is equally - if not more - important. Don't take a monkey-see monkey-do approach to designing your pedals.

Absolutely true.  My professional background is military and spacecraft systems which have to survive all kinds of environmental factors, but a guitar pedal has to survive being handled by roadies who say, "Ook, ook", eat bananas and drag their knuckles along the ground.  The stuff I worked on was never intentionally abused.  When we shipped the Hermes spacecraft from Toronto to Cape Kennedy, we paid Bekins Van Lines extra to get an air-ride suspension truck and trailer (quite rare at the time) and the best driver they had.  Your pedals will not get the same treatment, nor should they need it.

If you are just building a pedal to practice on in your bedroom, that's one thing - almost any construction technique will be acceptable because it will continue to perform the intended function and you will be able to fix any problems on the spot.  If your equipment is going out into the big bad world, it pays to use the best construction techniques possible, like when Traynor tested their amps by throwing them out of a second-storey window in their place in the Yorkville district of Toronto.  Peter Traynor was a guitarist who played with Robbie Robertson back in the day, so he knew the kind of things that mattered.

I often watch videos of guitar amp restoration from the likes of D-Lab, Uncle Doug, The Guitarologist and Fazio Electric (who doesn't love Coleen Fazio?) and whenever they get in an amp that has everything mounted to a circuit board, they always go over all the solder joints and retouch them because cracked joints are normal and they are worst around pots and tube sockets.

If you want an interesting insight into military standards, check this out:

http://everyspec.com/MIL-HDBK/MIL-HDBK-0300-0499/MIL-HDBK-454_9165/

which offers you the original version, notice 1 and a couple of updates, all for free.  Millions of dollars worth of engineering advice and experience for free.  And most of this advice is applicable to what we are doing.

[GGBB]

The best thing you can do is buy or borrow a flagship pedal from JHS, Wampler and EQD and open it up and see how they do things. Those three are the closest intersection of DIY and small/medium manufacturing. You won't get many practical ideas from bigger companies like Boss/Ibanez since they are on a whole different level that doesn't really scale downward, but those three in particular are great to learn from.

JHS and Wampler are pretty far removed from DIY IMO. They used to be close to DIY, but that was many years ago - they have "evolved" into small-scale mass production companies, complete with the typical manufacturing compromises associated with mass production. Wampler doesn't even make their own pedals anymore I don't think - they're made by All-Pedal I believe - who make for all kinds of brands (maybe even JHS - who knows).

Pay close attention to the pots you are buying.  I built one unit where I included a separate piece of scrap plastic on the inside to lock into the anti-rotation lugs on each pot.  But with some of the pots, there was barely enough threaded length with the pot going through the plastic plus the outer aluminum structure.  ALWAYS make use of the anti-rotation lugs.  If you use wide knobs, they will hide the anti-rotation lug holes in the structure, so they should be your first choice.  NEVER mount pots on the circuit board or you will crack the solder joints.  DO NOT mount the board by the pots.  Use a lockwasher on the pot and reef it down tight.

This will solve some of the recurring problems we often see - pots that turn beyond the end stops and rip out their wiring and cracked intermittent solder joints.

With much respect to Ron, I strongly disagree with all of this - at least framing it in absolute never/always terms. I've been designing and building pedals for over a decade and I do none of these things and have never had a single failure due to them. To reference JHS/Wampler/EQD again, all of them use PCB-mounted pots and none of them use the anti-rotation lugs. They've been doing it the same way for many years. These types of practices can cause issues in sloppy builds, but with good engineering (drilling, PCB layout, component placement, matched mounting heights) they are not risk factors and so are not inherently bad practice.

They may not be "high" risk factors when done properly, but they are quality compromises in favour of manufacturing cost reductions nonetheless. It comes down to warranty claims vs production cost - it can be more profitable to build a less reliable product. So I largely agree with Ron here, but you aren't wrong either.

DIYers shouldn't make the same compromises just because brand "X" does it, but they can do whatever they want. If you want to build something as reliable as possible - follow Ron's advice. If you want to make your builds a little easier - use board mounted pots and switches. The problem is in the suggestion or assumption that because brand "X" does things a certain way it must be the right way. You can only understand what is right when you know why things are done in a particular way, and you know what your own build goals are and how they align with brand "X" build goals. Also, some of the "good engineering practices" utilized to reduce risks in a compromise solution are subtle details that are completely beyond the grasp of many DIYers, so it's better for them if they stick to simple reliable solid build techniques like Ron suggests.

I recently saw a comment in a thread to the effect that the user liked using stomp switch breakout boards because they weren't any good at soldering wires directly to the switch lugs. I thought - if they can't do that well what makes them think their PCB soldering is any better? Newbies assume everything works great - they need to learn why that's not true.

[ElectricDruid]

DIYers shouldn't make the same compromises just because brand "X" does it, but they can do whatever they want. If you want to build something as reliable as possible - follow Ron's advice. If you want to make your builds a little easier - use board mounted pots and switches. The problem is in the suggestion or assumption that because brand "X" does things a certain way it must be the right way. You can only understand what is right when you know why things are done in a particular way, and you know what your own build goals are and how they align with brand "X" build goals.

+1 this. Commercial pedals have all sorts of aims, and being the best-built isn't always top of the list. And there's no reason why we necessarily need to build something that will stand being thrown out of a second storey window when we're just mucking about trying out an idea. So you need to understand *what* you're looking at and  *why* it was done that way, and *what* of that you actually want or need to copy.

All that said, there's a lot of learning to be had by looking at everyone's else stuff!!

[aion]

They may not be "high" risk factors when done properly, but they are quality compromises in favour of manufacturing cost reductions nonetheless. It comes down to warranty claims vs production cost - it can be more profitable to build a less reliable product. So I largely agree with Ron here, but you aren't wrong either.

I would agree that it's a tradeoff, though I'd nuance it a bit further to include other reasons besides just cost.

I fully concur with Ron and others that the absolute most reliable way to build a pedal, at least using methods within the reach of DIYers and common parts, is to have the PCB affixed to the enclosure with its own screws, no epoxy or adhesive standoffs, no board-mounted components, with all pots, switches and jacks wired offboard with stranded wire. There is almost no conceivable way the pedal could break short of running over it with a car. It would definitely be okay falling out a two-story window so long as it didn't land knobs-down.

But... these methods also require an enclosure that is much, much larger than the circuit would otherwise need. You pretty much couldn't fit it in a 125B if it had more than three pots since the PCB and pots can't really overlap at all. So you're usually looking at 1590BB at a minimum in order to follow these practices without compromise, or you're using a highly customized enclosure like Boss/Ibanez.

There is definitely the business consideration of, "Do the people using my pedals need it to survive a two-story fall? Will they pay appropriately for it?" But to me personally, the even more important question is, "Should I make all my pedals significantly larger and/or give them awkward control layouts on the off-chance that one of them might theoretically fall out a window someday?" And where I land on that is that I'd happily trade a sliver of theoretical reliability (which might in fact resolve to zero actual failures, as in my own experience) in exchange for something that's sized efficiently, has a good control layout, and that I'm proud to look at. For me the potential cost savings wouldn't come into account at all in that decision.

[R.G.]

Missing from much of this discussion are the ergonomic considerations.

Good points as always Mark.

I keep my own scorecard of these kinds of things - from the distance between footswitches - both within the same pedal and between two butted-up adjacent pedals - to the minimum spacing of pots based on maximum knob diameter and maximum candlepower of LEDs.  These are all subjective, of course, but there are some guides on this. NASA in particular has human factor guidelines and so do most large corporations.

Sometimes this is not a help. Case in point: footswitch to footswitch spacing. The super-micro pedal craze made wearing pointy-toed boots necessary for arrays of these. Another is protection of the knobs and such from stomping and falling speaker cabs. The company I work for brought out a line of pedals with a cast-in up-swoop that prevented knob and pot damage from things like boots and speaker cabs; even driving a van front tire over one of them didn't kill the controls. People groused and complained about it even after it was pointed out that it was to protect the knobs.

Many design shops hold it to be true that the biggest problem with designing controls and interface setups is that ultimately all the customers are humans. 8-)

[BlueLdr]

Many design shops hold it to be true that the biggest problem with designing controls and interface setups is that ultimately all the customers are humans. 8-)

As a web developer, can confirm. You can design something to be elegant and effortless to use, but once it's in the hands of the end user, none of that matters, because they'll always find a way to use it wrong.

Relevant meme:

[Mark Hammer]

Many design shops hold it to be true that the biggest problem with designing controls and interface setups is that ultimately all the customers are humans. 8-)

Aw man, humans are the WORST.  Annoying that we have no other species we can turn to for our customer base.

In the meantime, we always have to plan around, or at least content with, the "aw jeez, do I HAFTA?" reflex that exists in abundance.

[drdn0]

How I approached everything was to make my PCB in Kicad, import all the components in as .step models, export everything to Fusion and align everything in 3D. Once aligned, it's very easy to make drill guides/etc as you have good data points to take the origin from (I use the center of the face as you're not concerned about edges then.

Being able to mount things in multiple planes at once is the biggest concern I found - sure, you might be able to get everything scooched in, but if you physically can't put it into the enclosure you're going to have a problem.

[Viktor Von Doom]

I recently saw a comment in a thread to the effect that the user liked using stomp switch breakout boards because they weren't any good at soldering wires directly to the switch lugs. I thought - if they can't do that well what makes them think their PCB soldering is any better? Newbies assume everything works great - they need to learn why that's not true.

Respectfully, if I am the user you are referring to - that is not at all what I said. I said I hate to do it, not that I wasn't any good at it. If that were the case, none of my instruments would function very well.

And 'new member of forum' does not mean 'newbie.' Less assumptions all around, please

[drdn0]

I recently saw a comment in a thread to the effect that the user liked using stomp switch breakout boards because they weren't any good at soldering wires directly to the switch lugs. I thought - if they can't do that well what makes them think their PCB soldering is any better? Newbies assume everything works great - they need to learn why that's not true.

Respectfully, if I am the user you are referring to - that is not at all what I said. I said I hate to do it, not that I wasn't any good at it. If that were the case, none of my instruments would function very well.

And 'new member of forum' does not mean 'newbie.' Less assumptions all around, please

And, FWIW, soldering wires to switches is absolutely garbage when you've got a million options for breakout boards/etc that look significantly nicer, are significantly neater, and allow you to use all sorts of fancy stuff like ribbon cable. Costs you only a dollar or so extra, and saves you significant amounts of time.



Anything we can to do to make our life easier, we should.

[GGBB]

I recently saw a comment in a thread to the effect that the user liked using stomp switch breakout boards because they weren't any good at soldering wires directly to the switch lugs. I thought - if they can't do that well what makes them think their PCB soldering is any better? Newbies assume everything works great - they need to learn why that's not true.

Respectfully, if I am the user you are referring to - that is not at all what I said. I said I hate to do it, not that I wasn't any good at it. If that were the case, none of my instruments would function very well.

And 'new member of forum' does not mean 'newbie.' Less assumptions all around, please

My apologies - I misremembered your post - and from there assumed you must be new to building. I think my point is still valid in general though, we all need to be comfortable with all aspects of building especially those affecting quality whether we like doing them or not, if we want to produce high quality results. But that might not be a goal - anyone is free to do whatever they like obviously. For many of us I'm sure we're not that caught up with making "perfect" pedals and probably not touring with them (or perhaps even gigging), so finding ways to make building more fun is completely valid. I don't like painting enclosures, so I don't - so unprofessional! But I gig with mine so they have to be reliable, which for me means hand-wiring the stomp switch. When I started building that was something I too didn't like doing, but I got comfortable with and better at it over time.

[GGBB]

And, FWIW, soldering wires to switches is absolutely garbage when you've got a million options for breakout boards/etc that look significantly nicer, are significantly neater, and allow you to use all sorts of fancy stuff like ribbon cable. Costs you only a dollar or so extra, and saves you significant amounts of time.

Nice, neat, fancy ... looking "better" doesn't mean built better. And to me that stuff just looks the same as mass produced garbage, a well done imitation of a factory build but not high quality. Just my opinion on the use of cost saving manufacturing techniques in DIY builds - not a personal judgment. Also less repairable - try replacing that stomp switch when it fails.

Those things don't actually save you time when you are good at the alternative approaches - they actually get in the way and slow you down and compromise your build (more solder joints, more heat on the switch). They may make things easier for some though.

[davent]

And, FWIW, soldering wires to switches is absolutely garbage when you've got a million options for breakout boards/etc that look significantly nicer, are significantly neater, and allow you to use all sorts of fancy stuff like ribbon cable. Costs you only a dollar or so extra, and saves you significant amounts of time.

Nice, neat, fancy ... looking "better" doesn't mean built better. And to me that stuff just looks the same as mass produced garbage, a well done imitation of a factory build but not high quality. Just my opinion on the use of cost saving manufacturing techniques in DIY builds - not a personal judgment. Also less repairable - try replacing that stomp switch when it fails.

Those things don't actually save you time when you are good at the alternative approaches - they actually get in the way and slow you down and compromise your build (more solder joints, more heat on the switch). They may make things easier for some though.

If someones using a breakout board with a switch you still have to strip wire to wire to the board and you now have to solder twice per solder lug. That's not saving time.

[stallik]

I'd like to add something to the 'ergonomics of footswitches' list.
How will the footswitch operation be affected when the pedal is placed in close proximity to another?

Sticking a switch in each corner gives great separation on that pedal but places it alarmingly close to the one in the next pedal. Especially if you're using top mounted jack sockets.

My pedalboard is a testament to this error

[Viktor Von Doom]

My apologies - I misremembered your post - and from there assumed you must be new to building.

No apology needed. Assumptions can be easy to make, but they are not necessarily true.

When I started building that was something I too didn't like doing, but I got comfortable with and better at it over time.

To be slightly pedantic here - again; I did not say I was uncomfortable nor skilled at doing it. I said I found it to be an unnecessary chore and time suck that I did not enjoy.

If someones using a breakout board with a switch you still have to strip wire to wire to the board and you now have to solder twice per solder lug. That's not saving time.

That's where ribbon cable comes in. Widely available for next to nothing in 2 to 9 wire varieties (or whatever your heart desires), pre cut to length and pre tinned. Any well designed breakout board can use them. And it saves someone who is into the small details (like myself) plenty of time cutting up little bits of heat shrink tubing to size and fitting and shrinking them to size after the switch is wired up. Strain relief is a very real thing that contributes greatly towards reliability, and the inherent mass of a ribbon cable provides plenty of it.

After all, this topic is about mechanical considerations vs. 'mojo,' no?

[GGBB]

To be slightly pedantic here - again; I did not say I was uncomfortable nor skilled at doing it. I said I found it to be an unnecessary chore and time suck that I did not enjoy.

You actually just said you hated doing it - nothing about chore or time suck:

Does anyone know of a proper 'breakout' style PCB for stomp switches?

I don't mean the standard true-bypass types that are quite common. I mean one that has through holes for mounting the switch, but then 9 solder pads for connecting directly to each corresponding switch lug. I have been looking around the 'webs, but can't seem to find anything like this commercially offered.

I truly hate wiring switches, and have a few Deadend boards coming my way that don't seem like they will play well with the TB style switch PCBs. I'd rather use a PCB and a ribbon cable.

If anyone knows where something like this can be found, please let me know. If not, I guess this might be the easy project I was looking for to self-learn PCB layout and design.

Thanks!

That's where ribbon cable comes in. Widely available for next to nothing in 2 to 9 wire varieties (or whatever your heart desires), pre cut to length and pre tinned. Any well designed breakout board can use them. And it saves someone who is into the small details (like myself) plenty of time cutting up little bits of heat shrink tubing to size and fitting and shrinking them to size after the switch is wired up. Strain relief is a very real thing that contributes greatly towards reliability, and the inherent mass of a ribbon cable provides plenty of it.

After all, this topic is about mechanical considerations vs. 'mojo,' no?

Meh. Ribbon cable doesn't change the number of wire joints necessary to hook up two PCBs. Stripping it takes longer than stripping several lead wires IME - and don't nick the solid core wire or the self supporting function becomes a weakness! Neat solder joints on a switch don't need shrink tubing (unless you need to make poor joints look "pretty"), but that only takes about two seconds anyway - less time than the extra time to strip ribbon cable. If done right it probably is a little more mechanically sound than lead wires, but I doubt it's any faster - probably slower in fact.

EDIT: Note also that ribbon cable can be a bit of a nightmare for repair. If one connection breaks - from handling or what have you - you have to redo the whole cable - all connectors. And they do tend to break quite easily from bending at the joint.

[GGBB]

If someones using a breakout board with a switch you still have to strip wire to wire to the board and you now have to solder twice per solder lug. That's not saving time.

Also, a stomp switch breakout board requires you to make one more solder joint than necessary - more heat on the switch - because in typical 3PDT pedal wiring one lug is unused. I suppose you could leave it unsoldered - but that would be being smart about it, and it certainly wouldn't look as pretty.

[amptramp]

That is one thing I have noticed about soldering in the usual 3PDT switch - you had better bring your "A" game for that job because the pins may start to move around as the plastic starts to melt if you are soldering them to separate wires.  I haven't used a switch PCB but I assume this stops a lot of that problem.

So far, no one has mentioned prototyping boards as a halfway measure between vero or point-to-point and a printed circuit board.  This is typical:



This allows you to add IC's and have separate power rails under the IC.  You can get them with 3, 4 or 5 connected pads on either side of the rails.  You may have to space the IC's apart and jumper IC pins to other sets of pads but everything can be made to fit.  You have to do your own layout but you can use one style of board for most of the stuff you build.

[Viktor Von Doom]

You actually just said you hated doing it - nothing about chore or time suck:

That is a fair point- I did not state a reason for hating it (but now you know). So in hindsight your assumption might have been perfectly reasonable. I retract all prior statements.

Stripping it takes longer than stripping several lead wires IME - and don't nick the solid core wire or the self supporting function becomes a weakness!

I only use pre-perforated ribbon; just pull off the ends and solder. But that is another fair point. Stripping them all from, say, a 20' spool would be a nightmare.

Neat solder joints on a switch don't need shrink tubing (unless you need to make poor joints look "pretty")

No, I am not trying to hide anything. I only do it for the strain relief- one bit right after the joint that butts up against the lug shoulder and then another slid over it that is slightly longer and covers the joint as well. Likely unnecessary, but I am prone to obsessing over small details. In a manner of speaking, it is kind of my day job, lol.

But anyway, I don't want to detract from the OP's post, and I don't want to come across as argumentative. I just truly enjoy discussing itsy bitsy details with others, and the back and forth that comes with it. Hopefully the discourse is helpful to anyone reading with similar questions