PedalPCB 5-Loop True Bypass Switcher

[Giorfida]

wonder if it could be sandwiched into steel building studs? Could be some cheap enclosures of just the right shape

 

[mdc]

Tails, please.

 

[juice886]

If you end up with a non-programmable switcher it would be really good to have 2 different switching modes:
"and" - like a normal bypass when you can have multiple loops engaged at a time
"or"- only one loop can be engaged at a time, so when you engage some loop all other loops will be disengaged.

Extending my previous comment. I think such kind of a non-programmable switcher would be great to have in a smaller form-factor. It can be really handy in cases when someone wants to combine 2 or 3 different effects in one enclosure.
So if I want to put 3 different effects in 1590DD (or 2 effects in 1590BB) then all the switching will be done with a separate PCB that has two switching algorithms and optional 'high studio quality audio grade' input and output buffers in addition. I'd buy for sure

 

[Alan W]

I'm not opposed to multiple projects, in fact I'm planning on it.

I do want to do some Arduino projects as well, so it's certainly not out of the question.

Working up a design now for a fully passive 3PDT based version.

Not that I’m opposed to building several varieties of switchers (or anything else, it would seem), but knowing what some of the other upcoming options might be would certainly be great. I know it’s still early in the process, and your starting this thread as an organizing one for the possible line of switchers is great.

 

[Alan W]

I’ve built a few “dumb” loop switchers, and used 1x 2 inch aluminum tubing as a housing. I cut one of the 2 inch sides so that only 1/4 inch is left on each side, and screw in a thin plate of aluminum on the bottom, with folds at each end to close it up. (There is a slight notch cut into the 1/4 inch returns on the bottom, so that the bottom plate can fold in flush.)

Besides very low cost (especially if you have a stock of aluminum, which I used to), its pretty unlimited as to how many loops you can fit in, with as much space between the buttons as you want.

 

[CR0SSBL4DE]

I want to see how they're handling the effects order switching...

I drew up a relay matrix a while back that could do it, but it involved a lot of relays.

I suppose you could combine solid state routing along with mechanical relays to swap order while maintaining true-bypass without 50 relays...

My calculations are that for 100% mechanical routing and switching one would need:

[X = No. of loops]
No. of SPDT relays = 2 * X * (X-1) + X

1 loop --> 2 relays
2 loop --> 6 relays
3 loop --> 15 relays
4 loop --> 28 relays
5 loop --> 45 relays

I doubt that's what they are doing on the fender switcher. I wonder if jumping all the loops and measuring with input/output with a DMM would give you continuity. That's what I would assume for "true by pass switching". (also it has a volume control integrated)

 

[Robert]

My calculations are that for 100% mechanical routing and switching one would need:

I'm wondering if they're using a combination of mechanical and solid state switching...

Relays to bypass, but solid state switching to route the order when a loop is active. That'd still technically be true bypass ...

I might have to crack one open just to see...

 

[giovanni]

I think there must be a way to reduce the number of relays needed. I did a smidge of research and what we are looking at here is a permutation graph. I think the key is that whenever you want to change the order of a loop, that implies that all other loops change their relative order. For example if you have pedals in order A,B,C,D and want to move A between C and D, the relative order of B and C also changes, going from 2nd and 3rd to 1st and 2nd respectively. I feel like this property could be used to come up with some clever switching architecture but I haven’t had a aha moment to figure that out.

 

[Big Monk]

I've only ever wanted to switch the order of my Wah to be honest, from first in line to AFTER the dirt.

 

[Robert]

I'm mostly just curious how it works. I'm not really planning to do order switching in this project right now.

Without a true display like the Fender unit has I can't imagine setting that sort of thing up would be much fun. I know the GigRig does it, but that's just too much stomping and clicking for my tastes.

"Hold this switch now tap that one three times, loop A and B have now swapped order unless you were holding down C at the same time. If so, now D is A and B is C and E... well I don't really know where E went ..."

 

[Big Monk]

I'm gonna see if I still have my graphic design files for when I made my switcher mockup. @Robert I don't know how you plan to lay things out mechanically but maybe seeing a mockup, front, and sides, etc. could be helpful when planning features.

 

[giovanni]

Now I’m down a rabbit hole of permutation network switching and whatnot. Somebody stop me!

 

[CR0SSBL4DE]

I'm wondering if they're using a combination of mechanical and solid state switching...

Relays to bypass, but solid state switching to route the order when a loop is active. That'd still technically be true bypass ...

I might have to crack one open just to see...

True by pass implies no buffers, but a direct cable from input to output, when the effect is off.
That said, it would not surprise me if marketing would coin the term even when it does not apply. (there has been Class AB amps marketed as Class A, because of the term "Class A" thought to mean higher quality than B or AB)

 

[Giorfida]

True by pass implies no buffers, but a direct cable from input to output, when the effect is off.
That said, it would not surprise me if marketing would coin the term even when it does not apply. (there has been Class AB amps marketed as Class A, because of the term "Class A" thought to mean higher quality than B or AB)

I’ve also seen MXR/DOD style SPDT “output bypass” switching (where the input is always connected) called “true bypass” since it used a click switch.

And don’t get me started on folks calling eyelet/turret/strip board “point to point” amps!

 

[Robert]

True by pass implies no buffers, but a direct cable from input to output, when the effect is off.

Right, what I'm suggesting is additional solid state switching when the effect is on.

For example, let's say we have five relays, the relays never change order, they're always hardwired in direct sequence.

When Relay 1 is switched off it is indeed true-bypass, a straight wire connection across it's contacts passing on to the next relay in the chain.
If they're all off it's a straight shot from the Input to the Output... True-bypass.

However, when Relay 1 is switched ON it doesn't connect directly to Loop 1, instead it goes to a matrix of switching ICs that can route that relay to Loop 1, Loop 2, etc.

Still technically true-bypass, but only requires 5 relays.

 

[CR0SSBL4DE]

Right, what I'm suggesting is additional solid state switching when the effect is on.

You mean like when all the effects are off? One effect off still means to route the signal from the effect before to the effect after. It still goes through the routing.

 

[Robert]

When any given relay is off it would bypass that channels switching matrix entirely.

I'll draw a diagram of what I'm describing when I switch off my soldering iron for the day.

 

[juice886]

I guess most likely in this case when the channel is on it needs a buffer to support such an electronic switching. So it is 'true bypass' only when the channel is off but involves additional components in the signal path when the channel is on. I guess that's what Robert means.

 

[Robert]

Exactly. I'm not saying that's how the Fender switcher works, just that it's a possibility.


This is extremely simplified, but this is what I'm describing:

One relay for each spot in the signal chain. When that block is active the relay is on and the digital controlled switching routes the signal to the appropriate effect.

When that block is bypassed the relay is off and that entire block of circuitry is true-bypassed.

If the desired order is [Loop 1] > [Loop 2] we (by way of the microcontroller) enable U1.1 and U2.1, and U3.2 and U4.2, then turn on both relays.

If the desired order is [Loop 2] > [Loop 1] we enable U1.2 and U2.2, and U3.1 and U4.1, then turn on both relays.

True-bypass doesn't necessarily mean "True-active".

This similar scheme could be used to add switchable buffers before and/or after each effect, a movable volume control, etc.






It's also entirely possible that Fender has a matrix of relays in there, I'm fairly sure that's how the GigRig G3 works.

 

[benny_profane]

Wild speculation is so unlike you, @Robert

True-bypass doesn't necessarily mean "True-active".

The hiccup there is that the copy makes a point to discuss the switchable input/output buffers. Any transistor switch will amount to a buffer of sorts.