ilovefuzz.com

Mike wrote:Compared to current switching methods, is this cheaper? Simpler? Smaller? More reliable? Easier to service? Unlikely.

Relay switching has two things over traditional switching that I can think of:

1. Nearly silent. Very important when you're doing ambient sets at cafes or small venues. The K-CHUNK of a traditional switch is sometimes cringeworthy in these cases.

2. Allows electronic control of bypass. Especially useful if you have a microcontroller. Easier to set up trails on a delay.

I know I'll be moving my pedals over to a momentary stomp and relays eventually. A good momentary stomp costs abotu half as much as a good 3PDT, as well, though using a relay uses up that difference..

cloudscapes wrote:A good momentary stomp costs abotu half as much as a good 3PDT, as well, though using a relay uses up that difference..

That's just it though. If you read the documentation for the AMZ relay, a lot of people were reporting problems when using cheap momentary switches. AMZ is recommending a $15 switch with that circuit. Plus the cost for the chip and the board.

If we are talking real-life as opposed to theoretical life, the lifespan is only as long as the weakest component. A lousy switch is a lousy switch. And with relays you also have to look at the quality of the chip and the other additional components.

I understand the mechanical advantage of non-physically latching switches. But at the same time, I also know I've bought numerous cheap consumer electronic devices where non-latching buttons stopped working for some reason or another. So I don't know. I'm not saying relays are bad, just that I don't feel like there's enough real-world testing to say for sure they are the way to go.

If I were a gigging musician, I would probably go with relays and shell out for the best switch possible. I think then you are getting the best of both the theoretical and real-world. I mean, i feel like one reason 3pdt switch are bad is because no one really uses them anymore outside of a few specialized electronics-hobby type. So you could probably make a decent 3pdt switch that would cost $10 and last nearly a lifetime but there's no incentive for companies to do so. And there probably never will be as it's just old technology. So while relay switching is the way in the future, as a bedroom player/hobbyist not sure if it is worth it at this time.

I'm not even sure that old-school analog-ish stuff is even going to be around much longer. Everyone will just plug their guitars into computers (if they play guitars at all) and just use VST's.

eatyourguitar wrote: we do not need to multiply (footswitch failure)*(relay failure) to get the total failure probability.

woah easy. that would not be the way to do it anyways. ie if p(fail) of the switch was .5 and p(fail) of the relay .5, you say the p(fail) of the system is only .25? i don t think so. (if they are independent it would be p(fail) = .75 in this case for example).

the cases for failure are: just the switch fails. just the relay fails. both fail. with the numbers above each of those occur with .25 probability. so the probability of failure is .75 not .25 (under the assumption of independence).

its not just both pfails multiplied together, although if you think of probability of success, in this case it would be the 2 probabilities of success multiplied together, then you can do 1 - p(success) = p(fail) and get the same result. Because for a successful switch the relay has to succeed AND the switch has to succeed, if independent -> implies multiplication.

do some Bernoulli trials and find out what the expected trial number of the first failure is....

the relay is not a human interface device. it is also protected from accidental damage since it is inside the case. we can not compare them like apples to apples. so with all that in mind, the 400,000 cycles for the relay is most likely the only spec that matters. since we can not in our life time ever use a pedal that much, you can pretty much assume that it will not fail at all. so now if we can find a footswitch that is more reliable than the blue 3PDT,

yes i think you see my point then. if the footswitch you have sucks, why does it matter what the spec of the relay is? you would still be repairing that 36 year old pedal. you sort of say well 400,000 is alifetime, if only we had a foot switch that would last that long. total circular reasoning. i mean sure the setup might last longer but you are still at the mercy of a mechanical component, which probably wont go 400,000. that is all i was saying.

crazynoises wrote:That's just it though. If you read the documentation for the AMZ relay, a lot of people were reporting problems when using cheap momentary switches. AMZ is recommending a $15 switch with that circuit. Plus the cost for the chip and the board.

This fault is not due to switch failure, or cost of the switch.
This fault is the result of a poorly "designed" debounce/filter routine of the trigger (switch) input, which is evident by the hardware realization of the circuit.
A fancy pants switch will only mask this design flaw, which will return as the switch ages. Rats.
Need more detail on faults in the "design?" I probably know of more than the "designer."

Those who know the material get A's on their homework. Those who don't copy the answers from a peer, crossing their fingers and sweating when their paper is returned in the hopes that their peer found the correct solution.

10:1 says the same switch will perform flawlessly with the Click-Less™ circuit. I got $5 on it. You?

I think implying that the total system has the same failure rate as the relay is a bit misleading. You could use a great relay and have an awful latch circuit or actuator. You'd have to be pretty terrible to design an unreliable relay circuit. Cars and many other things work perfect everyday with relay switching. Why not guitar effects too? But even if your design is fantastically amazing. The total system failure will be greater than whatever individual part is the most prone to fail. That relay might be great but do you trust the capacitor next to it? That being said, bypass systems should evolve. We can't design the pedals of tomorrow with the bypass systems of yesterday.

Full Disclosure:
I am currently using the 3DPT-PCB(Edit: 3PDT-PCB)switch from CHK Electronics in all my designs. Which is how I new right where to find the datasheet. I love datasheets, sometimes I read them just for fun. Everyone wanted to copy my math, science and electronics school work. Engish... not so much.

Narwhal-Industries wrote:Full Disclosure:
I am currently using the 3DPT

three dogs purple trousers? tripple dro perfect thro? I think you mean 3PDT!

eatyourguitar wrote:

Narwhal-Industries wrote:Full Disclosure:
I am currently using the 3DPT

three dogs purple trousers? tripple dro perfect thro? I think you mean 3PDT!

No way, man, I've heard in central PA they use triple double pole throwers.

Just ask Tom, I think he uses 'em in all the FuzzHuggerFx.

No, what I meant was, Lebanon Bologna 4-Life.

Jack Deville wrote:This fault is the result of a poorly "designed" debounce/filter routine of the trigger (switch) input, which is evident by the hardware realization of the circuit.

im not sure if im following you totally. are you saying from the circuit, you can tell that the code inside does not use a great debouncing scheme?


i mean i would say from the description on the site that the people 'designing' it are not coders, just from the way they talk about the device but id be curious as to any technique that can deduce the debouncing scheme from looking at the hardware layout. (other than perhaps its all done in code, or it has some analog support).

multi_s wrote:

Jack Deville wrote:This fault is the result of a poorly "designed" debounce/filter routine of the trigger (switch) input, which is evident by the hardware realization of the circuit.

im not sure if im following you totally. are you saying from the circuit, you can tell that the code inside does not use a great debouncing scheme?


i mean i would say from the description on the site that the people 'designing' it are not coders, just from the way they talk about the device but id be curious as to any technique that can deduce the debouncing scheme from looking at the hardware layout. (other than perhaps its all done in code, or it has some analog support).

I don't really understand either, there's a cap over the switch to form a low pass with the pullup in the micro. But that's it. As long as that low pass is in some kinda sensible range there's nothing wrong with the hardware debouncing.

Yes. I can tell from the hardware realization of the circuit (the parts selected and their placements) that the debounce is shitty. The tell: a redundant hardware low-pass filter. Of course, its only redundant if the routine was well designed to begin with.
And why is there an external pull-up? Oops. I've said too much. Now there will be a revision based on my observations. AGAIN.

They say we are truly our own worst enemy. I should just shut the fuck up.

id like to see a relay switch back and forth between 2 pcb's with a pot to control the speed and a momentary to juice the whole damn contraption
kind of like a double distortion build

Jack Deville wrote:Yes. I can tell from the hardware realization of the circuit (the parts selected and their placements) that the debounce is shitty. The tell: a redundant hardware low-pass filter. Of course, its only redundant if the routine was well designed to begin with.
And why is there an external pull-up? Oops. I've said too much. Now there will be a revision based on my observations. AGAIN.

They say we are truly our own worst enemy. I should just shut the fuck up.

ah ok. i didn't look too hard at the pcb myself and when i made my previous comment, I was under the assumption that there would be no analog de-bouncing. i guess ya if they have some, maybe thats another tip off they have not so much idea on the coding side.

pedalmania wrote:id like to see a relay switch back and forth between 2 pcb's with a pot to control the speed and a momentary to juice the whole damn contraption
kind of like a double distortion build

you could do that all with a relay and a 555? (: