Debugging amusement

[paul.creedy]

So, I finished a Parallel Universe II build the other day, which works but only partly.

I read that it can be sensitive to which LM386 is in it, so I ordered a few more to test.

While I'm waiting I started another build, then discovered I'm lacking some caps that i thought I had, so I've just ordered some, plus a whole load more other bits and pieces for the next few builds.

Having done that I looked round to see what to do next (I do have a little work to do today, but not much) and gazing round the room I spotted the open case of the PU2 and thought to myself "I suppose I might as well double check the board, even though we all know I built it very carefully so there's not going to be anything wrong there".

The first thing was to check the outboard wiring and . . . there are five wires which are next to each other that are in the right order, but all one strip out of place.

I'm going to drink a cup of coffee and think about this and then go and do something about it - I'm beginning to think that it might be this that's part of the problem, and might not need to swap LM386s after all

[MrStab]

nice! that sounds like a real pain in the arse to undo. on the upside, at least it's not some obscure problem that'll take ages to figure out!

the other day i managed to reverse the input and output jack wires for the first time. if i'd plugged the pedal in backwards, it would've worked.

[PRR]

> "...we all know I built it very carefully so there's not going to be anything wrong there".

We should not tolerate such language. Especially from ourselves.

"The first principle is that you must not fool yourself—and you are the easiest person to fool." — Richard Feynman

[paul.creedy]

It's better, but there's still work to do.

The overall volume is better now, and the gain and tone are working fine, but if I switch the oscillation on it needs a big strum to get any sound at all (and even then there's not much) and the oscillation control appears to be doing nothing.

I'll check it over again when the LM386s arrive (or before if I have time). 

It's all part of the fun though, and when I get something to work finally it's almost better than having it work perfectly in the first place!

My Happy Little Accident circuit being discovered through failing to get the circuit I was trying to create working was a huge bonus, but that didn't stop me from having another attempt at the original, and last night I filled one of the original two boards (the other is in the completed HLA) with the appropriate parts - this time it worked

Because I used transistor sockets I did a little swapping and then put it away until I have a spare enclosure to put it in, and now that I've learned a little from getting things working I can relax and move on

[paul.creedy]

> "...we all know I built it very carefully so there's not going to be anything wrong there".

We should not tolerate such language. Especially from ourselves.

"The first principle is that you must not fool yourself—and you are the easiest person to fool." — Richard Feynman

I'm always happy to expose myself as a flawed human being (I left an exam in my youth proclaiming that anyone who didn't get top marks (an A) was an idiot. I got a B) for other people's amusement.

[PRR]

I don't care if you expose yourself. And it is a good lesson.

I do object to language like "I am sure it is wired correctly" because my own experience, and observations in chatrooms like this, is that 97% of the time it is NOT wired correctly.

Full Exposure: early in my career I faced a failed power supply in a broadcast console. Pulled it out, easy bench-fix. Put it in, didn't work. Too-long story short: for 2 weeks I could not figure it out. A friend came by and ignorantly questioned which wire was which. I had ass-umed I knew, but when I Opened My Mind to the unlikely possibility that I Could Be Wrong, it was clear: I had been connecting it wrong all along.

Naturally I told my co-workers "It was a tough repair but I got it!" rather than admit I was stupid for weeks. But I drag the story out occasionally when it may help others.

[paul.creedy]

I'm beginning to think we've misunderstood each other - my original post was meant to show that despite my best efforts, I clearly didn't wire it correctly, as it didn't take long to spot a pretty large error

The "I suppose I might as well double check the board, even though we all know I built it very carefully so there's not going to be anything wrong there" bit was meant to be shot down by the next sentence where I showed that making such assumptions is often a foolish thing to do, and as with your example, I'm always happy to be the one providing the lesson via my mistakes

[MrStab]

"The first principle is that you must not fool yourself—and you are the easiest person to fool." — Richard Feynman

as a recovered addict, this statement is very true. i also find myself applying it to electronics - i think about how many stupid errors i've made before, and go into debug with full expectation that I'M most-likely to be the bug. it's a very reassuring outlook when you've soldered a load of IC's without sockets!

[duck_arse]

My Happy Little Accident circuit being discovered through failing to get the circuit I was trying to create working was a huge bonus, but that didn't stop me from having another attempt at the original, and last night I filled one of the original two boards (the other is in the completed HLA) with the appropriate parts - this time it worked

might I enquire - how does the sound differ from your happy little accident?

[paul.creedy]

This is the original demo - it's got some stutter in there, but it's more of a glitchy, starved-signal sound than mine has with the MPSA13 in.

[pinkjimiphoton]

haha i spent 2 whole days debugging a phaser, only to realize as i was about to start desoldering chips that i had put in DUAL opamps instead of SINGLES.

i feel your pain!!

everyday

[paul.creedy]

More debugging fun.

With the wiring in the right place and my "best" LM386 in, I tried a couple of pedals in the loop (an EHX Pitch Fork and my RNG) and that worked ok, but still one or two bits weren't working as I'd hoped.

I'm still waiting on some more LM386s, but I do have a couple that I bought cheaply with a little amp kit from China, so I thought I'd try them while I wait for the others.

The first one sounded much like the one it replaced, if a little noisier.

The second one turned the pedal into an extremely startled rodeo pony that had just eaten a number of super-hot chili peppers.

The starve switch works, the oscillation is mental and I can barely turn the vol pot up before a huge sound bursts forth.

I think the suggestion that the pedal may be a little sensitive to which LM386 is in it could be considered an understatement.

[paul.creedy]

Here's another shockingly bad photo of the finished pedal - if there's some sunshine tomorrow I'll have to try and take some better ones



So that's five finished, one complete but unboxed and one waiting for components.

Nearly time to consider what to build next, not that this pedal building lark is remotely addictive . . .

[paul.creedy]

My other thought, if I don't build anything new next, is to remove the transistors from a Dwarfcraft built Devi Ever BDSM that I own and replace them with sockets so I can play around with other transistors and see what happens

I've tried selling it once or twice with no interest (I'm not a huge fan of it, but it's not a terrible pedal) so tinkering with it might make it something I'd happily keep.

I've also got an Ampeg Scrambler I built from a kit which never really did what I expected it to, so that might be fun to tinker with too.