Building a semi-modular and need help with simple VCO/LFO circuits... & is my case trying to kill me? (beginner stuff)

[josh_truant]

So I decided to get started building a synth, but wasn’t interested in modular necessarily; I own no other instruments, and don’t plan on buying any. So the first promising enclosure that popped up was this vintage aluminum case for storing old slides, 13x7x3.

My plan was to slowly build and drill holes as schematics caught my eye, just using small stripboard that I was treating as though were eurorack modules (run off of a eurorack PSU, individual power cables [not any synth standard, just 3-pin keyed 2.54mm], and all standardized sizes so I could stack, swap, etc).

Case issues:

I was planning on using the AI Synthesis design and used the BOM from there specifically, but realized I didn’t feel comfortable putting the secondary of the wall wart through the metal case with the jack internal, and built the MFOS wall wart schematic with a plastic insulated jack external instead. Was I right to be nervous about that, or am I being paranoid? (The outside of the secondary was my virtual ground in both cases, but my assumption is that a virtual ground fails to work as such if the + or - is lost.)

My lingering concern is my virtual ground is still through my case due to contact with pots/jacks after power up, and last night I noticed touching a jack socket affected the sound…

Conveniently, the same thrift store that had this thing, also had a brand new 4u 84hp rack yesterday, so is an all-metal case with a virtual ground, wall wart-based PSU not possible to design safely and I should scrap it now? This is more like what I want to build than a CGS/Serge clone, but thrift fate is weird like that.

Ok that was serious safety stuff, but on top of that, I’ve built a dual VCO based on Moritz Klein’s (exactly his, minus thermistors and plus one circuit using an open input on the 40106), and noticed the frequency pots affect BOTH oscillators, to a lesser extent on the non-connected pot, but substantial.

I decided to build a simple dual version of the LFO by Nicolas on electro-music, having built one already on non-stripboard before I settled on format & hoping this would be an easy success after the VCOs… only to find that it, too, has surprising results.

With a cheap scope, the saw is OK sometimes, but the skew pot response is extreme and collapses the wave with a lot of rotation left; the pulse has a harsh saw spike at its start, and has similar response to skew changes, except reaches a point after the pulse collapses that it produces negative voltage spikes. Which is extra confusing because it’s 2 copies of the same previously tested circuit, just cleaner now that I’ve gotten better at soldering & on one strip board (completely separated vertically, no missing cuts between them or anything) and they both behave the exact same (wrong) way.

Which brings me back to my DIY enclosure/initial plan for the whole thing, because the common things to all 3 circuits are the case, the power supply, and the power filtering circuit I copied and pre-loaded onto DIYLC for my stripboard layouts, which is 1N4007 diodes for polarity protection, 2 10ohm resistors for short protection, 47uf electrolytic caps and 104s for decoupling, with 104s to ground at all ICs power, as close as possible.

North Coast’s article on design mistakes mentions that the resistors as fake fuses are obviously unnecessary for diy if you check for shorts before powering, and might actually cause issues with power regulators, like in my DIY PSU, so my thoughts so far are take the resistors out, but I’m still too worried about the enclosure to want to power it up and test.

I know that’s a lot, but I generally get a little too ambitious and rotate around projects if I’m stumped and the problem seems liveable; the case/ground anxiety has that rotation stopped, so if you only had time to answer one (and a half) question after reading that novel, it’s: is my enclosure dangerous? If so, how is all that hardware mounted to an aluminum modular panel any different?

If I know the enclosure is safe, I’m going to hook up a meter and scope and watch for fluctuations at the regulators on the PSU, and remove the 10ohm resistors if so; barring missed cuts/bad joints/misplaced components, which I periodically re-check and haven’t yet noticed any issues with, are there other things I should be checking? My intuition still says it’s a greater issue than the circuit itself.

Thank you for reading, and for any insight y’all might have to share! I’ll be reading up on esoteric 4U stuff in the down time…

[analogoutput]

You seem to be using the term “virtual ground” to mean something other than what I understand it to mean, and I’m not at all sure I understand what you mean. Maybe if you draw us a picture? A schematic or at least some sort of diagram to make it clearer what you’re doing and where your concerns are.

But something like the MFOS wall wart supply is quite non-dangerous, if by “dangerous” you mean to your physical safety. (Your circuits’ physical safety is less easily assured.) The dangerous voltages are all confined to the wall wart. What’s on the wall wart’s 2-conductor output cable is 12 VAC, with neither conductor connected to earth. In the MFOS supply one of these conductors connects to that circuit’s ground, and hence to the ground of each module. If you have jacks mounted to an uninsulated metal case (or faceplate), and the jack sleeves are connected to the modules’ ground, then the case will also be connected to ground. If the case/faceplate is insulated (e.g. if the metal’s painted) it could be floating and you should connect it to ground. (Scrape a patch of paint off and solder a wire from there to ground on the MFOS circuit board. Oh, did you say “aluminum”? Forget I said “solder”, use a nut and bolt or something to make physical contact to bare metal.)

The cases of the pots (if they’re metal) should contact ground too. If the case is uninsulated metal and the pots are mounted to it, they will be grounded. If the case is insulated or non metallic, you may need to solder a ground wire to the case of each pot.

To reiterate, all this is not to do with your physical safety — that’s not in danger. It’s to do with making sure the electronics work.

It probably will not be necessary to connect the synth’s ground to wall voltage’s earth although it should be okay if it is (do it safely if you do it, and do it only if you’re sure you know what you’re doing. That has to do with your physical safety.)

You can use the continuity setting of your multimeter to verify contact between the case, ground for each module, ground for the MFOS supply, jack sleeves, and pot cases.

@josh_truant

are there other things I should be checking?

General advice for troubleshooting a module and asking for help

(This is a wiki post, please add and correct!) Look and see if others have had the same problem. Use the Discourse search facility to find topics where the module you’re building is discussed. Inspect your work. Verify all components are correct values. Look for solder connections that are rough, irregular, skimpy, dull, or missed completely (we all do that!) and reflow them. Look for solder bridges between pads and remove them. Check your power connection. If it’s a Eurorack style power cabl…

[josh_truant]

I’m talking about the reference 0v created by the dual regulator design, but probably not understanding enough mixed with reading far too many new terms to be using “virtual” properly; as far as troubleshooting, I’ve seen many times that the wall wart should be safe, I guess I needed to hear it personally! (I’m not even particularly safe or worried about the circuits, just my neighbors and the old wood building, so anything that won’t incapacitate me enough to prevent me using the fire extinguisher is where I’d like to stay.)

[analogoutput]

Yeah, that’s just “ground”.

Having the jack external to the case (or mounted to the case for you to plug the wall wart in externally) is a good thing, but having it internal isn’t particularly dangerous, other than if it’s going in through a hole with exposed sharp edges that could nick the insulation. But if even if you short the 12 VAC wires the worst that should happen is the wall wart trips off. If it’s designed right. Or the wall outlet’s circuit breaker trips. Best avoided but it’d be hard to get that to rise to the level of personal danger.

[josh_truant]

Thank you for the response, especially so fast; my real job will not be as thankful since now I’ll be up all night working through troubleshooting with the power on, which will hopefully lead to tools coming out, etc.

[luminiferous]

@josh_truant

Conveniently, the same thrift store that had this thing, also had a brand new 4u 84hp rack yesterday, so is an all-metal case with a virtual ground, wall wart-based PSU not possible to design safely and I should scrap it now? This is more like what I want to build than a CGS/Serge clone, but thrift fate is weird like that.

I would highly recommend getting the 84hp rack, if it’s affordable, for the following reasons:

1. Easier to build/debug. If you build a module at a time, and then install it in the rack, when something goes wrong, you will know which module is to blame, and you can easily remove and diagnose/repair that individual module, whereas (in my experience at least) the slide box will gradually become a chaotic rat’s nest that is very difficult to work on.

2. Easily expandable. You may find, as you go along, that you run out of space. If you start with the rack, you solve this problem by adding another rack, whereas with the slide box, you need to build a new power supply, make sure that there is no voltage drop between your enclosures, and so on.

[josh_truant]

Of course it was a simple pot wiring error with the LFO (a signal wire was untrimmed at a pot, contacting case, which explains why touching ground was affecting sound and making me nervous); added another cap to power to fix a spike in the square wave and it’s running as expected.

Still not sure why my dual 40106 has so much interference between their tune knobs, but decided to give up on tiny cases for the sake of troubleshooting ease per advice, especially cause I realized I had severed a very hard to trace wire during some other repair.

Moved everything to an IKEA cube for now while I figure out what to do with the 4U rack; it should be deep enough to hack into a matching pair, so I’ll probably do 168HP of 1U & 3U, but there’s so many options when you do it yourself.

Thanks again!