[sdiy] Plug in-breadboard power supplies, non-CV simple audio signal gens, and more!

[lanterma at ece.gatech.edu]

I get to teach my "Electronics for Music Synthesis" class again this semester. Yay! And yes, I do plan on taping the lectures again... ;)

Anyway, having gone through four semesters of helping people with their final projects, I've been thinking of ways of trying to streamline the process of Synth DIY experimentation (both for my students, and for myself as well.) I've been thinking about the various fiddly bits that take up time in lab, like dialing in the desired supply voltages, hooking up power, etc. I've also been trying to think of ways that students can do more building & testing & debugging on their own, in their dorm room or something, without necessarily being in lab - i.e. if they just have a cheap multimeter, but don't have immediate access to a bench power supply, signal generator, scope, etc.

[First off: yes, I know breadboards suck, for myriad reasons. Moving along...]

So I'm thinking about: 

1) Power. I'm intrigued by the little breadboard plug-in power supplies,.

---> One example is this one the MakerShed, that gives +3.3V or +5V depending on a switch setting from a somewhat arbitrary typical DC wallwart. I notice the switch changes the resistances around a LM317; that's how it can get the two voltages. It uses a slide switch which makes sense; the voltage is something you wouldn't want to be able to accidentally change from bumping the switch, so it's something you'd want to make the user have to put a bit of effort into doing.

http://www.makershed.com/ProductDetails.asp?ProductCode=MKSF5

http://makezine.com/images/store/Breadboard-PowerSupply-v2.pdf

Here's one from Sparkfun, using the same trick:

http://www.sparkfun.com/products/9319

http://www.sparkfun.com/datasheets/Prototyping/General/Breadboard%20Power%20Supply%20-%20SMD%20v11.pdf

Here's another variation on the same them, this time by Curious Inventor:

http://store.curiousinventor.com/media/upload/bread_board_power_supply_instructions_vc.pdf

And here's another unipolar one:

http://www.daqstuff.com/ebaydata/400039/400039-rev3-PDF.pdf

http://store.curiousinventor.com/media/upload/bread_board_power_supply_instructions_vc.pdf

And here's a skinny one: http://gadgetgangster.com/find-a-project/56?projectnum=358

And finally, here's an adjustable supply from adafruit industries: http://www.ladyada.net/make/bbpsup

---> OK, that's enough examples... moving on... It would be great to have one that plugged into a breadboard and put ground on blue rail, +15/12/9 on the right red rail, and -15/-12/-9 on the left red rail, and have a separate set of breadboardy-type connects on the PCB with +5 V or maybe +3.3 V to power a microcontroller or whatever else digital logic.

A three-way switch could be used to select 15, 12, or 9 volt bipolar. I require my students to conform to the MOTM/Blacet power format, which is 15 volts, but sometimes they choose to adapt a 12 V synth module design or 9 V guitar pedal design. I tell them to get things working with the original rail voltages first, and then we can fiddle around with adapting it for 15 volts. A second switch could turn everything off and back on. It could also have an MOTM style connector (and maybe connectors for a few other manufacturer types) for providing power to a *single* module on the bench, so don't need to run another power cord from your big fancy MOTM Power One that's powering the rest of your synth. 

Does such a beast exist? I'm thinking probably not, but it sure sounds useful. 

If I were to design such a think I'd probably want to start with a +/- 15 V DC (or maybe +/- 18 V DC? I have seen 18 V used in old Oberheim designs, I recall) wall wort (or preferable wort-in-the-middle to avoid taking up space on a power strip), rather than having the students fiddle with transformers and mains voltages. Anyone have a favorite wall wort-type supply?

The wattage wouldn't need to be very high - just enough current handling to juice up typical "single" modules. No one will be trying to power a whole wall of modules with this.

Anyway, I doubt such a three-voltage beast as I describe above already exists, so I'd need to design one. Would it be better to try to use a 7815, 7812, and 7809 "in parallel" (same for the 79xx side), and use a dual-pole switch to switch between their outputs? Or perhaps use the 7812 to get 12 v from the 7815, and then use the 7809 to get 9v from the 12 volt (i.e. cascading them), and use a dual-pole switch to pick the voltages? Or would it be better to use a LM317 and use the switch to switch in different resistors to choose +9, +12, or +15 (that's the trick used in most of the circuits listed above), and similarly use an LM337 for the negative side. I could use a dual-pole switch to pick the different resistors (since the positive and negative sides would need to switch together.) The more characters I type in this paragraph the more it sounds like the LM317/LM337 approach would be simpler and less costly. (Even if I am using a bipolar 15V DC wall wort, I was thinking regulating it - with the 7815, or appropriate resistors for the LM317 - would be a good idea in case someone was using an extra crappy wall wort).

For the +5 and +3.3 volts - again, is it better to say get the +5 volts from the +9, +12, or +15 supply, or get it directly from the original DC coming in? 

2) Test signals. It would be great to have some sort of uber-rough, minimal parts count sawtooth, square, and maybe also triangle wave generator (or generators), that output say -5 to 5 volts (MOTM oscillator standard), not CV but maybe with a knob that changed the frequency in a super-rough untuned exponential sort of way. Any favorite non-CV oscillator circuits? There are probably hundreds of such designs out there...

Maybe also a lamp-based oscillator for sine waves (I recall Grant R. was a fan of these).

3) CV sources. For easily emulating typical CV sources, I'm planning on just using post and an op-amp buffers - one could run from 0 to 10 V (emulating MIDI-to-CV pitch info), and another from 0 to 5 V (emulating EGs), or something like that. The student can just turn the knob up and down and be a human LFO or whatever. ;)

4) Output. Students often look at the scope traces long before they actually plug in a speaker. Generally, there's a few PC-type speaker sets in lab, and students swap these around, but they tend not to stay penned down for very long... also it would be good for students working in their dorm or whatever to be able to hear their circuit working (or not as the case may be) without needing a scope. Maybe something like this?

http://www.sparkfun.com/products/9151

I've often seen students try to stick their module output straight into a tiny speaker like that, which messed with the circuit in all kinds of ways. I then explain that they need some sort of appropriate way to drive the speaker, which is *not* part of a standard synth module, and which most computer speaker sets have built-in... so I'm thinking of sticking in an LM384, although maybe that's overkill.

Anyway, any thoughts are very welcome - and if I come up with any cool PCBs or something I will make them available.

http://www.national.com/mpf/LM/LM384.html#Overview

- Aaron