Using synth components/designs that normally run on +/-15V power in stomp boxes

[dxinteractive]

I'd like to make a state variable filter (which must have a voltage controlled frequency!) in a stompbox and found what looks like a winner here:
http://www.birthofasynth.com/Thomas_Henry/Pages/VCF-1.html
Schematic: http://www.birthofasynth.com/Thomas_Henry/pdf/VCF-1/Sheet_0002.pdf
Schematic for expo converter: http://www.birthofasynth.com/Thomas_Henry/pdf/VCF-1/Sheet_0001.pdf
I intend on using an LM13700 instead of the 2x CA3080 as the author suggests.

Anyway building that shouldn't be an issue, but powering it as a stomp box raises some questions. I could always just try to find a +/-15V wall wart, but would prefer to make it use a standard +9V / +12V / +18V power source just so I can power it from my power supply like all my other pedals (a cheap but useful Joyo Power Supply 2). If that's an option of course.

Seems like my options are to use charge pumps, and potentially introduce noise / current sourcing issues... or modify the schematic to work across narrower supply rails, but it's really quite beyond me what I'd need to change in the schematic.

Any other ideas? Anyone tackled this kind of thing before? Thanks.

[MaxPower]

The datasheet I looked at didn't list a minimum power requirement for the LM13700. It only listed +/-18V (or 36V single polarity). You might look at various datasheets and see if one of them lists the minimum power requirements.

You could always try operating it with a lower voltage and see if it works. Try a couple of 9 Volt batteries set up to provide +/- 9V.

Edit:
D'oh! voltage controlled filter. I have no idea if changing the voltages will affect the performance of the circuit. I was just thinking about powering the ICs.

[StephenGiles]

Possibly the main difference is that synth modules are designed to run at line level and so need the higher voltages to allow for the extra headroom, whereas stompboxes are at instrument level so only need a small voltage.

[nocentelli]

I'd use a voltage inverter chip like the 7660 and try it on +9v / -9v. Headroom is likely to be the the drawback, but guitar level signals shouldn't cause a problem as mentioned. If you search for Eric Archer's Vcf, he's done a similar circuit, but only using 9v/4.5v/ground as a quasi bipolar supply and a 13700 chip. It's only Lpf, but you could tap the BP/HP points, and it has the same expression pedal voltage control circuit. Works great, self oscillates at max resonance, doesn't overload with humbuckers.

[UKToecutter]

You could use an AC supply and regulate on board.
This has been used on a number of boards I've made and was used extensively by Shumman (i.e. PLL, Lion etc.)



IC1 is an LM317 and IC2 is an LM337
D1 - D4 are 1N4001
Ignore the +-12.47.  That was the voltages for the PLL

[Strategy]

Rarely looked at on this forum is that a long-time synth DIY maker, CGS (Cat Girl Synth) actually makes a PCB all in one project that allows you to use synth processing modules as a pedal. It provides I/O with appropriate level matching, and regulated power.

http://www.cgs.synth.net/
Search his list there for CGS46, Stomp Box Adaptor.

One multimode filter that I have done as a pedal straight up, powerable from +9V, is the Wasp VCF. There are many versions of it. Even the one I did needs a little input gain to properly interact with guitar or line input. Build thread here (with shameful lack of audio samples from me!) Bonus: the version I did has some interesting switchable overdrive! It's CMOS and CA3080 based.
http://www.diystompboxes.com/smfforum/index.php?topic=102714.0

[PRR]

> main difference is that synth modules are designed to run at line level and so need the higher voltages

A secondary difference is the demand for "voltage control". For single-ended controls like Vol and Freq, "zero" is a perfectly valid input. However single power supply circuits usually do not handle zero gracefully. They need to come above their negative supply a bit before they really get a grasp of the input.

For example: the '3080/'13700 types will run on 3V total (with accordingly small output). However their Iabc input is "dead" until it comes above 0.6V/1.2V, because it has to turn-on a transistor (or two).

Sometimes this is acceptable. In fact it may be wise to have a Volume that shuts-up if the input is only "near" zero, so small leakages in the control path don't cause sound-leak when you want *no* sound. However a 1.2V threshold is a large part of the 0V-5V control-range suggested by 9V supply. If you do a simple Linear pot, it is mute the first 25% of rotation. An Audio pot is dead over maybe 2/3 of rotation, very awkward. You can pre-bias for this, but that spoils some of the universal-control concept (each amp/filter may need different pre-bias and thus can't share a control signal).

At some point it really does get easier to put your WHOLE voltage-control system on bi-polar power, so that "zero" is not an extreme but a middle. I can't really see a bunch of separate pumps-- much work, poor tracking. Put all your VC stuff together in one box on one +/- supply. I like simple supplies (such as AndyM posted) but I suppose you could have a On-Spot or similar powering one -9V pump or a +9V:+/-12V converter.

[dxinteractive]

Thanks so much everyone. Some of the 9VDC ideas look interesting, I think I'll give the +/-9V charge pump idea a go first and see if it works, and work out what to do from there.

PRR you've got some really good points I didn't consider - looking at Eric Archer's VCF it runs into that exact problem of having a 0V CV at an extreme, and therefore looks like it requires a rail to rail op amp to cope with that. The only other things I've built that make use of CVs in the past all had some kind of microprocessor like an Arduino to listen to the input, which made responding to 0V a cinch. I don't know if I'll be making any other pedals that actually require a bipolar supply, but if so then perhaps it would be better to group them together to share a supply.

http://www.cgs.synth.net/ is great, haven't seen this before. I totally forgot to consider the fact that guitar signals are so much quieter than what a synth would normally work with.

So another question - in getting this charge pump going to provide a negative rail, is noise a potential issue, or is there anything else I should watch out for when setting that up? Haven't needed to use a charge pump before. I had a quick search and I can get some TC1044s which I've heard are like the MAX1044s (that have the boost pin to raise the clock whine above audio).

[alanp]

There's a semi-related thread on e-m -- http://electro-music.com/forum/post-405563.html#405563 -- check out the post by JovianPyx I linked to. The thread is about a 9V standard for synth-diy newbies, and he touches on some stuff about it.