My First Homemade Synth Module!

Last modified date

Comments: 0

My first attempt at a homemade Eurorack-compatible module.

This is part 1 of 2, next is : Amplifier/Attenuator Synth Module

KiCAD schematic & front panel templates are on GitHub.

So I recently got a Eurorack frame and one commercially-made module (see My First Modular Synth!). Coincident with this a Behringer Wasp synth arrived. It very soon became apparent there where compatibility issues with my existing music room gear, most significantly, signal levels :

  • instrument : mic, guitar etc : 100mV
  • line : mixer, computer soundcard etc : 1V
  • synth : 10V

Very ballpark, some mics produce about 10mV, but I’ll usually have a pre-amp (or some kind of booster pedal on guitar/bass). That 1V (give or take) is really the studio lingua franca. For synths, there doesn’t seem to be any strict convention, could be 0-5v, +/- 10v, but around there somewhere. 10V is hellish in comparison to the rest (but very nice for signal/noise ratios).

In addition to this, my low-budget setup mostly uses 1/4″ jack connectors. Euroracks (typically) use 3.5mm jacks.

So what I need is a –

Signal Amplifier/Attenuator

I didn’t bother looking what modules were commercially available, I already have a well-defined problem. But I did do due diligence with half an hour’s googling for homebrew circuits. Something very close to what I want is Ken Stone’s Stomp Box Adapter. The name suggests this is designed mostly for integrating external guitar effects pedals into a modular synth setup. He has a fixed, DC-coupled resistive รท20 for attenuation, a fixed x20 signal amp for boost. While I do appreciate the minimalism of this, I think I would benefit from a bit more flexibility.

So, this first pass at least, I decided on a passive resistive attenuator with a potentiometer wired as a divider (a 1k resistor on the wiper output to avoid any shorting problems). A x10 amplifier with a pot on the output, switchable going to another x10 stage. Offering variable fine, x10 / x100 coarse.

Circuit Design

Ken Stone uses a TL071 op amp. These are just the go-to for (relatively) low noise, inexpensive op amp audio designs. FET input, so effectively infinite input impedance, gain-bandwidth is about 5Mhz so plenty for audio, I’ve found them a brilliant little workhorse. Got a bunch of TL072s in my drawers, so here we go :

It’s a cross between the datasheet circuits and Ken Stone’s version, tweaked a bit.

Each stage is a x10 non-inverting amplifier. The 1M resistor on the input avoids the capacitor that follows picking up a bit of charge and popping when you plug in. The 10uF is primarily DC-blocking, but with the 150k to ground will act as a filter to any way-low freq stuff. It does pull the input impedance way down from what the op amp offers, but tying that down is usually a good idea. There might be issues with very high impedance sources (like piezo pickups), but usually they need calming anyhow.

The gain of an ideal op amp in this non-inverting configuration is 1 + R4/R3 = 11.

I’d have probably first reached for a 100k/10k selection, but only out of habit. [ I took the 47pF from Ken Stone’s design. I’ve not done the sums/simulation, but that in combination with the resistors should cut the bandwidth well down to say 200kHz. (Probably 1/(2*pi*R*C) is a significant factor, but I really can’t be bothered, it looks about right). High-end limit is likely to be important, even though I’m only aiming for max gain of the whole lot to be 100. Experience has taught me any non-inverting amp likes HF oscillation given half a chance.

Ken Stone’s amp has a big (10uF?) capacitor in series with the ground resistor. Again, I’ve not done the analysis/sim, but I want to play it safe, making a DC grounding.

Second stage is just another of the same.

I’ve put a 10k level control between the two stages. Yeah, it means any circuit noise gets boosted no matter what, but there shouldn’t be much there, undoubtedly a tiny fraction of that generated by everything else in the chain.

So each block should work, more or less, as a filter with bandwidth from well below to well above audio range, gain of 10-ish, fairly flat in the middle.

Circuit Board

I am a huge fan of stripboard for one-off projects like this. The cost of getting PCBs fabricated is neglible now, but drawing them out takes a lot of time, then the delay of getting them back would mean me losing any momentum.

I really don’t know how people can use the boards with a single copper ring per spot, make their own lines between. Strips! 90% of the connections done already. (I’ve never tried the spotty boards, also never tried wire wrap, that also looked unnecessarily difficult for no gain, lots of expense).

You don’t have to plan layout with something of this level of complexity. Just copy schematic to soldering. Hah!

I did forget to tick off components as I soldered them, very good practice, hey ho.

I wanted it to be as small as possible without making it difficult. Thinking narrow Eurorack spaces.

Make it dual.

I could have left myself more space on the LHS (longer board, cut down after soldering). But still the only jumpers I needed were for the chip power lines, yellow to simplify connectors. Similarly little for chopping tracks under.

They do say to put decoupling caps close to the devices, so I’m not going to feel to bad how I ended up mounting them :

It is, as yet, untested.

The Eurorack thing.

So the one module I’ve got :

16 HP (80mm-ish). Does so much! (Far less than many Eurorack modules, that’s a rant for another day).

I mentioned the connectors?

Two attentuators – small jack in, pot, big jack out

Two amplifiers – big jack in, pot, switch, small jack out

Height is nominally 128mm (3U), width 80mm (16HP).

I need to think a bit more about the layout before cutting/drilling. I think for this one I’ll use LibreOffice Draw (like Powerpoint) because it takes no thought. It is now possible to use KiCAD to make a PCB design, have screenprint on 2mm aluminium, drilling done, a week turnaround, reasonably pro-quality front panel, maybe 10 for $100

But one-off.

I haven’t tested the circuit yet, but easiest way seems like building the whole thing, fixing from there. If need be, replace the whole circuit – the low effort bit.

Part Two : Amplifier/Attenuator Synth Module


Leave a Reply

Your email address will not be published.

Post comment