Last year, around June maybe, I bought a spring reverb tank — which I then did not do anything with for months. Well, that was pretty much the plan. I bought it when I did because it was inexpensive and I didn’t expect to be able to get a similar deal later.
At the time we were awaiting a spring reverb driver module from LMNC. It was originally supposed to be released in February 2022 but what with one thing or another it wasn’t. I was starting to wonder if it would ever be released, and I knew even when and if it was it might not be what I was looking for, so I got to work on a Kosmo format adaptation of the Music Thing Spring Reverb (Mk II). The LMNC module did come out in September. I took a look at it. It’s called “SPRING SPRING CRUNCHY SPRING SPRING” or “CRUNCHY SPRING SPRING CRUNCHY CRUNCHY” depending on which page you’re looking at, which gives you a hint some distortion might be involved. Indeed, it seems from the schematic the first thing it does with the input signal is put it through a comparator to turn it into a square wave. I was more interested in a more clean reverb, though, so I went ahead and finished up the Music Thing adaptation. I got the boards fabbed a month or so ago and finally built it this week.
(Once again the 3D printed front panel is a stand-in until I get an FR4 one fabbed.)
The circuit is just about identical to the original. One thing I noticed is R5, a 100k resistor to ground on the second input (which is normalled to the wet output, for feedback). It’s parallel to several other ~50k resistances to ground! That means the input impedance is well below the usual 100k, and adding R5 just makes it lower still. I don’t know why you would want to do that. I left it in the schematic and put a footprint on the PCB, but with a note saying “OMIT?”, and I did leave it out in my build.
Other minor tweaks were that I replaced the single rail-to-rail bypass caps with dual rail-to-ground ones, I added power reversal protection diodes, and I decided the 10 µF AC coupling capacitors C1 and C2 could perfectly well be 1 µF, and so could be film instead of electrolytics.
The only other change I made, and it’s not even really a change, is that the expander module — which consists only of a pot and two on-panel phono jacks — is incorporated into the base module. The jacks are so you can switch out reverb tanks without having to delve behind the front panel. (I haven’t bothered mounting them on this temporary panel.) The pot is to mix between the spring reverb and the (optional) Belton Accutronics BTDR-2H reverb module, a.k.a. the Brick, a PT2399-based electronic reverb circuit.
I’m using this with an Accutronics 4EB2C1B tank. The behavior’s VERY sensitive to the trim pot, there’s a small window between basically dead and out of control where it’s all good.
The wet/dry mix is voltage controlled, via a vactrol. I used an Xvive VTL5C3. Yeah, I know, real DIYers make their own vactrols, but I haven’t been very happy with my attempts, and this commercial one’s been sitting in my stash for a long time waiting for me to use it. Or to find it. When I say “in my stash” I mean I’d put it in the box marked “voltage regulators”. As one does.
That’s another difference between this driver and the LMNC one; the latter has voltage controlled everything. There are two LM13700 handling four different control voltages. But Tom Wiltshire deliberately didn’t do that: “The input signal has a huge impact on the output. Sometimes, feedback is immediate and uncontrollable, other times it never comes. That’s one reason why feedback isn’t CV controlled; it normally seems to need a human ear and hand as part of the circuit (though it’s easy to patch CV-controlled feedback by putting a VCA in the feedback loop.)” It’d be interesting to compare both approaches. Maybe someday I’ll build the LMNC module and give it a spin. For now, though, it’s this one.
Schematics, layout, Gerbers, and documentation may be found at:
Analog Output 