This is an envelope follower (EF) module in Kosmo format with mic and instrument preamplifiers.
I looked, in simulations and on the breadboard, at several envelope follower designs (you can read about some of that by following the envelope follower tag) and ended up deciding to go with a design based on the ARP 2600 — there’s a schematic in the service manual. It’s a rectifier-and-filter design, as were some of the ones I decided against. I suspect part of the reason it performed better was its 4 pole, 24 dB/octave filter. The original design used LM301A op amps, but substituting a TL074 seems to work just fine. I did reduce the gain of the output stage since it seemed needed on the breadboard, partly I suppose due to the ±12 V instead of ±15 V supply. Aside from that, I followed the ARP design, including a head scratching input attenuator that’s wired sort of backwards from what I’d expect.
That’s the envelope follower proper, but the ARP EF also has a preamplifier section for input of external, non-synth level signals which, of course, are what you’re going to use an EF on a lot of the time. It’s designed around a Teledyne 1339 op amp, which good luck finding. Rather than trying to work out a suitable substitute I decided to follow the lead of Eddie Bergman, who built a stripboard envelope follower based on the 2600 design but replaced the ARP preamp with two different preamps: One to boost instrument levels to synth levels, and one to boost mic levels to instrument levels — which then goes through the other preamp as well.
I liked the idea but not so much the implementation. I don’t know anything, really, about mic preamps, but some googling turned up some disparaging comments about a mic preamp similar to the one Bergman used. Those commenters seemed more favorably inclined toward a different mic preamp design by Andy Collinson. Again not knowing anything about mic preamps, I could only go by their opinions and by my perception that Collinson’s writeup seemed to suggest they knew their stuff. So I breadboarded it… and it didn’t work, or at least not well. More googling and I found this discussion of Collinson’s design. The OP was having different problems with it than I was, but I nevertheless tried adding the 1k resistor mentioned there and that seemed to fix it. So I went ahead and used that design. It uses BC549 transistors, which I didn’t have in my stash and had to order. It’ll work with 1N3904, if you take the reversed pinout into account, but the BC549 has a noise figure about three times lower.
As for the instrument preamp, Bergman’s is built around an LM386, and while that may work well enough, what I’ve been told is the LM386 is meant to drive low impedance speakers or headphones, not high impedance synth modules, and that it should not be used in preamp designs. But I knew of another option. Ken Stone’s Stomp Box Adapter is meant to boost instrument levels to synth levels (and attenuate them the other way), and Stone used nearly the identical circuit for external instrument inputs in his Active Real Ring Modulator. I felt more confident about that, so that’s what I incorporated into my design, with the fixed feedback resistor replaced by a pot for gain control.
Rather than toggle switches to choose mic, instrument, or no preamp, I decided to use the tip normal switches on the jacks. If nothing’s plugged into the synth input it uses the output of the instrument preamp, and if nothing’s plugged into the input of the instrument preamp it uses the output of the mic preamp. And if nothing’s plugged into the input of the mic preamp it uses the condenser mic capsule mounted on the front panel! Yes, this module is always listening, so be careful what you say. Besides going to the EF input, the instrument preamp output goes to a front panel jack, so you can use this module to feed external signals (or the front panel mic) to other modules too.
Bergman made another couple of modifications on the other end of the EF: A comparator to generate a gate and trigger when the envelope is on, and a switched capacitor acting as a simple lag processor. Both are useful, I think, but too useful to just hard wire them to the EF. Instead I’ll be building a separate comparator module and a 2600-inspired voltage processor including a lag processor, to which the EF output can be patched if desired.
I drew up the PCB based on the results of breadboard tests, and then (as usual) after building the PCB discovered things I’d missed. Such as, the output amplitude depends not only on the input signal amplitude but its waveform. A square wave signal gives a larger output amplitude than a ramp wave, for instance. I ended up changing the values of two resistors to increase the gains of both the preamp and the EF output stage.
I tested it by using the ADSR envelope generator on my B2600 to modulate an oscillator signal, then fed that into the EF. Here you see the original B2600 envelope (yellow), the input signal (magenta), and the EF output (cyan). Not perfect, but no EF is.
For input frequencies below about 150 Hz, you start to get audio frequency ripple on the output.
And if the original envelope has very fast attack and decay, the EF is noticeably slower.
But those are typical EF issues.
This module isn’t really a plug and play kind of thing. With an external signal source you will want to use the gain knob to get the preamp output up to around 10 Vpp, and then use the signal level knob to get the output to the right level. It’s probably a good idea to use a scope to check your settings.
Schematics, KiCad design files, Gerbers, and documentation in the GitHub repo: https://github.com/holmesrichards/arpenvfol.