The Mixer is a 5-input mixer with a transistor-based make-up gain stage and topology based on the model CP3 mixers from the modular electronic music systems built in Trumansburg NY by the RA Moog company. Its features are low noise, minimal slew, and massive asymmetric distortion on demand.
Each input has an input amount level, and there are three outputs. Two outputs are unattenuated positive and negative outputs, and the third is an attenuated output controlled by the sixth knob on the panel. Its normal operation is bipolar from negative to positive, but when the "negative kill" switch is engaged the attenuator works from null to full positive. In addition, channels four and five are normalised to positive five volts and negative six for CV offset or distortion "pulse width" effects.
As calibrated to factory standard, 2 o'clock on the dial is approximately unity gain, at 3 o'clock it begins to clip and then distortion presents itself above that input level (from 10V p-p input signals.) This is adjustable to the taste of the customer and if your application needs a predictable unity gain position for clean mixing applications it can be provided for you on request. (obviously if you want to live dangerously it can be calibrated in a more extreme fashion.)
The positive and negative outputs are not only necessary for control signals and other phase-sensitive applications. Because of the circuit topology of the gain amp the positive and negative outputs actually have different distortion characteristics. The negative output is more of a straight clipping, but the positive output has some interesting waveshaping to it which also means that when the output attenuator is in bipolar operation it is actually possible to isolate only the difference component between the two different types of distortion. This is a feature which was not available on the CP3 (or any other mixer that I'm aware of based on it, which until now have only been privately built for the most demanding of enthusiasts.)
You may ask "why on earth would you want such an old circuit? why go so far as to model the power rails of that old stuff for what anyone else would just throw a cheap dual opamp chip in there? are you nuts? the cheap dual opamp would give you another 10 volts of headroom and would bring your circuit cost to a tenth of what it is now!!"
The answer is of course, the sound. The sound of those old modular systems (the ones on albums like Tarkus, Rubycon, Mirage, Halo, Switched-On Rock ... those albums that make us want these sounds) wasn't just a matter of a certain lowpass filter which everyone and their brother has made a copy of. The sound started with simple animal-like unijunction-core featureless oscillators, was mixed in distortion-heavy mixers, filtered with the aforementioned filter (and other filters people never bother to copy), and then level-managed by a transistor-based VCA. one of those four stages is the mixer, and in my opinion it's more than a quarter of the sound.
Here is a meandering sound example with 2 Synthesizers.com Q106 VCOs into the Mixer, then the MOTM-490 LPF, and an STG Soundlabs VCA: