The Triad drive and its variants are driven by miniature motors from Namiki. These take +/-3V sine waves with three phases and a common return (probably Y-topology, but not ruling out delta-plus-ground).
The controller boxes we have are ones that we can't get more of, and the cabling is unreliable, so I started designing a replacement.
This was never finished. A prototype of the front panel was designed but the shield wasn't.
The tenatative plan was to use a 5V supply and use buffer amplifiers to produce 0V, 2.5V, and 5V. The 2.5V line would be the common return, and the motors would be driven with square waves at +/-2.5V.
There's no guarantee that this would have worked; the motors sometimes got stuck even at 3V. But there's high confidence that it would be safe to use with the motors (de-rated voltage).
It's questionable if the AD86xx amplifiers would have been able to source and sink enough current.
The original plan was to have separate analog and digital headers, to avoid injecting digital noise into the motor cabling (which is right next to ephys cabling).
The revised plan (at least for the prototype) was to put the analog circuitry on the mainboard (Arduino shield), and have analog and control signals both on the microcontroller header.
The microcontroller header therefore has two different use-cases for the same pinout. In digital mode, per the schematic, the "motor drive phase" signals are driven at TTL voltages and the "Com/NC" signal is disconnected (or better yet, grounded). In analog mode, the analog buffers are before that header, and the "motor drive phase" signals are fed directly to the motors, with "Com/NC" being the common return (shared between all motors, which may cause problems with ground loops).
(This is the end of the file.)