The image above shows the mechanism to dynamically position the tracker bar. This idea is nothing new, conventional pianolas use a vacuum system to achieve the same effect. Part of the array of 88 (plus) phototransistors can be seen between the HDPE rubbing strips.
Completed and functioning unit. The strip LED light for illumination of the sensors is above the paper. The leads are 12 volt in and audio out. (18/04/18)
The electronics end (yes, it does have an experimental prototype look about it). The diminutive rewind motor (on the right) is viable because during the rewind process the takeout motor runs in reverse, feeding the paper to the rewind spool. The rewind motor current (and hence the torque) is limited by the resistors in the right foreground. This arrangement avoids the need for a clutch.
The hardware end. The 1:1 chain drive makes little sense (though it is laterally compact). The idea was to have some flexibility in selecting drive ratios, however, although the sprockets, hubs and chain are good quality the supply is problematical and the flexibility was not realized. The knob, despite its knurling, has a pull out, not a rotate function, it is a part of a few bits that locate the non-drive end of the spool.
"The printed circuit board we didn't have to have." (Oz political joke, reference.) This connector intensive board principally turns devices on and off according to the input conditions. It would have been more economic, connector and MCU wise, to incorporate its functions into the CPU board. The slide potentiometer is the tempo control.
Use: The download material on the Optop subdomain is available without restriction. (This is not the case for other cartography.id.au domains or subdomains.) However it would be welcomed if (in the case of Optop material) the name Optop, and the URL, optop.cartography.id.au, were preserved.
Optop contact: email@example.comJase.