Bing has the parts as pointed to by Mike above at
microsmt.com.cn Also on his aliexpress store.
Re Bing's comment on spacing, I settled on 20mm spacing to get some space between them, to make it easier to handle component changes, or feeder changes. But, I have a feeling I am not going to be changing the feeders themselves, except for maintenance. It is so easy to load with new components, with the feeder staying in place.
However, my feeders are super easy to fit and remove. To fit, I position it above and a bit behind, then I drop it onto the rail, and move it a bit forward, the rear locks in place. A single 3mm screw secures it (screw inserted in feeder in top pic, lhs.). Even with feeders on either side, it is easy enough. In the pic, the feeder on top had a lock that I glued in place when setting everything up, but final version has an extended tape chute with locking key built in (bottom feeder in the pic).
The 150rpm N20 motors are perfect with the standard nylon worm gear. I did not like all the old designs of the tape peeler, where the worm gear is unsupported on top. Most of the load on the worm is vertical, driving the gear, but there is some lateral loading, just over half way up, so I designed my own tape puller with a drop in support axle for the top. No bearings, its just acting as a bush, but I got some 3mm bronze welding rod so friction and wear is hopefully ok. Obviously bearings are better, if I could have got the small bearings Mike used, then I would have used the design that he has in the grabcad files linked above.
The other 2 gears are PETG also running on 3mm bronze axles. To make sure everything was OK, I used an old school analog ammeter to check the current used by the motor, under different conditions:
1. free running, no top axle or gears
2. free running, with top axle but no gears
3. free running, with top axle and gears
4. actually pulling tape, with different tensions and thicknesses
I also fed +- 1000 0603 resistors through a feeder continuously using my little machine I alluded to above, all good, I monitored servo and N20 temperatures very scientifically with my fingertips, they got a bit warm, but not majorly so. This test drove the feeder much harder than OpenPnp ever would, with no delays between feeding.