I had a few minutes today to dig out this bike, and take some quick pics of possible motor and NuVinci hub placements. Keep in mind that most likely each of these will hang below the frame on dropouts/mounts made specifically for each one, rather than sitting along the tubing like the pics show. I may yet add the bottom framing to enclose them, too, that is shown in a pic/drawing on a previous page.
First, one of the Fusin motors from Dogman's melt-off, just forward of the NuVinci. Note the NV is larger than the Fusin!
Close up of same.
Probably I would run a small sprocket bolted thru the rightside cover of the Fusin, into a freewheel on the NuVinci's input splines. I might instead put the freewheel on the Fusin, and use a thinner single sprocket on the NV splines, to save room for a pedal freewheel. The motor chain is shorter than the pedal one, so if I can only fit one freewheel on, I'd rather it be the one on the longer chainline.
If I have to I can run a jackshaft to combine motor and pedals, but if avoidable, it'd be nice.
I am considering something like John in CR thought up, of running the Fusin core, minus planetary and shell, and putting a very small sprocket on the magnet bell's side, which would normally drive the sun gear of the Fusin's planetary, to a very large sprocket on the NV input. This would eliminate both the plastic gearing *and* the heat-retaining external shell, allowing the motor to be easily ventilated and cooled, and thus able to run harder and longer than it can be right now.
It's really only a 300-500W motor as-is, and pushing 500W into it constantly for long periods is not a happy thing, either, as I have shown myself on DayGlo Avenger with the trailer trips--it just gets so hot that the halls start intermittently not working, but fortunately they came back after cooldown. Probably won't always do that.
Alternately I could use the GM/9C hybrid, which can be a 1000W motor, minimum, as-is:
Again, bolt small sprocket to rightside motor cover, large one on NV drive input.
On both of these, the NV's brake rotor mounts would probably be used to drive the output sprocket, but I might use the spoke flange instead, depending on the size of drive ratios needed at the rear, and amount of chain engagement I want.
One more thing--since I don't want to deal with chain growth/etc. on the rear swingarm/frame, there will have to be a new custom swingarm built that will accomodate a jackshaft as part of it's mounting pivot point, so the NV output can go to the left side of that jackshaft, thru the js/pivot, then out the right side of that to drive the normal freewheeling input of the wheel (because I don't want to back drive all that chain up to the NuVinci's output side, including the NV housing).
That custom swingarm pivot is going to take some thinking.
Most likely I will need"
--core shaft (small electric scooter rear axle, from ScootNGo or the like) that will support the weight and stress of the pivot,
--bearings on that *just* inboard of the dropouts the shaft will mount into.
--Possibly a spacer tube between the bearings to keep them at the proper distance, riding just on the outer diameter of the axle.
--Large spacer tube to ride on the outer surface of those bearings. Tube will need to be etched on inside diameter a bit to make recesses for the bearings so it can't slide on them side to side. Might eliminate need for small spacer tube on axle. This tube will be the torque transfer from left to right at the pivot; keyed sprockets with spacer rings on either side will fit over it.
--bearings on outer diameter of above tube, just inboard of the sprockets, will support the pivoting of the swingarm.
--Tube on those bearings will be part of the swingarm.
Something like this:
Anyone see why that might not work? Or have somehting more practical and easier to make?
I also have the rear swingarm and belt drive/etc from a Honda Spree gas scooter from the mid-80s, but it's for a much smaller wheel than I want to use. I am sure I could lengthen it; I'd have to take it off teh scooter to see exactly how it's setup.