Vroom Vroom...putt putt

It's getting pretty hot again now, so I decided that even though I don't have an ideal solution, I do have a potentially workable one, with small motors and friction drive. I don't like friction drive for several reasons, but it's easy to do, and it does work, and I expect it to be pretty dry for my rides during the weeks (?) I will need it, which was one major problem trying to do it earlier in the year--too much on-and-off rain that made it not work reliably. One oily puddle and it just slips for several hundred feet or more, instead of providing any real motive power--and oily puddles abound when the rains hit, since they're infrequent and too little water to actually wash the streets clean of it--especially the edges and righthand lanes, where bikes generally must stay.

After having looked at a number of various ways of doing friction drives, especially with smaller motors, I found this idea:
http://www.psnw.com/%7Ejmrudholm/elecbike.html
It appears to have bigger ex-radiator-fan motors than mine by a bit, and he's definitely using bigger batteries, but the principle works, and is easy to build. I even have an improvement on it. His drivewheels are better than mine, too, being made from old skateboard wheels, while mine are just the ex-fan-rotors (for a couple of reasons). My smaller batteries means I don't need to use my trailer but I still could use that with the car batteries I've got around here if I needed the range (assuming the car batteries are up to the task--they might not be, as they're a few years old).


These two pics show my implementation of his idea. I didn't have any angle-bracket I could use (as he did), and decided I didn't actually need anything more than the flat plate, since I have the baskets already there for some support. I used an old rackmount faceplate, since it's aluminum and over 1/8" thick, quite stiff and sturdy, but easy to work with the tools I've got.

Cutting the motor holes was a bit of a challenge, as I had no hole-saw large enough, nor any jigsaw blades that would work with this aluminum. I used the doorknob hole-saw I did have to make the core hole, then used a 3" cutting disc tool to expand that to a hole big enough to pass the motor thru, with some play around it for positioning and fine-tuning the tension of each motor if necessary (turned out it was, due to the play the baskets have around the wheel/frame from their shape and mounting design).

I simply used machine screws to hold the controller and MOSFETs (already on their own little heatsink) to the ex-rackmount plate. The motors are held on with the same screws they had holding them into the plastic fan cages from the car.

A couple of angle shots of the rollers, which are simply the plastic fans with the blades cut off, then sanded to a rough finish for traction against the tire tread.

The darker yellow one is from the old Tercel motor I tried earlier this year, because after a very very bad pothole strike (over 6" deep!) due to a car that would simply not let me move farther into the lane even with my turn signal going, one of the newer ones came off it's aluminum mounting point:

Unfortunately there's no way for me to push the little metal hub back into the plastic--it has offset grooves in there that show me it was molded together as one piece, but can't be forced back together now that it's come apart. I consider myself lucky it didn't damage anything on the bike itself (especially considering I just rebuilt the wheels to try out various truing methods not two weeks ago, and they ride better than even when I first got the bike--a fair trade for about 5 or 6 hours of work on them).

This does tell me that these things aren't going to last, so I will need to find an alternative method of transferring the motor power to the wheel--I can still use friction, but I'm going to have to find something I can mount to those metal hubs (or directly to the motor shaft, which is very short and relatively thin). Probably skateboard or rollerskate wheels, as they're also smaller diameter which would improve the motor performance--they'd be able to spin faster while still getting the same wheel speed, so the motor wouldnt' be under as much of a strain being held back from it's "natural" speed at the 24v I'm running them at (assuming full controller output). I did some calculations a few days ago for what motor speed I should have for various size friction drive rollers vs the desired wheel speed / vehicle speed, but now I can't find the clipboard I had the notes in, or I'd put that info here. :-(

This is what the whole bike looks like with the motor system on it:

A bit uglier than it used to be, but it's sort of hard to tell that these days. :-)

That improvement to the idea I borrowed is simply a way to disengage the rollers from the wheel for various reasons, when not using the motors.
I added a notch in the plate for each bracket wire of the basket it had to cross, so it could seat down in between them, and have room for an adjustment for tension against the wheel. The front edge of the plate is notched to fit under the basketwires at the front, so it will act as a pivot, and the rear edge can be pulled down by this:

I took a rear-shifter off of the very first donated bike I got for this project (the green Huffy), along with it's cable, and mounted it as close to the inboard right side of the handlebars as I could get and still easily reach it with the righthand. The cable passes to the back of the bike, then (since I didn't have an actual mount to use for the end of the cable sheath) thru the slots on a hose clamp that's around the back end of the basket. That gives the cable something to pull against. When the shifter is up, as in the middle pic, it pulls the back end of the plate down which puts the motors' rollers in contact with the tire tread. When it's down, like in the last pic, it releases the plate which then doesn't pull the motors against the tread, and will (when I put a return spring under the plate) eventually actually pull them away from the tread altogether.

That lets me use the motors when I need to, but disengage them completely so they don't interfere with pedalling when the batteries die or something else goes wrong (as it inevitably will). I can also do it on the fly, while riding, at speed, so I don't have to stop to disengage (which might not be possible in traffic, sometimes) because of a problem. I can also use this feature for testing motor/controller/etc issues, without having to either prop the bike up to keep the wheel off the ground, or taking the assembly off the bike. Kind of like a "neutral" gear. :-)


There is also a throttle, which at the moment is just a slider potentiometer, because the Hall sensor broke during installation, and I didnt' have time to dig out my other salvaged ones to replace it, but I did already have the slider pot right there in a junk bin from something else I'd been fixing the day before. I do want to put the Hall throttle back in, so I can use it more like a typical handlebar-throttle again--that's a lot easier to use than this one. If I could, I'd like to get a throttle off a motorcycle, or something else, that's made of metal, instead of the plastic ones I've run across so far. Similar to what was done in the design I based this motorizing scheme off of, where he used a dirtbike throttle, and used it's cable to actuate the thumb-style throttle he has for his controller. I'd probably put a magnet on the throttle and a hall sensor on the body, though, as is often done with the cheap plastic throttles on ebikes and scooters, mostly because the controller I'm using was designed with that in mind (the resistive throttle I'm using now had to be put in a voltage divider network to get any actual throttling range out of it, and it still doesn't control over the full sweep of the slider--min and max are less than 1" apart despite the nearly 2" of travel).

I put the scooter's "ignition" switch and power meter in a much better place this time:

I still can't see the LEDs in daylight at all, but in deep enough shade or indoors they're easy enough to see. Knowing how much power I have left is not critical anyway, but nice to know. This time I put the switch/meter down at the frame's headstock tube--this means the cables from it don't have to flex, and it's one less thing cluttering the handlebars (they're pretty full, something I don't like). I don't need access to either of these things during a ride, only at the beginning and end of it, when mounting or dismounting the bike, really, so they're fine out of the way here. I can still reach and see them if I have to, during a ride.


I considered cutting the plastic shell that the scooter used to house them in and mounting that there, too, to keep the dangling wires better protected, but realistically it's just more stuff on the bike, and isn't tough enough to protect it from anything serious anyway--not even weather. I might still do it eventually, if it turns out that I keep this setup for any length of time. Depends on my mood and my "copious spare time". :-)