It’s tough to compare bikes qualitatively so I mounted some guages on the bike to grab some actual hard data. I took a piece of plastic and dremeled two holes in it so I could attach it to the front boom with a hose clamp. I then held a voltmeter agains the plastic and drilled two holes through the plastic and the bottom of the meter and screwed them together. I wired the meter to the generator output. I also added an ammeter to measure the current flowing from the generator. I left room for another meter to measure battery pack voltage and then finally mounted the bicycle computer.

I added two Sterilite trash cans under the driver’s seat for cargo carrying. I placed them low enough to avoid interfering with the handlebars. Where to put the batteries now? They will likely hang behind the trash cans, but for now they are IN the trash cans.

Speaking of batteries, charging 12 batteries one by one is no fun. Each battery takes about 20 minutes tops– depending on how far it was discharged– but that’s still 12 batteries that need to be baby sat. I inquired on the Endless-Sphere forum about whether the Makita fast charger allowed charging of multiple packs in parallel and it does work. Fellow e-biker Jonathan D. hooked me up with 2 Makita LXT shells– battery packs with no batteries inside. I opened one up and soldered a wire to the + terminal and another to the – terminal and routed them out of the shell through a couple of holes near the front of the shell. I attached an anderson connector. I then tortured myself trying to work with the security screws that hold the shell together before I just dremeled their heads to turn them into slot head screws.

I then rewired the 2 packs I built to be 18v each– all batteries are in parallel. Now I can put the shell on the charger and attach it to one pack and it will charge the whole pack– theoretically. It should take 4 x 20 minutes for each pack, but at least I don’t have to baby sit it. Turns out, however, that the 7 little pins in a yellow connector on the battery report some sort of status information to the charger and the charge always fails after two minutes with the shell. DoctorBass on endless-sphere suggesting opening one of the working Makita’s and soldering the anderson connectors to it’s terminals.

Finally I fixed the kick stand; leaning the bike against stuff was getting really obnoxious as its now about 40-50lbs fully loaded and would probably get damaged in a fall.

Testing

Former partner in crime Eron V.’s holiday party provided another opportunity to test the bike. We took turns giving party goers rides and then turned the bike over to the more adventurous. So far about 12 people of varying stature and weight have ridden the bike, mostly in the tandem configuration, and I haven’t found any problems except that the seat post keeps coming lose and twisting in the socket. High speed ride and stability is actually better with two riders, but a very low speed oscillation in the steering gets worse with two people. If you ignore it, the bike moves in a straight line, but at about 2-3 mph there is a feeling that the bike wants to turn back and forth and it’s unnerving.

The greatest load carried so far has been 380 lbs with a 240lb rider in the back.

I rode from home to work twice, once alone and once with my significant other on the back.

Average cruising speed (alone): 17.5mph. The bike seems to hit some controller limit at 18.5 on the flat
Average cruising speed (two riders): 16.5-17.0mph (depending on rider). Max speed drops also by 0.5-1.0 mph.
The motor got much hotter with two riders.

Ride time (alone): 38 minutes
Ride time (two riders): 40 minutes

Battery state was

One rider: v₀=39.3V, v_f=35.3V
Two riders: v₀=39.3V, v_f=34.5V

Low voltage cut off is set below a safe level right now, but 28V would be a sensible limit. Voltage sag on full throttle acceleration is about 1.0-1.5v.

Odometer currently reads 43.8 miles.

Next upgrade is a head light! Very hard to ride through the marshes with no light.

I got the hall sensor wires worked out– the +5v wire was rubbing against a sharp part and actually got cut. I fixed it and the bike worked. I took it to work and to the San Jose EAA meeting and everyone took turns riding it and then going 2-up. Initial results are very promising: the suspension works great, even with two riders weighing 350-400lbs total the ride is very good. Acceleration, especially up hill at the low end, however, is poor. I’m not sure why it’s as bad as it is because the previous version worked quite well in San Francisco. Is the bike that much heavier now? Could there be some mechanical drag like the coaster brake engaged on partially? Or is it due to the hall sensor wiring?

I removed the now useless 2nd bottom bracket and drilled holes in front of the rear wheel in the triangle for an aluminum tube. The aluminum tube is super lightweight and should work just fine as a foot peg.

Best upgrade so far: I fixed up the driver’s seat. I used two L shaped pieces of pipe, attached the back rest at lumbar height with a 1/4″ bolt and large washers, then attached a padded seat to the bottom by using a piece of pallette wood. The real bastard was attaching the seat to the frame. This I accomplished by welding two tabs onto the frame that fit between the two L shaped pipes that are the frame for the seat. I welded 3 small pieces of 1/4″ steel plate together and drilled a 3/8″ hole through them to make the tabs. I then drilled 3/8″ holes in the two L shaped pieces of pipe and stuck bolts through. 1/4″ is not enough! I tried and the bolts simply bent and had to be removed with the Dremel.

I placed a large exercise machine seat for the rear passenger. To give them something to hang on to, I welded some kids bike handlebars (the type with a big U shape to them) onto a seat clamp, which I then clamped onto the original seatpost.

I added under seat steering. Tom Kabat of Woodenbikes.com had this on some of his bikes and it is super comfy. I cut some kids handlebars in half and welded a 90 degree angle in between. Tom bent some pipe into a large U shape, which would be preferable, except that I can’t bend pipes (yet). I think a V shape would be ideal. Basically, with a SWB layout, the steering tube is just a little bit forward of where it would ideally be. You can put above seat steering but it has to be high up to not interfere with your legs (if you use standard handlebars). With under seat steering, the handlebars have to come back far enough to reach your hands and have to avoid interfering with your legs when you are stopped but at the same time the closer in to the frame you bring the bars, the less you can turn the front wheel at low speed (because your hands hit against the frame).

Unfortunately the Kollmorgen’s hall sensor board did not like me resoldering the wires (which were flimsy after all the work done around the motor)! Two of the traces flaked off. I had to Dremel around the board to remove it and try some surgery. So far, the fix is not working, but I’ll give it one more shot before swapping motors. Never again will I buy these sensored BLDC motors/controllers! So much trouble for too little gain.