Tribrid bike

Making it a Tribrid

(Gasoline, electric or pedal power)

Why?
The electric motor gets me through the gravel pit and long driveway from the shop to the road without totally exhausting me, gets the gas engine started without undue stress on my old back and knees. Plus I have backup for the 7 mile trip to town and back in case the gas engine fails. The peace of mind is worth a lot.

20 May 09
The bike is now a hybrid (or maybe tribrid) with the addition of a 600 watt, 36 volt electric motor that I bought from this fellow: http://shop.ebay.com/merchant/qvea
It comes with battery pack, headlight, horn (only one I have seen with these). On top of the headlight shell is a semi-circle that is acts as a state of charge indicator, a power on indicator, and a headlight on indicator. Also features electric braking and even came with some half-way decent tools. The e-braking is so feeble, I could not justify the added home brew switch bracket on the V brakes, so I clipped the wires.

Hybird

The wheelbase is 2" less than most bikes. The controller has been moved to the front of the frame.

The 4 red wires sticking out of the battery box allow checking the individual battery voltages, desulfating, or a 12v charger. Desulfating the new batteries increased the voltage by .45 volts each, which may indicate they sulfated slightly in the warehouse before shipping. Not really a problem, but I hope to make these batteries last a long time.

I opened the battery charger that came with the motor kit and turned the tiny pot next to the output cable so it shuts of at 41-42 volts. It had been a bit high @44 volts. 41 volts may mean I have to let the batteries rest overnight and charge again, but perhaps I am being overly cautious by reducing the voltage.

The controller main chip is a HA3089, which appears to be a PIC microcontroller. Installation went smoothly in spite of having a gas engine to deal with. First test I got it up to 17 mph on a slight upgrade and a good headwind. It pulled the bike through the gavel pit of a driveway. Update Top speed on this bike is 17 electric. With the front wheel raised it reads 26 mph.

I could not think of a way to make a dual throttle, so I replaced the gas throttle with a thumb lever I made from an old gear shift. It is mounted under the bar so it can be activated as the e-throttle is twisted.

The wiring looks similar to this

Now the problems... As soon as I switch the ignition on and get the gas engine turning, the e-motor dies. I think this is why... The electronic ignition interferes with the Hall Effect sensor lines that tell the e-controller the status of the motor. Wired as it was designed, the wires pass directly above the ignition. The cure should be to shorten the lines, twist and braid them and encase them in shielding. Not sure yet, but I may have to do that to the throttle too. I made a bracket that is riveted to the front of the diamond, just behind the front fender for the controller in order to have the control wires as short as practical. In the photo above the controller is at the rear below the battery box.

I feel that the connection to the motor should be at the axle instead of at the battery box. This will make changing a flat far easier. As it is, you have to take all the wiring loose to get the front wheel off. Not something I would like to do on the road.

2 June 09
Okay, I have shielded the control wires and it seemed to help. When I am able to use the e-motor with the g-motor, it really accelerates. Last thing I did was shield the spark plug wire. It did not make any difference at all.

The bike does what I originally wanted; a way to get it through the gravel to the road with exhausting me, allows back up in case the g-motor fails, and store-to-store hopping on the e-motor.

I looked at the blue wire. It goes directly from the magneto to the CDI. The ground side of the CDI went to the kill switch, where opening the switch kills the engine. That is now replaced with an inline mini phone jack at the CDI. The phone plug is shorted, and attached to a SS lanyard and ring I happened to have. Locktite was used on the screw that attaches the lanyard to the plug. This is not as handy as a kill switch, but should cut down a little on electrical noise.

Now I can use both motors at the same time up to about half e-throttle. Maybe with some more thinking I can solve that.

7 June 09
I checked the power transistors and they are insulated from the heat sink, which means the controller case has no ground to the circuit at all, so I will not insulate the case from the bike frame. Not sure yet about the hub motor. Will be difficult to isolate from ground.

Tonight I thought it might be a good idea to braid the power wires from the battery pack to the controller. Plus 35v, ground, and the power from the key switch. Only took a few minutes, and there was quite a bit of difference. Most of the e-throttle can be used now especially when the g-motor is at a higher RPM than idle. Getting closer to what I want. Next will likely be a sprinkling of filter capacitors. I have some 600 volt paper .1µf. Really showing my age with that statement.

Bought an adjustable center kickstand on eBay today and it should be here in a few days. That should allow me to experiment in the shop with the bike stationary.

9 June 09
I had the thought last night that shielding the CDI unit might help a lot. I know there are at least two coils in it that are pulsed making a likely source of RFI. I wrapped aluminum foil around the CDI and part of the sparkplug wire, then jumper wired it to ground. There is a difference. I managed to get two of the e-speed indicator lights to go out. Before could only manage one. (More speed, more go out). Aluminum can shield RFI, but not magnetic interference, so next step is an iron shield. Probable cut from a coffee can and formed around a wood pattern and the seams soldered.

Ok, made the can cover and soldered it to the plug wire braid. Results are not as good as I hoped. Looks cool though.

10 June 09
Aha! A major clue. I measured the resistance between the hub and the bike frame and found it to be 70 ohms. I thought a jumper from the front fork to the frame would help, but to my surprise the e-motor would not run at all at any speed or rpm of the g-motor. I may have to insulate the hub from the fork, but will wait until my center kickstand arrives so I can make voltage measurements.

11 June 09
Speedometer died. I am not going to invest in a third one.
I tilted the bike slightly to the left so the front wheel was off the shop floor and opened the e-throttle. There was quite a bit of out of balance, so I installed one of the white side reflectors on the spokes and by trial and error managed to balance the wheel. Never really thought a bike wheel would need to be balanced, especially since the front wheel weighs 15 pounds, it makes a difference I found.

Very nice kickstand arrived and was easy to install, although it clears my tire by only 1/8". Got the g-motor started and parked on the cattle guard with the kickstand on the pipe so I could run both motors at the same time.

15 June 09
I made a bracket for a microswitch on the front fork and have the electric brake connected now.

Frame to fork .17 volts DC
Battery negative to frame 4.5 volts intermittent
Battery positive to frame 5.5 volts "

I tried a jumper wire from the fork to frame... no difference.
Tried a jumper from negative on the battery pack to ground, still no difference.

Next step is to run a wire from the front axle to frame, a wire from the battery pack negative to that same spot, and another from the magneto.

29 June 2009
The problem with the e-motor not working if the gas engine is running has been solved.

It took shielding the CDI and braided shield on the spark plug wire connected the shielding on the braided power wires between the battery pack and the controller, but NOT grounded to the frame.

The controller is also electrically isolated from the frame (and not connected to the plug shielding), and a resistor spark plug is required. To test it with a resistor plug I used an old Autolite plug from my Lincoln Towncar.

This was a rather tedious trial and error discovery with a lot of strange things happening. What finally led me to the solution was discovering that it worked smoothly if I had my hand on the spark plug shielding while riding, and opening both throttles (while watching for traffic). I don't recommend trying that.

Now it really accelerates hard if I want it to. The front wheel occasionally spins accelerating from a stop. Hang on to your hat!

Next project is a NiMH (nickel metal hydride) battery pack I have 216 cells on the way that I bought on eBay. They are rated at 2.5 amp hours. Feedback from other customers are pleased. The cost was $8.88 per dozen with shipping paid. 144 cells could replace the original lead acid pack and reduce the weight by about 21 pounds.

They advertise the NiMH cells can be recharged over 1000 times versus about 300 times for the lead acid batteries that came with my motor. Taking the number recharging cycles into account, I figure the lead acid batteries cost at least 4.5 times more. Also, NiMH cells put out nearly full current until they are nearly dead. Lead acid batteries lose voltage as they are discharged, and it is not a good idea to discharge them below 31.5 volts to avoid shortening their lifespan. Already the lead acid batteries are showing a little loss of performance.

While waiting for my cells to arrive, I made two "saddle bags" to replace the school bags I had mounted earlier. The new ones are meant as portable file cabinets made by Sterlite. The markings on the bottom indicate they are made of polypropylene (PP). They were $3.50 each at Walmart.

Scrap angle iron from an old bed frame reinforces the sides riveted to a back up plate of aluminum, measuring 1/8 x 3/4 x 9". Wide head 3/16 pop rivets were used.

4 July 09
The new battery pack was tested last night. It goes much faster, and has enough torque now to keep the engine turning over for starting if necessary. I expected that, because the NiMH cells are better at maintaining the voltage under load. And they were not even charged up yet!

180 cells made three 12 volt packs, in series that makes 36 volts. There is a down side (of course). Since the NiMH cells maintain the voltage until almost dead, there is not much warning.

8 July 09
The AA packs are not doing well. Need to come up with a better charger. Yesterday I tested the original lead acid pack for distance. Taking it very easy, I managed 17.2 miles before it looked like the batteries were low. No load they measured 11.8 volts. 17.2 miles is more than enough to get me home if the gas engine fails. I may try it again but at about 3/4 throttle since that is where the best efficiency is supposed to be.

21 July 09
The AA cells I purchased are supposed to be the same ones used in the Toyota Prius. These are a few technical links:
http://www.eaa-phev.org/wiki/Image:EAA-PHEV-PRIUS-ControlBdSchematicV2f_060410.png
http://www.cleangreencar.co.nz/page/prius-technical-info
http://answers.yahoo.com/question/index?qid=20090510110216AAH8aH2

The 10 year/100,000 miles warranty on the batteries is encouraging, but I was surprised to learn that the state of charge is held between 40% and 80% in order to do that. I have 100 three amp diodes on order to make charging a little "smarter".

27 July 09
Trying to trickle charge the packs is very strange. One row of 10 cells read 40 Ma, Adding a second it still read 40Ma. Switching to another pair or even three rows, it still read 40Ma. Small diodes and limiting resistors to each makes no difference. I read contradicting info that they can and should not be trickle charged for extended periods. I will be on the safe side and limit the trickle charging time. If the 40Ma still holds true when my 3 amp diodes arrive, I will use that as a benchmark for full charge.

____________________________

My mind is still on building a Velomobile someday, however I keep hearing more on lack of stability of three wheels. Four wheels would be best, but legal issues arise. Depending on the state, licensing, insurance, safety inspections and possibly a VIN number may be required.

____________________________

Weight breakdown (I weigh 170) oops, I guess I weigh 188
Bike           39.7# (from the shipping box)
E-motor    14.5#
Batteries    28.5#
Gas engine ~20#
Total    102.7# with the original battery pack
Total 82.7# with the NiMH battery pack

Front weight 37.5
Rear weight 76.5 56.5 with the NiMH pack and tools

Cost breakdown
Bike $66 from Walmart, but not really a good bike.
G-motor $149 including shipping
E-motor    $375 including shipping
Total $590 (however there were many items that were added because I wanted to)

Range
electric        17.2+ miles (27.7 km)
Gas            ~70 miles (112.6 km)
Total            87 miles (140km)

Links
tribridbike.com Innovative ideas using a portable generator

Back to 80cc page