Dirt-E Bike

Kawasaki KE-175 Conversion - Part 1 ...

Nov 2007

As with most projects featured on theworkshop.ca this one is based on an item that I salvaged from our local scrap yard.

A 1980 Kawasaki KE-175 rolling chassis, though as I got into it I found that it was just barely "Rolling"...

Again my thanks to Jeff @ Yolkowski Lumber & Salvage, regardless of the condition $20.00 is a great deal for a project like this.

The bare frame is shown here after I plasma-cut 13 bits & pieces off that were for the original ICE (Internal Combustion Engine) hardware.

Although I had no idea what direction I was going to take for mounting parts, I wanted essentially a blank canvas from which to work.

With the bike all over the shop I started with the front-end.

The wheel was seized due to the rust that caked the inner drum and shoe lever.

Some sanding, lots of WD-40 and elbow grease to free everything up and it's as good as used. 

The front-end had a tremendous amount of slop in it and felt quite rough when turning the handle bars from side to side.

It seems that the Head-Tube bolt was loose or loosened, such that half the ball-bearings were missing.

I found great replacements from 2 crank bearings found on an old mountain-bike. The bearings are identical in size or within a couple thou read on a micrometer.

It started to become a hassle to keep the bike set-up on a milk crate, so I fashioned a simple kick-stand and return spring to replace the one that was missing.

Not really a big deal but something so simple can take me a day to fabricate, between finding suitable stock, misplacing tools, and losing the actual parts as I go along.

The brake control lever, control arm and pedal were also missing, so those items were fabricated from scratch over a week or so. Similarly the rear hub was in need of a good cleaning though the control itself was in better shape than the front.

With this the 5th in a series of progressively more advanced Electric Vehicles built at theworkshop.ca, I learned that if I don't resolve fundamental issues such as brakes I will regret it later as the electronics and drive components are installed and the itch to go for a quick test rip becomes too strong to resist. 

Disclaimer!!!

The following section outlines procedures and practices that void warranties, stress components beyond the manufacturer specified limits and may prove hazardous to the operator and bystanders... 

In an ideal world, I'd be outfitting this bike with an "ETEK" 7 to 15HP pancake motor, but my world is less than ideal, so don't email about how great the Etek is, and that it's the only way to go etc, unless of course you have an Etek that your wanting to send my way.

Once I was satisfied that I had enough of a basic frame to work with, I got into the conversion parts.

This test bed is made up of 4 (four) 12V 18Ah Sealed Lead Acid (SLA) batteries, hooked in series to total a 48V Pack.

The Controller is a 48V 50Amp Yi Yun LB-37 from TNC Scooters. This is the same controller that trashed the gear reduction unit on the LEV-1 project, so I know it has lots of power.

The throttle is a standard Hall Effect Sensor twist unit, while the motor is a Unite Motors MY-1020 750Watt unit.

Although the motor is rated at 750Watts the Voltage parameter is less clear, UNITE Motor spec's it at 36V and 27Amps. But after doing lots of research on various forums it seems lots of folks are posting good results by "Over-Volting" the motor to as much as twice it's rated Voltage (72 Volts), though the most common is either a 48V/50Amp or 48V/100Amp configuration.

Certainly I'm most grateful for the pioneering work being done by others, but I can't help but feel that I'd best learn as much about the motor myself and try to work some of the simpler maths as I understand them.

As shown above this has a 4-pole commutator, 2 positive and 2 negative brushes. With a Fluke digital meter, I measured between 0.9 and 0.5 ohms on the power leads very consistently. Though with the armature removed from the brushes I read exactly 0.9ohms on the commutator segments that are spaced 180 degrees apart and 0.45ohms on the segments spaced at 90 degrees. 

Also I noted that the brushes have a serrated curvature that is parallel to the axis of rotation of the commutator and will likely perform better as the brushes wear-in. I found this interesting as I'd read a number of postings that stated they that their motor performance increased after an initial break-in period.

The image to the upper right shows the brushes all pre-loaded with the springs hooked down to allow re-assembly, if you've got 4 arms it's not that useful a trick, but otherwise it's a bugger to re-assemble by one's self.

I measured the diameter of the armature wire at 0.030" and found that it lines-up to an AWG # 20 or 21. From the AWG chart there are resistances and current load parameters based on 1000ft, but without unwinding the motor to measure the length of the wire, I'd have to work the measured resistance forward to determine the length and calculate the max current based on the chart. Given that everything hinges on the resistance and it could be off given that it is such a low resistance I didn't bother.

While I had the motor apart I opted to add a series of holes to the front casting to aid in venting the motor when I get into progressively higher voltages.

I'd had great success in adding active cooling to far more expensive motors than this in the past and have no reservations about voiding warranties or the like.

As an aside the gear housing was covered while I was drilling to keep any metal shavings out.

The integrated gear reduction unit has a 9 tooth drive machined directly into the armature shaft, that couples to a 60 tooth output shaft.

The armature and output shafts are fitted with sealed bearings on both ends of each.

Beside the 750Watt unit that I'll be using is a 1200Watt MY-1020 that is rumored to be able to withstand close to 100 Volts, though I wouldn't want to be around if it lets loose.

The 9 : 60 ratio on the output shaft measured between 670 and 690 RPM at full throttle no-load. This caught me off guard as the motor is rated for 2800 RPM, and for it to hit over 600 RPM on the output shaft would require over 4,000 RPM on the motor. That's why I counted the teeth and even measured the motor armature at 4,400 RPM with the gear reduction unit removed.

Given that the Output shaft is set with a 10 tooth sprocket, that drives a 44 tooth sprocket on the rear wheel I thought I'd run some numbers through a spreadsheet to see how the bike would perform in theory at least.

The range between 1500 and 2500 RPM is what I'm hoping to hit as a realistic top speed, though speed is really not what I'm personally after, this bike will have to be able to push through tall grass, and up steep inclines if it will be of any use to me around the farm. Given my poor vision, I could likely break my neck just as easily at 5MPH as at 30 so a slow ride is a design criteria.

This is the underside of the motor mount that has been welded to the swing-arm.

The 3/16th" plate is stiffened a bit with a chunk of angle iron.

The holes have to be reamed straight, and even stiffer bracing added (as I threw the chain twice during a test ride).

Even with the misplaced holes the motor does line-up enough to determine the chain length.

I should note that the wheel is adjusted all the way forward prior to welding the bracket and then cutting the chain.

The main benefit of side mounting the motor on the swing-arm was to have the best chain alignment over the travel of the arm without the complexity of adding a tensioner.

The 48V Pack was measured as 7" by 12" for this frame that they will be bolted into place with.

The framing is bolted to the KE-175 chassis through the front motor mounts that I didn't cut off as I knew they looked handy.

The batteries will be held in a block by a ratchet strap around their outside, and held by the 2 lengths of 1/4" #20 all-thread welded to the battery pack frame.

The batteries are installed and secured by a length of 1" by 1/8" steel strap over the threaded rod.

While trying to find a good spot for the electronics I realized that I should fabricate some sort of seat.

The base is a piece of 3/4" thick pine, under it are 2 lengths of 1/8th" thick strap that were drilled and welded to the frame.

The actual seat cushion is made from a base of 1/2" thick Styrofoam and 2 layers of sponge foam.

The cushion is covered by a gray vinyl type upholstery that I saved just for this type of job.

Every thing is sandwiched and wrapped tight and stapled underneath.

It's not the most comfortable seat, but is head & shoulders above the bare steel tubing of the frame.

The seat is held in place by wood screws and seems quite secure. With the seat being flush with the tube framing, it's impossible for someone to try to lift the bike by the seat, which would be the only way to break it free by mistake.

At this point I'm in a panic to test the bike and opted to fashion a plate where the gas tank had been to mount the electronics on.

Although there is plenty of room under the seat, the long term aim is to hop this unit up to a 72V machine and I'll have to stuff 2 (two) more batteries somewhere.

The base plate is aluminum and will help act as a massive heat-sink for the electronics... Not really needed at this point but as parts start to get pushed beyond their design limits it certainly won't hurt. 

Last details included wiring the throttle, charger port, and a brake switch that cuts the motor power when the front brake is activated.

The bike has a hint of a trials type machine, though I doubt that I'll be riding along downed logs or nimbly hopping up and over boulders and creeks.

I'm hoping for a nice 2 or 3 week thaw just so I can put the proto-type through it's paces before tearing it down for paint and upgrades.

This action shot actually has the rear wheel spinning out though it looks like I'm passing a kidney stone.

Over-all I'm happy with it's performance over a 10 minute session.

It readily spins the rear wheel given the snow and ice, donuts are easy, but the chain was thrown twice and re-enforcing the motor mount is at the top of the list.

There isn't enough level driveway cleared to determine a top speed, of even a feel for how fast it can really go.

And the same with hill climbing, the rear wheel just spins when I get into more than 6 inches of snow (which is everywhere).

If the weather improves fine, otherwise I'll plow a test road of at least 1000ft, as there is no way I'll be able to wait till spring to continue with this project.

Dirt-E Bike (KE-175 Conversion), 2, 3, 4

Back to Energy

Disclaimer (an unfortunate necessity)
All Rights Reserved theworkshop.ca © May 13, 2008