Electric chopper trike build journal

cboy

100 W
Joined
Sep 18, 2017
Messages
179
Location
CA Central Valley
My goal for this project is to build a commuter vehicle capable of highway speeds (60 mph), a reasonable range (25 miles) and a budget that won’t require a second mortgage ($5K).

I have chosen to use a three wheeled trike design for the following reasons:
a) I am familiar with it from prior builds.
b) It is relatively light weight.
c) It is more stable (for me at age 72) than a two wheel “motorcycle”.
d) It is easier to license here in CA as a motorcycle than if it were a hand-built four wheel vehicle which would require licensing as a car.
e) although the trike will be “open air” I am fortunate to live in California’s Central Valley where year-round motorcycle riding is quite common.


The trike will be built using two 3000W, QS Motors hub wheels mounted on the rear. The hubs have 16" rims, 273 ring size, 40H magnet size and are dual shaft, requiring drop outs on each side of each wheel. The other major electronic components include:
*Two Kelly controllers model KLS7230S (includes regeneration, electronic reverse, and cruise control)
*Two controller heat sinks
*Cycle Analyst V3 instrument cluster
*Dunlop Motorcycle tires (150/80/16)
*Kelly 400A72V contactor
*Holdwell ED250B-1 emergency shut off
*Hydraulic disc brakes on both rear wheels
*Dual disc hydraulic brakes on front wheel
*Mechanical parking/emergency brake kit
*6 - 12 volt EverStart 29DC deep cycle lead acid battery pack rated at 125 amphour.
*4 awg welding cable for all main battery wiring.


I will be posting my progress to this thread as well as a somewhat more detailed version on my own site, https://hotrodjalopy.com/. Once on that site just click on the Electric Chopper Trike button to navigate to the build. You can follow the build and ask questions on either site but probably better to ask here...much more traffic on this site.
 
Donor Bike

Here is the donor bike I'll be using for the project, a 1989 Kawasaki Voyager 1200 which fit my needs and my budget ($500).

kawa-2.jpg


The bike was in good running condition when I bought it and it had a clear title and the registration was up to date. So I will be able to use this title as the basis for legally registering the vehicle after the modifications are completed. The parts I intend to use from the Voyager include:

*Front fork assembly
*Front wheel and tire
*Steering head
*Front dual disc brake assembly including hand controls and hydraulic system
*Seats
*Luggage carrier including adjustable (sliding) base and rack
*Progressive coil over shocks (Series 412 Model 4221)
*Headlight switch and dimmer
*Turn signal switch
*Rear brake pedal assembly
*Horn(s) and horn button
*Rear brake light, turn signal and running lights
*12 volt charger (for my separate 12 lighting system)
*I also intend to use the 97 HP Voyager engine and transmission in a shifter cart project which is next in line after this trike is completed.
 
Stripping down the donor

The first step of the build process is to totally strip the Kawasaki down and set aside parts and pieces to be used for the trike build.

002a-1.jpg


The main pieces I am after on the donor are the front fork/wheel/ and dual disc brakes assembly. On the Voyager the front fork is welded rather than bolted to the frame so it must be carefully cut away. I left the "junction box" welded to the steering head so that I have something to weld the frame too and not mess up the steering head itself.

009a.jpg


016a.jpg
 
Fabricating the Base Frame

The pieces for the base frame are cut from 1x2x.090 wall rectangular steel tubing.

004a-1.jpg


The frame pieces are squared up for welding. The red arrows note where the side rails were pie cut and bent inward.

010a.jpg


The base frame welded up.

012a.jpg
 
Mounting the Front Fork

The front fork will be set at a 28 degree angle, or rake (matching the angle of the original voyager) and held in place using two front upright supports and two angled supports. While being set up for welding, the fork is held in position on an adjustable tripod (the orange legs in the first photo) to keep it at the correct rake angle. Note the lengths of 1x2 tubing (red arrows) which have been clamped to the side rails of the base frame, squared, and extended forward. These rail extensions act as guides for centering the front wheel and fork with the frame base.

017a.jpg


The steering tube and front fork have machined “stops” (red arrows in the photo below) and a centering pin (white arrow) which prevent the front wheel and fork from being turned too far left or right thereby creating a dangerous or unstable condition. By fitting identically sized wooden shims between the center pin and the stops on each side, the fork and the “junction box” can be clamped in place at dead center to insure the junction box, steering head and front wheel will be pointed straight forward when the fork is welded to the frame.

001a-1.jpg


023a.jpg


The two front uprights and the two angle supports to the frame are then welded in place to permanently mount the front fork.

010a-1.jpg
 
Have you determined CoM yet? A common issue when duplicating a bike's steering geometry, is that it typically places the CoM too far forward, and too high.
 
amberwolf said:
Center of Mass

Short answer, no. I stayed strictly with the head angle, fork offset, trail and any potential wheel flop type calculations. In the end these all led me to just stick with Kawasaki's original specs. Based on your post I did take a look at the center of mass equation and found it somewhat above my pay grade. I can understand what it is attempting to calculate, but getting there is another matter. Just too many variables in a project with the size, shape and weight distribution of this bike. Hats off to those who can understand it and use it effectively.
 
Some ES members here seem convinced that the 3-wheeled (non tilting) vehicles are inherently unsafe. Fact is, when the 'trike' builder employs a few basic design principles, the end results can be, and often is,... just as safe and dynamically effective as a 4-wheeler. But if you ignore the well established design principles, you could be unknowingly building yourself a coffin.

Start here: https://www.rqriley.com/3-wheel.htm

Then examine the image below:

https://www.rqriley.com/images/fig-3whl.gif
 
Basically keep the weight (batteries, controller, seat/rider) low and between the two rear wheels. The geometry of a two wheeled motorcycle is not suited for a single front wheel trike. It will give very nervous steering at speed. You need way more head angle / rake and trail. Look at the old VW trikes and their geometry:
6968377101_2270d3e040_c.jpg
 
Papa said:
Fact is, when the 'trike' builder employs a few basic design principles, the end results can be, and often is,... just as safe and dynamically effective as a 4-wheeler. But if you ignore the well established design principles, you could be unknowingly building yourself a coffin.

My motto is, never argue with advice. So in keeping with that I would encourage any would-be builder to read as much as possible regarding center of mass, as well as all other motorcycle/automotive geometry and determine for themselves its usefulness and applicability for their particular design and fabrication issues.
 
SlowCo said:
Look at the old VW trikes and their geometry:

Yes, some trikes do have extended front forks or a more extreme rake. But you can also find plenty of photos like this:

trike-c.jpg


this

trike-b.jpg


or this

trike-a.jpg


which appear to be stock or nearly stock rakes on custom fabricated bikes.

Don't get me wrong. I'm not arguing one way or another. And perhaps you are correct about more rake. But the photo evidence we can all look at would indicate trikes have been built in all sorts of configurations and any one or two photos may not tell us much...at least not without a lot of hard evidence regarding the handling characteristics of that particular bike/trike. On the other hand, I totally agree with your premise of keeping the "load" low and centered.
 
Battery and Electronics Enclosure

The batteries and electronic components will be located between the rear wheels and behind the seat. The box will be 24″ x 29″ x11″ (external dimensions) to accommodate the six lead acid batteries. 1x1x.0625 square tubing is cut for the perimeter of the battery box.

001a-2.jpg


The tubing is squared up and welded to the frame base to form the box.

005a.jpg


Additional 1×1 tubing is cut and welded to form the box’s angle bracing. This bracing provides support and strength to the box and to the frame itself.

010a-2.jpg
 
OT from your build, but on the topic of trike turn-roll characteristics:

FWIW, I was advised for SB Cruiser it should have a much straighter (more vertical) steerer, but as it worked out I ended up with a much more relaxed angle, and it handles very well this way. It might handle as well or better the other way; maybe someday I'll build a version like that just to try it out.

It's not a high-speed trike, 20MPH max allowed here, but it corners at or nearly at taht speed.

It's got almost all it's weight low, below the tops of the tires, and most of it is at axle height. I plan to move the batteries from the seatbox down to under the cargo deck but have to build a box for that first (actually a few boxes that will bolt on independently for easy service); this will probably help improve turning radius (without rollover).


BTW, regarding pics of various builds: While I have little idea which rake and trail is better in the discussion above, I never trust that a certain method is more commonly found in photos or build-pages, etc., unless performance information is also present to show me how it worked for them, good and bad, so I can compare them to what I want out of something.

A few examples: I see a lot of pics of soldered-together 18650 packs, but most of the builders never show how it performs over time with reliably-gathered data, so I wouldn't be able to judge if they're any good or not. I see a lot of photos and videos of various lithium battery packs on fire and having burned, but it doesn't mean that just because they're lithium they're bad. I see a lot of pics of hubmotor wheels with really thick spokes and crappy bicycle rims...but I know from experience and experimentation of my own (as well as that of others proven reliable to me) that that is counter to the way a "good" wheel is generally built.

So I've learned (the hard way) to not just trust what I see a lot of pictures of. ;)


Plus....sometimes a particular image search on a particular day may bring up one set of things more than another, on a different day it might be the other way around. :(
 
Swing Arm Fabrication


The rear wheels will be mounted to the frame using U-shaped swing arms which support both ends of the wheel’s axle shaft. Fabrication of the arms begins by making drop out brackets for the axles. The drop out plates are cut from 1/4 inch steel plate. Two holes are drilled in each plate. The larger whole is drilled to match the axle diameter. The smaller whole is for mounting an axle torque plate. The holes are drilled using a jig on the drill press to insure all four drop out plates match up. The torque plate is also made from 1/4 inch flat stock and is designed to prevent the torque of the hub motor from twisting or spinning the axle within the slot of the drop out. The QS hub wheels come from the factory with prefabricated torque plates which have a very precise fit on the axle shaft. The axle holes in the drop out plates are cut open to allow the axle to slip out the bottom.

004a-3.jpg


The side arms for the swing arms are cut from 1x2x.090 rectangular tubing. The tubing is first cut to TWICE the length for each arm. My arms are going to be 13 3/4″ long but this will vary depending on wheel/tire size and other design differences. Draw a line exactly half the length of the tubing and then drill a hole 1 ½” in diameter in the center of the arm. I used a metal cutting hole saw in the drill press to do this.

010a-3.jpg


Cut the 1×2 tubing exactly in half based on the center line which was drawn earlier. This cut will also be directly in the center of the 1 1/2″ hole.

012a-1.jpg


You now have two arms notched to fit quite nicely around 1 ½” O.D. tubing and ready for welding. The pivot tube for the swing arm is cut from 1 1/2″ O.D. tubing. The length of the pivot tube will vary depending on wheel width and mounting position considerations but for this project the tubes are 13 3/8″ long.

015a-2.jpg


The drop out plates are tack welded to the side arms and then the side arms with drop outs are bolted to each side of the wheel axle. The pivot tube is lined up in the side arms, clamped in place and tack welded.

021a-1.jpg


The completed left swing arm.

026a.jpg
 
SlowCo said:
Basically keep the weight (batteries, controller, seat/rider) low and between the two rear wheels.
To avoid unnecessary surprises at speed, the trike's laden weight distribution should be in the neighborhood of 33f/66r (placing equal weight on all tire contact patches). So if mounting the batteries between the rear wheels satisfies that criteria, then fine.
SlowCo said:
The geometry of a two wheeled motorcycle is not suited for a single front wheel trike. It will give very nervous steering at speed. You need way more head angle / rake and trail.
From a dynamic prospective, the steer tube axis angle is largely irrelevant. However, relaxed steer tube angles of less than 50 degrees is quite common - but its primary purposes (combined with suitable tiller), is to ergonomicly position the handle bars for the repositioned pilot... with the option of placing the pilot further rearward if the laden WD and/or vertical CoM deems it necessary. Yes, relaxed steer tube angles alter steering response somewhat, but if a suitable wheelbase is achieved, then it's largely irrelevant.
 
Papa said:
From a dynamic prospective, the steer tube axis angle is largely irrelevant.

The head tube angle is one of the largest influences on how direct the trike will steer.
Imagine a non leaning trike with a 90 degree (vertical) head tube. Then one degree of rotation of the handle bar results in one degree of steering. Now imagine a non leaning trike with a zero degree (horizontal) head tube angle. Then no amount of rotation of the handle bar will result in any steering (not 100% true as the front wheel will tilt and result in some directional change).
So the steeper (towards vertical) the head tube angle the directer the steering will be. The faster the trike the less direct you'll want the steering because at speed it will result in very nervous behavior. And then there is trail. Too little will result in the trike wanting to find it's own way (into a ditch...) and too much of it will result in the flopping shopping cart effect.
 
Brake Caliper Mounting

By pure dumb luck it turns out the brake caliper can be mounted in the correct position on the disk by simply dropping a piece of 1/4" flat stock directly down from the swing arm drop out plate and then bolting the caliper onto the plate. I made a pattern on card stock paper to get the bolt holes right and then drilled the plate for the mounting bolts.

002a-3.jpg



Here's a shot of the simple plate/bracket bolted to the caliper. Not very pretty, but functional.

004a-4.jpg


The wheel and swing arm are positioned on the trike's frame and clamped tightly in place. The caliper and bracket are fitted in place with the brake pads slipped over the rotor to insure everything lines up correctly and that the caliper does not interfere with any frame components. The bracket (see arrow in Photo 3) is clamped to the swing arm and then tack welded in place. The caliper is unbolted and removed and then the bracket and swing arm are removed from the wheel and trike as one piece for final welding.

013a.jpg
 
SlowCo said:
And then there is trail. Too little will result in the trike wanting to find it's own way (into a ditch...) and too much of it will result in the flopping shopping cart effect.

Steering trail is a huge can of worms with trikes. With two-wheelers, you can pretty much home in on a desired range of steering trail just by knowing how fast you want to go. But on a trike, trail is almost an enemy. Instead of being the lever that helps a two-wheeler stay underneath you without any effort on your part, it's the lever that causes the trike to actively steer itself downhill whenever the surface is unlevel. It's closely correlated with chassis drop when turning, which makes the trike not want to go straight at low or moderate speeds.

With a trike, you want just about as little trail as you can get away with-- basically, however much trail keeps typical bumps and roughness from turning the trail negative. That's a function of what kind of bumps and roughness are typical where you ride.

Real world example: Most pedicabs use forks designed for bicycles. Also, most of them have very bicycle-like head angles. This results in bicycle-like amounts of steering trail. When a pedicab like this is left to its own preferences, the front end falls into a roughly 90 degree turn. This tendency is what the rider has to fight against at all times. It's tedious.

The pedicabs I design and build have relatively steep head angle with unusually long fork offset. This minimizes trail, downhill steering bias, and wheel flop. Longtime riders usually have a mental adjustment to do when they switch, but it's much easier to ride one like the ones I build.

Another real world example: one of my good Seattle buddies had a chopper trike with 20 inch wheels, roughly 50 degree head angle, and a stretched fork of close to 30 inches axle to crown. It was breakage-prone and an odious chore to steer. By doing a geometric analysis of the steering trail characteristics of the trike, I discovered that the correct amount of fork offset was in the range of 125mm or 5". I made him a triple tree fork with crowns in the shape of equilateral triangles. It cured all the steering problems and beefed up the fork enough that the frame became the new weak link (which expressed itself later).
 
Chalo said:
With a trike, you want just about as little trail as you can get away with--
Yup.
Chalo said:
The pedicabs I design and build have relatively steep head angle with unusually long fork offset. This minimizes trail, downhill steering bias, and wheel flop.
I commonly refer to this as "neutral" steer.

Because all of my hybrid HPV trike builds are 'feet forward' recumbents, I've grown accustom to 50-57 degree steer axis angles to achieve acceptable WD and ergonomics. In addition to those slack angles, I habitually apply 4 to 5.5 inches of fork offset, specifically to suppress positive trail, (exact amount is also dependent on front wheel OD). The example linked below has 4.75" of offset w/18" (355) front wheel, and it's barely sufficient.

https://endless-sphere.com/forums/viewtopic.php?f=11&t=92604&p=1354360&sid=83ba4b01bf95252f5dd16cd0e9cd73c2#p1354360
 
A swing arm mounting bar which spans from wheel to wheel is cut from 1x2x.090 rectangular tubing and clamped to the existing frame of the battery box with the 2″ dimension in the horizontal position. (See upper arrow in Photo). A second bar is cut from perforated channel strut and is welded with the long dimension in the vertical position to the underside of the 1×2 tubing (see lower arrow in Photo). This prevents the mounting bar from flexing either vertically or horizontally.

026a-1.jpg


Brackets for attaching the swing arm pivot tubes to the mounting bars are cut from 1/4″ flat stock and drilled with ½” holes. There is a size and shape difference between the inside brackets and outside brackets but all four bracket holes must line up in the same position. So as each plate drilled it is marked for proper positioning. The exposed corners of are trimmed away and then ground to a smoother shape.

020a.jpg


Half inch steel rod is cut to length to fit the pivot tube length plus bearings plus brackets plus end collars.

022a-1.jpg


Flanged bearings with ½” I.D. and 1 3/8″ O.D. are fitted into each end of the pivot tube (see arrow in Photo). These particular bearings are rated for a dynamic load of 4900 lbs.

024b.jpg


With the swing arm, wheel/tire, and pivot tube square to the frame, the pivot tube brackets (arrows in first Photo below) are clamped and tack welded to the horizontal mounting bar. Once it is certain everything is square and true, the brackets will be permanently welded to the mounting bar. The passenger side swing arm is then welding in place in the same way. The 1/2″ rod is capped on each end with a 1/2″ I.D. collar and set screws.

016a-1.jpg


015a-3.jpg
 
It's a Roller

This is a fun day in the progress of any project. The day you can roll the beast out of the garage on its own two, three or four wheels. The chopper trike has no rear suspension yet but with the swing arms clamped in position the bike sees daylight. These photos also begin to hint at how the trike will look in a more finished state.

005b.jpg


003b.jpg


010a-4.jpg


008b.jpg


009b.jpg
 
Rear Suspension

Positioning the rear spring/shock mounts and setting them at the right height can be a daunting task on any scratch built project. During the fabrication process there is no weight on the frame and the mounting points for the top and bottom of the spring/shock are just theoretical points in space. Once the vehicle is completed and under full weight, including the rider, the suspension is going to compress and the vehicle is going to end up at a different ride height than during construction.

The Kawasaki donor had a nice pair of aftermarket Progressive 412 Series, adjustable coil over shocks on the rear which will be used on the new trike. Calculating final ride height and shock compression can be even more of a challenge when using progressive coil springs such as these. With a progressive spring, the spring rate increases as the coil is compressed. My 412-4221 coils, for example, have a spring rate of 140 lbs per inch when fully extended. But they have a spring rate of 200 lbs per inch when fully compressed. And that rate is constantly changing and getting greater over the full 3.52 inches of spring travel. Theoretically the spring rate increases with every incremental increase in compression distance. In actual practice, however, the increase is not perfectly linear. But for estimating ride height one can assume a linear rate increase and get things relatively close.

For this trike I wanted the springs and shocks to be compressed approximately ½ way when the trike is completed and sitting at ride height with the rider aboard. So I calculated an average spring rate over the first half of the spring’s compression (155 lbs per inch for my shocks). I then estimated the total weight of the finished bike and rider AND the estimated weight distribution front and rear. In my case the battery pack will weigh 360 lbs and I estimated 85% of that over the rear wheels and 15% over the front. The rider weight will be 170 lbs distributed 50/50. The frame, seats and all other components were estimated at 150 lbs also distributed 50/50. These calculations put the total weight over the rear wheels at 466 lbs. That weight will be distributed 50/50 between the rear wheels so each spring/shock will support 233 lbs. Dividing that number by the spring rate of 155 lbs per inch indicated each spring would compress approximately 1.5″ with the bike finished and the rider aboard. I wanted the finished bike to sit level at 6″ above the pavement with the rider aboard so I set the frame on 6″ blocks at the front and 7.5″ blocks at the rear. The swing arms, however, are positioned absolutely horizontal to the pavement. With the spring/shock fully extended and the lower shock eye bolted to the swing arm, the position of the upper spring eye can then be identified relative to the frame and future shock mount.

Since my knowledge of spring physics is quite limited and my mathematical and theoretical calculations always suspect, I allowed myself a fudge factor during fabrication by making my shock towers adjustable. I cut the 1×1 tubing for the towers about 2″ longer than my calculation and then I drilled a series of 11 holes, at half inch intervals, in the tower. (Photo below) This will allow for adjustment of the upper spring brackets either up or down should the spring position calculations be in error. Once the trike is complete any excess holes at the top of the tower can be cut off and removed.

021a-2.jpg


The upper shock mounts are created by making 3″ x 4″ plates cut from 1/4″ flat stock. Using a simple jig in the drill press, the four shock mounting plates are drilled out.

007a-1.jpg


The large single hole in the plate is for the shock eye bolt and the two smaller holes are for bolting the plates to the shock tower. Each plate must also be notched on the bottom to clear the top of the coil/over shock. A total of four matching plates must be made, two for each shock tower.

011a.jpg


The plates are bolted to the spring tower at the estimated height indicated by the spring rate calculations. Note the extra mounting holes for adjusting the ride height.

024a-1.jpg


The lower spring mount is much easier. It is a 5/8″ grade 8 bolt inserted through the swing arm and drop out plate.

017a-1.jpg


To align the top mounting plates with the lower mounting bolt, a spacer cut from 1″ x1 1/2″ rectangular tubing is welded to the shock tower. (See red arrow Photo below)

030a.jpg


The spacer and shock tower will be welded to the frame base and battery box in a vertical position. The tower, however, will be positioned so the upper shock mounting hole on the tower will be a little forward of the lower shock mounting hole on the swing arm. This creates a forward shock angle of approximately 10 degrees. Putting the spring/shock at an angle reduces the effective spring rate. I used an angle similar to what was on the donor Kawasaki. If the ride is too harsh or to spongy once the bike is completed, the Progressive 412 shocks can be adjusted to compensate for the miscalculation. With the swing arm parallel to the ground and the lower spring eye attached to the lower mounting bolt, the shock tower with mounting brackets attached can be positioned on the frame so that the upper spring eye, with the shock fully extended, fits exactly between the holes of the shock tower brackets. If things are “off”, the shock tower brackets can be moved to a higher or lower hole in the tower until everything lines up. When completed the shock tower should be in a vertical position with the bottom of the tower even with the bottom rail of the battery box. The photo below shows the shock tower welded in place after the position had been established with the swing arm and spring/shock in place.

004a-5.jpg


With the shock towers welded in place the swing arms are once again installed and the coil over shocks can be bolted in position.

013a-1.jpg


To see if our calculations, estimates and guesswork for positioning the springs is anywhere near correct the batteries are loaded into the battery box . At this point the trike is still about 3/4″ above my design ride height. Once the rest of the framework, electronics, seat, and driver are added, the trike should be within 1/4″ of where I want it to sit. So I may have lucked out with my original calculations and hopefully I won’t have to make any major alterations.

020a-1.jpg


With the upper shock mount position now established and reasonably tested for accuracy, an angle brace is cut and welded to each shock tower to triangulate with the frame and stabilize the towers.

026a-2.jpg
 
It's coming along nicely. With those big motors and the heavy load you want to beef up the dropouts, and if you plan to use regen braking a clamping type arrangement is necessary to prevent the motor axle from rocking back and forth and wallowing out the dropouts as well as deform the axle flats. All of the forces to move the trike are transferred from the motor via the very short radius of axle flats, so it's literally thousands of pounds of force...if powered adequately those big 273's may push past 10,000lbs of force. You seem to like to build strong stuff, so don't ignore the one part where there's no such thing as overkill.
 
Back
Top