How to build Moonbeam, a 100 MPG microcar

How to Build Moonbeam

one gallon challenge 2011
Summer 2009: The One Gallon Challenge
The Microcar Concept
Why three wheels?
Why Not Electric?
Street Legality
Test Drives
How to Build Moonbeam
photos 1 and 2
Photos 3 and 4
Photos 5 and 6
Photos 7, 8, and 9
Some more Pictures
Improvements you might make
How you may use this information
Links and Videos to Check Out
Maine to Santa Monica at slow speed
Report from Santa Monica's Altcarexpo
Progress Report Jan 8, 2009
Moving On! August, 2011
The Sequel to Moonbeam: Sunbeam

A step-by-step journal



1: CHOOSING THE DONOR VEHICLES  (It takes two,remember?)  


        It’s good to think well before you choose which motorcycles or scooters to chop up for your microcar. I decided that I wanted the following characteristics: 100 miles per gallon, a four-stroke engine with water cooling; an occasional small second-passenger capacity, but usually one passenger and 6 grocery bags; no gear shifting with hand controls only; an enclosed vehicle with a heater for all-weather operation; easy interior access with lots of light; and finally, a nice looking machine, that you looked back on admiringly as you walk away. All in a budget of $2000, including the donor vehicles and 400 hours of labor. A half-time, half-year project. Ha! What an underestimation!

You, yourself, might choose different vehicles than I did, but I think you will conclude that you need a 4-stroke engine for clean emissions, a 150 cc engine for good pep in the 0-40 MPH service range, water cooling for heating the cab in winter, variable speed drive for simple driving, and a 3-wheel configuration to avoid the many 4-wheel vehicle regulations. There are many Japanese and Chinese motorcycles, scooters, and moped possibilities out there, both old and new.

Of course, an important part of chosing donor vehicles is  thinking out as best you can, the overall design.  It's hard for hammer and tong mechanics like me to resist banging away, but a good month of study is worth it.  There are four websites I keep coming back to:   www.maxmatic   and   These will lead to many others.  

I'm too cheap to buy Riley's book "Alternative Cars for the 21st Century"  but I'd bet that $50 would save you hundreds in heartache.

Yes, do study and engineer; but don't get bogged down!  Possibilities can be paralysing.  The ideal can eclipse the real.  I need company!  It's lonely out here, driving the only microcar in the US in daily, year-round use.  So make a deadline and stick to it: I'm gonna be driving this sweetie next June!  And make sure you read the page on 'street legality' before you begin building.

I chose two motor scooters: a Honda 1987 Elite 150 in excellent condition ($400) and, not finding a duplicate, a 1984 Elite 125 parts bike ($100).

I will be quite interested to hear what vehicles you, yourself, choose.  Using 80's scooters involves some antique parts hassles, but by the 90's, water-cooled scooters had grown to 250cc's, which I think is too large.  Honda Helix, for example. The Honda Ruckus, I think is watercooled, but only 50cc's.  I don't recommend an engine that small.  

One long-shot  is the Honda SH150i which is a modern scooter very popular in Europe, and--I love this--has 16" tires and fuel injection. The PH150 has smaller tires.  You could buy a wrecked one, or order a new engine unit,  which is the complete drive from the engine hinge point to the rear tire.  I'm pretty sure, though, this new piece, would cost more than the whole scooter.   It weighs about 90 pounds. 

 Here in Maine, we have a great weekly swapsheet for getting older stuff.  And of course, we're generally frugal as heck.  So it's tempting to put a "wanted" ad for the bikes you want,  in your swap sheet over a good period of time and see what comes up, as well as keeping track of Ebay.

Having driven antique cars and motorcycles most of my life, I see no reason to be deterred by older machinery.  It is simpler, huskier, and the parts chase has been vastly simplified by E-bay.  Many Chinese clone scooters could involve more parts problems than the Japanese brands.   The main consideration--and for me this is important-- is to respect the engineering and quality of what you choose.   Then you will enjoy working with the pieces.

Working with Honda stuff, I am constantly in awe of how thoroughly everything is engineered and thought through.  Who else would have a dashboard warning when it's time to change the oil?

I wasted some time getting two Honda CT-110's in my garage, and was just about to destroy them when I realized that air-cooling, manual shifting, and 110cc's were dead ends I didn't want to travel.  In retrospect, I might have added that large, spoked motorcycle wheels are not strong enough for the side loads of a non-leaning vehicle.

The scooters I chose,  I completely stripped, carefully bagging and labeling all parts, and then sawed through the frame tubes right where the tube enters the rear subframe. I used a reciprocating hand power saw, commonly called a Sawzall, and kept handy a large pack of 14 tooth blades. Gasp! It was hard to destroy a beautiful red motor scooter! See photo 1, which also shows what I am calling the subframe.


     This is a big project, which can easily get bogged down. It’s finally taken about 1000 hours, that is, a year of 20 hour weeks. I work in the world half-time as a handyman, so I could devote lots of time to the project. More likely, building such a car might involve two years.

It involves almost as much time thinking as doing. You need to learn about bending plastics and steel, about steering geometry, how to weld, and many many other things. The project will sometimes catapult you out of bed with inspiration, and equally have you in pits of despair and stalemate. Using used motor scooters and watching costs carefully, about $2500 will do it.

The good news is that, although the project is long, you are never far from driving the vehicle while in the building process. Using an unmodified drive system, the car is soon on the road. Neighbors will get excited, will give you ideas, and before long your whole community will have adopted the project as their pet.


     I wanted to build a minimal frame first and test the vehicle on the road before I went too far with building the body. As you will see, my idea was to join the scooter rear end into two front ends of the same scooter.  See the 'improvements" page for, in retrospect,  an easier way.

I bought an 8’ length of steel rectangular tube which was 2" X 4" in section and an eighth inch thick and sawing 45 degree angles created a "U" shaped piece of chassis. I chose 40" inches as the car’s width, so the sides are 40" on center and the arms extend 18" forward, with caps welded on the open ends. This strong main frame shows in Photo 3, the first road test.

A 40" width, with a wheelbase of about 57", turned out to be a nice size. But when 2 adults are seat belted side-by-side, THEY NEED TO BE ON FRIENDLY TERMS! It’s better if the second passenger is a child.

I would strongly recommend that you think in terms of a 1.5 passenger vehicle. These are only 10" tires. There are drum, not disk, brakes. Especially important, the front suspension, which mainly supports the passengers, has limited travel. Two adults going over a large pot hole might well bend something.

You might choose a larger format, but my interest was always to see how small a vehicle I could use with dignity. I might have gone to a 63" wheelbase and used the extra length for more legroom. But remember: size makes weight. Moonbeam weighs 112 on each front wheel, and 162 on the back, for a total of 386 lbs. It accelerates quickly up to 40 MPH, then slowly on up to 52, but with two adults aboard, it does labor up steep hills.

I didn’t know how to weld, so bought a Hobart Handler "MIG" welding set with helmet, gloves, cart, etc. and had the salesman give me a crash course in welding. Before I started welding the chassis, I forced myself to spend a day practicing on all types of welds on all thickness of steel. Even so, my welds were always amateurish. The MIG welder, which uses inert gas, does make welding a lot easier.

I then welded this "U" chassis to the scooter rear sub-frame, using scrap flat 1/8" metal gussets to strengthen all connections. On the sub-frame, I also lengthened the rear springs by 1" to raise the height a little, and then re-installed the motor unit in the sub-frame.

To begin understanding some of the 3-wheel technical stuff, read everything in this site: Especially note all the front end geometry stuff, and the fact that: "The center of gravity should no farther than 35 percent of the wheelbase from the side-by-side wheels of a three wheeler". This means that the driver will sit further forward than you might imagine.

To position the two front forks, I built a stand, shown in photo 2, which supports both forks at 40 inch spacing, angled together at the top 1-2 degrees (camber) and leaning back 10 degrees (caster). The motor scooter caster of 27 degrees would make steering too hard. With this wooden stand screwed with dry wall screws to the rectangular plates which  already exists on the Honda fork tubes, and which show in front of my right shoulder in Photo 1; the stand supports the forks as I eventually wanted them. I then removed the forks, bearings, tires, etc. and sawed off the level part of the round scooter frames parallel with caps on the front of the chassis I had just made, and welded them to the chassis arms. The round scooter down-tube is also an eighth inch thick, which makes for easy welding. Then I put the forks back in, cleaning and greasing the steering head bearings, removed my wooden stand and jumped merrily on the chassis to test it. Hooray! A rolling chassis.

When Photo 2 was taken, I was thinking of using round tubing to be the main frame, but a savvy neighbor talked me into the square section stuff. I never regretted having super strength in that area. You can also see I was experimenting with the handlebar position.


     I wanted to steer with handlebars using all the original Honda electrical controls, brakes, throttle, as well as the speedometer cluster. This is such a major simplification! So I welded a temporary steel box channel between the steering heads, and pivoted the old Honda handlebars in the middle. I welded flat steel 'steering plates' leading forward from the scooter’s forks right below the lower bearings, spacing them outward 23 degrees from straight ahead. These show well in photo 4. This would give correct "Ackerman" angles to the wheels when fully turned, the wheel on the inside of the turn needing more angle than the outer. 

     Another way to calculate this 23 degrees, is that the outer ball joint end of each radius rod, sighted straight through the lower steering bearing, should point exactly to the 'contact patch' the rear wheel makes with the road.  On your car, using a different tread and wheelbase length, it won't come out 23 degrees.

Later in construction, when I fine-tuned the passenger position, I removed the crossbar mentioned above, which was too obstructive, and used a post jutting out toward the driver from the curved forward frame member.  See Photo 7.  This maximized the ease of getting in and out.  The radius rods themselves are the limiting item for legroom.

Then, after welding in the crossmember,  and reassembling the forks, with upper and lower bearings well cleaned and greased, I created adjustable "radius rods" using 3/8" hardware store rod, which I threaded to match the spherical ball joints, called Heim fittings",  which I bought at the local auto parts store. ( Dorman 116-203, box of 5) I carefully drilled out the plates leading forward from the forks, using a 6" radius and 23 degrees outward spread and assembled the radius rod to two back-to-back Heim fittings on an arm from the handlebars. These fittings are mounted exactly one above the other in order not to change the toe-in length when the wheels are turned.  See Photo 7

To set the correct toe-in, I then lashed two sticks along the outside of each front tire and adjusted the rods until the separation of the sticks behind the tire was 1/8" more than in the front of the tire. Hooray! The wheels turned smoothly together!


      The beauty of this cycle-car, is that it uses so much of the wonderful engineering of the original Honda. I simply needed to reconnect the wiring harness, reattach the speedometer to the handlebars, then attach the horn, ignition switch, fuse box, and radiator to my temporary front cross member, put a battery box near the engine, and press the starter button. VROOM!

But I needed at least one brake for the road testing, at best a rear brake. So, from my local scooter repair man, I got a Honda Aero 80 rear brake cable which was long enough to go to a modified bicycle hand brake which i clamped between the left side handlebar electric cluster and the rubber hand grip. I knew I wanted left side to be the rear brake, and right to be front as on most mopeds. This allows you to blip the throttle while braking the rear wheel. Once I had a good rear brake functioning on the left side lever, I donned my warmest clothing (on Groundhog’s day here in Maine) and pushed the beast out in the weak winter sun. Three intense months of building had passed! See photo 3 for the original road test.

I had registered and insured the vehicle as a motor scooter, using the donor vehicle information,  so with new plates, I slowly circled my immediate block and gradually traveled 10 miles. The steering was far too twitchy, but otherwise, given the lack of weight, which the eventual body would provide, the car handled beautifully up to my personal limit of 40 MPH.

    It was amazing to be driving a vehicle you had created yourself.  There was little feeling of safety or creature comfort.  The wind chill was bracing.  But what a great boost to morale!  Now I could again engage in such a long-winded  and humbling project.

Back in the garage, I shortened the radius of the handlebar steering arm from 6" to 3" and tested the car again. This time the handling was steady and predictable and the car could still "U" turn in the width of a road. The handlebars moved a quarter circle each side of center. I now felt confident enough to begin on the body, so I removed all the stuff I had installed for the road test. You might be able to see in the picture that I was using conduit for the passenger foot support, held up by red hold-down straps. Not reccommended at 40MPH!

 V:     BUILDING A BEAUTIFUL BODY (well…nice enough)

      I had a few constraints for the body: It should be strong enough to roll over at 45 MPH with the occupants safe. It should be nice looking, aerodynamic, with lots of sun and light, zero maintenance, no leaks, easy entry, and be able to be left outside in all weathers. No "garage queen", please. This is to be a daily driver!

For the structural members, I bought four 20’ lengths of 1.5" EWR steel tubing, two in .083 and two in .065 thickness. Someone might say I was overbuilding the frames of this little car with such strong tube.  I don't care: the car weighs only 400 lbs.  I have no desire to have it weight less!  It accelerates well, but is quite subject to wind.  I have some confidence about surviving a roll-over.  Thus, given that the car is very small and light, I don't mind that it is super strong. 

    For mocking-up the body as a whole, I bought 8 ten foot lengths of half inch conduit, fifty conduit connectors, and a conduit bender. Conduit, which is easy to bend, is a designer’s dream, perhaps the fun-est part of the whole project. I decided to use 1/8" Lexan for the clear parts and 1/8" high-density polyethylene for the opaque parts. The brand I used is called "Versadur". I don’t know yet how it stands up to years of sunlight. See for the Versadur. Both types of panel, I thought, would be bent as necessary with hair driers, but in the end, I used 'stress bending', which means torturing them into place! See photo 4 for some sculpting .

Bending the heavier tubing was definitely a job for an automotive shop, so I had two "U" shapes bent which would form the front and rear roll bars.  For these and the  higher stress situations, I used the heavier of the two tubing thicknesses.  The front  "U" shape I clamped firmly in place to give the basic occupant space without welding it. I installed the large child car seat you see in the photo and, cramming myself into it, moved around to a position where there was good visibility, leg and headroom, and handlebar position. Then, with those human constraints, I began bending conduit.

I think you’ll be surprised to find that the central design problem is an unexpected one: an easy means of access which is leak-free and good looking. I won’t take you through the obsessive thinking, the blind alleys, the thrown away tubing, along the path to the body you see. But I will mention some of the initial helpful directions.

The rule is "start with what you know for sure", so I made the car's floor.  I used some of the heavy tubing to extend the frame forward to be an impact front bumper, giving the front wheels enough room to turn in, and room for a comfortable sitting position as far forward as I thought I might need. I then skinned this the new frame member  from below with 1/4" polypropylene sheeting to be a strong floor pan. The chassis was first primed with metal primer and painted with a white enamel. Then the floor pan was screwed from below with stainless sheet metal screws, drawing the floor pan up on a bead of silicone caulk.

With a strong floor now in place, and a movable large child’s car seat sitting on it, the design rule was clear: don’t weld anything. Don’t bend any big tubing. Stick with the 1/2" conduit until the body is complete with all the constraints met. And so I came again to the central design challenge: how to get in and out!

To make the opening leak-proof, a side entrance is perhaps best, or even better is lifting up a large section of the body as a whole using a counterbalanced hinge. Yet both designs limit the ease of getting into a very small car, which is best accomplished with simultaneous side and top removal. I wanted a car which would not be sidelined by the yogic contortions necessary to get through a side entrance nor by the muscle-straining lift of a large body section. I also wanted an opening which could be left open in summer and yet was not removed from the car and stored.     

     So I decided to take the time to design an easily-opened top-and-side opening which looked good when left open, backed up by a surefire rain gutter system to catch the inevitable leaks in the joints. It would take extra time to deal with the complexity of one panel passing over another, but in the end, what a joy: from the driver’s seat to hinge the canopy back easily  and talk to a curious bystander.

     As I mentioned, using two ten foot lengths of 1.5" thick-wall steel tubing, I had a local auto shop bend two identical arches. They must be identical because the overhead hatch will hinge over both of them. With the front arch welded to the frame, I then went ahead and had the curved dashboard member bent and then welded it between the steering heads, so I could proceed with fitting the front plastic panels. I added some 2" X 1/8" flat bar between the upper and lower frame members to form junctions, so the plastic panels could be fitted in smaller sections. This made bending easier.  See Photo 7

It’s hard to see the structure which evolved in front of the driver, but thinking of impact, I built things strong there.  The additional photos may show  the structure better.  For correct handling, the passenger in a three-wheeler has to sit well forward and thus is in a very vulnerable position, so  strength is important. Also weight in this area is good for getting that center of gravity within 35% of the wheelbase midpoint.

I heated my Versadur panels gently with a two hairdriers to help with the bending.  It helped a little, but not much.  Its a great material, milky white, and only $40 for a 4 X 8 sheet. I then fastened each panel to frame members with self-tapping screws and washers.  I later used flathead 6-32 screws tapped into the steel to give a flush exterior surface. Hooray, my beast was starting to look like a car!

Then I took a detour: If I completed the body now, I would have to contort myself to do all electrical hookup inside. So I installed the lights, turn signals, horn, fuse box, speedometer, brakes, etc. while I had such good access. For the front brakes, I simply added a second cable connection to the existing right hand lever.  See photo 8.   I found that the rear brake cable for a Honda Aero 80 is a nice long cable to use for each front wheel. It is still available from Honda.

I used 4 inch car double filament headlights which are squeezed into 4" rubber pipe connectors from my hardware store. They were then aimed at night and siliconed in place. On low-beam, they total the wattage of the Honda headlight. On high beam, the slowly run the battery down.  I eventually installed a 12 ampere-hour battery instead of the original 9 AH.  One could switch to LED taillights, turn signals, etc. if current draw is a problem.  My eventual entertainment system will also be humble.

For the rear brake, I ordered a type of bicycle brake lever which has a push-button to hold the brake on.   This is made by "Tech-77".  This gives you an emergency brake and well as the rear brake. This must be kept on, for the Honda’s starter to function. To give more power to the rear brake, I extended the lever arm back at the rear drum, so the tire would skid when the left hand  is gripped well. Since it’s a long run to the rear, I used heavy duty bicycle brake cable for this rear brake. Equalizing the front brakes was easy, using the existing adjusters at each front drum. In practice, braking is intuitive, with a right-handed person favoring the front brakes, which do most of the stopping. It might be worth it to buy fresh brake shoes, which are still available, if you check numerous dealers.

Once the brakes were functioning, I tackled the wiring. I bought a wire crimping tool and 100 butt connectors and set about extending the Honda’s wiring harness. I resisted all temptations to simplify the electrical system, and used the engineering of the scooter exactly. Why mess with success? The speedometer ran as usual off the right front wheel. I positioned it with good visibility, but not obstructing anything.

One of my design parameters was hand controls only, which is a wonderful system: Front brakes by the right hand: they do most of the braking; rear brake by left hand. Turn signals, horn, starter, dimmer switch, are all right there: no need to change anything. And with a twist-grip throttle, driving can be learned quickly. Since there are no foot pedals, with a small second passenger, both can position themselves easily.   I eventually used the engine kill switch work the windshield wiper.


Now I came to the toughest part, the central body panels. If I could do this well, the rear panels would be easy and the project finished! At this point, I had exhausted the $2000 budget, so I E-bayed the extra motorcycle parts, and thus had $500 in hand to finish the car. Labor was at 300 hours, so I hoped another month of part-time work would largely see me through. Ha! What a dream!

First I tested the car for 500 miles, still using the child’s car seat for testing, along with a seat belt and a helmet. I found that it would go 105 miles on an exactly-measured gallon in country driving, and 88 MPG in stop and go. The car would go 52 MPH, and but felt more comfortable at 40. I made further changes to increase the steering ratio to take more of the twitchiness out of the steering. I named it ‘Moonbeam’ and got lots of nice publicity in our small Maine town.

Then I bought a 4 X 8 sheet of 1/8" Lexan, and used part of it for a windshield. Stress bending panels in place is certainly the trickiest job of all. It takes many clamps. It’s best to have the panel completely cut to size and clamped in place before drilling and tapping for the 6-32 stainless flathead screws . I got very little help out of the hair driers to aid in the bending! With the Lexan, there’s the additional challenge of not scratching the panel. It is possible, however, to buff out slight scratches with Novus. I used lots of high quality white silicone caulk. Now for that overhead door!

I was tired of paying for bending the 1.5" round tubing, so for the door, I decided to use eighth-inch flat stock which I could bend myself. I used the front arch, which you remember was already welded in place, as a form to bend all three flatbar arches for the door.  The two  outside ones are 1.5" wide and the middle one is 1". See photo 6. (I don’t know who that bald guy is!)  The photo shows the form-fitting of the flatbar to the existing front arch, using a short section of conduit between the pieces to help bend.

On the lower edge of the car’s main chassis, I set up a temporary hinge point using a hardened bolt clamped to the chassis and set up all three of the flatbar arches on that hinge so that each would just clear the permanent front arch with a thickness of a body panel taped to the arch for spacing. Then I bent a center fore-and-aft overhead brace using the radius of the door’s swing, and used it to clamp all three door arches in a trial position.

To now position the rear main arch, I needed to find out the biggest door size, so that when the door was hinged open, it would leave the same size opening as that provided by the fixed arches. Once that was found, with the door in the open position, I placed the other identical structural arch under the front flatbar arch (now in it’s rear position), again adding the thickness of body paneling taped to the top of the permanent arch. Then I cut off and welded the rear permanent arch in place. As always, I double primed and painted the new structures with rust-o-leum metal primer, and then with gloss white paint.

I welded the center brace to the three equally spaced door arches, then cut out the overlapping metal at the hinge point, so they laid flat, and welded the three pieces of flat stock to come together as a single thickness, then drilled thru the main chassis for a permanent 1/4"  hardened hinge bolt, and: Hooray! my overhead door would hinge, somewhat wobbly without the panels to stabilize it, from open to closed, fitting snugly in each position.


     The pattern of the windows of a car, as much as the overall shape, create something pleasing to the eye or an ugly duckling. One of my basic parameters was a good looking machine, or at least good enough. So once the hinging roof was functioning, the next job was to figure out the rest of the car’s windows.

I didn’t want to make a ‘garage queen’ a car which has to be garaged. But to be out in the sun, with the roof closed, creates a problem: Located beneath angled windows, a car’s interior surfaces need to be dark , or they will reflect distracting images in the driver’s front or rear view. But a large lens-like window with a dark surface below it is a natural heat generator. Such a car will soon become an oven on a sunny day.

     So I needed enough window area to feel airy inside, and certainly enough for the good visibility necessary for defensive driving. But otherwise, I needed as much cooling white plastic panels as possible. A ‘bubble car’ though interesting in principle, is not a practical car under sunny skies. You need a good amount of opaque surface overhead.

Given this tension, the first step is to find out what is necessary for unobstructed vision. So I set up the rear view mirrors in their locations and began driving with the roof both open and closed, using masking tape to try out window patterns, keeping in mind that the car needed to look streamlined with both roof positions. See Photo 4.

What you see is what emerged: enough clear Lexan, but not too much. I then welded flat steel one inch wide on the sides, to be junctions of Lexan and body panels. I also chose the placement of these junctions where I knew the panels would resist bending in two directions at once. My efforts with a heat gun for bending, especially on the clear Lexan, had been failures. So I stuck with stress bending. See photo 5.

In retrospect, making that overhead door was the central challenge of the whole project, and though it does look clunky in the open position, it  gives easy access and visibility. It weighs about 30 pounds, and has a nice counterbalance spring below the hinge point. (zoom in on the home page photo)  Nevertheless, I would be strongly tempted to use the same concept next time, but to work in canvas instead!   In mine, there are hundreds of screws holding the panels on:  it strained my patience to the breaking point.

 VIII.   FINALLY FINISHING THE BODY  (rather than going swimming)

     Now the project began to bog down. Summer was here, I was driving the incomplete car daily, and finishing the rear body panels was like pulling teeth.   I first took Moonbeam to a motorcycle inspection station for a preliminary inspection. They insisted on the two vertical structural members you see behind the driver, which tie the scooter sub frame nicely into the arches. They also insisted on a windshield wiper and an arched central round tube connecting the arches. I was bummed by the additional work, but I also didn’t want to be crushed in a roll-over, so....

Since the rear body panels however are not structural, but only esthetic and aerodynamic, I could use eighth inch by three quarter inch flat bar which can be bent easily between the thumbs to sculpt the body. I moved the fuel tank 9 inches forward to give myself more design space. I also wanted storage space on each side of the engine for groceries, etc. Once the polyethylene panels were bent over these small flat bar frames, and screwed tight with 6-32 flathead screws tapped into the steel, I was amazed at how strong the panels became. No need for more structure. I then built a plywood box around the engine area, and soundproofed the inside of it with marine soundproofing. Driving was then much quieter, but still quite annoying. I  painted the engine box with 'Quietcar', a sound-deadening paint, and added some fuzzy carpet backing behind the driver's seat and below the back window.  Eventually, I got the car pretty quiet.

Seeing that wooden  engine box, by-standers ask if that is where I keep the gerbals powering the car!

The final side panels must be removeable as are the sides of the soundproofing box, for good engine access.  See photo 9.  The panels slide under 1.5" polyethylene strips on top and sides, and are attached only at the lower edge with 1/4-20 nylon pan head screws. Photo 7 shows me tapping these screws into nearby 3/4" flatbar framing.

Moonbeam was then quite close to done. I located that overhead arched 1.5" round tubing a little off center, so a single passenger wouldn’t hit it if the car overturned, and getting out of the car to the left was easier than a centered member would permit. If two passengers were aboard, this new beam is between their heads and well above them. I installed two seat belts and a simple double seat, covered with foam and naugahide.

You might want to make a more elaborate seat. I do lots of yoga and am used to spartan sitting arrangements.  Even though 90% of the use was to be a single passenger, I still wanted space and a seat belt for a second passenger.  The project seemed more sociable that way!  Still, I cautioned myself to be very careful about bumps and high-speed travel whenever a passenger were aboard.  After an hour of driving, your butt gets sore and your crossed legs go to sleep.  It's a short-distance car.

     I ordered two trailer fenders and attached them to the front scooter forks, filling in the sides with body panels to cover some of the suspension hardware. You may notice that the two front forks are slightly different, coming from 1987 and 1984 scooters. One has leading arms, and one trailing arms. Thankfully, this seems to make no difference in handling!

By  September  I was up to 1500 miles of use. Gas economy varied between 85 and 105 MPG depending on conditions. 105 is when driving a steady 40 MPH. 85 is for total stop-and-go town driving, both with one passenger. A reasonable claim might be 100 MPG. The car’s heater, which is actually the scooter’s radiator mounted in the front of the passenger space, was beginning to be much appreciated. In summer, I had to leave the overhead door open to keep cool.

Later, I may use the second scooter’s radiator somewhere outside the cab, in the airstream, but I can’t imagine where. Then plumbing valves would choose between the two radiators, according to the season. I still have further soundproofing to do, since the steel and plastic interior bounces the sound around mercilessly. I’ve added a latch mechanism, based on snap fasteners,  for the overhead door to be fixed in various positions. 

In many states, you need a motorcycle endorsement on your driver’s license to drive a vehicle like this. You may want to check your state. You have to take a day-long safety course, which is probably a good idea anyway for driving such a light car which, while far safer than a motorscooter, is still less safe than a larger, heavier car.

Alas, the car is a little homelier than I had hoped, although a mother always loves her child! I hope yours will be prettier. Stress-bending those panels in compound curves was such a chore.   I just didn’t have the patience to really go to the limit!    I'm sure a smoother body would come out of using fiberglass over urethane foam as in the website; but that too is very time consuming and in somewhat toxic working conditions.

Please keep in touch if you undertake your own project, and we can all make improvements together. Good Luck!

Jory Squibb, email:

P.S. This manual intentionally doesn’t have any limiting copywrite. See the page, "How you can use this information".  So please pass it on freely.  Isn’t it sad that such a simple vehicle, which could be mass produced for $5000 or so, which awaits no fancy fuel cells nor high-tech breakthroughs, isn’t available to the public? My thanks to you for helping to bring an obvious concept to light! As you watch gas prices move on to $4 and $5 a gallon, your long labors will seem worthwhile, and you’ll be glad to be rolling light.

PPS.  I am often struck at how impossible it must be to follow these directions. I can barely follow them myself!    You'll just have to hop on a bus and see the real thing here in Maine...