[8.5.14]  Build of the Chibi-Atomic-Jeep


In the weeks before the Detroit Makerfaire, myself Ben, Mike, Ciaran and Rob got the chance to putz around in the just-about-complete chibi-miku-van. It was glorious. At the last moment we decided to build our own entry for the Detroit makerfaire, using whatever parts were lying around miters.

What ensued was chaos driven gloriousness

Sensored Alternator
Frame and Steering  Helical Gears
Wheels and Brakes
Conclusion Image Directory

The chibi-atomic-jeep, (red jeep-like contraption shown right) completed in less than a week and a half, raced and completed the endurance portion of the Detroit makerfaire 2014 race. We burned through NINE 8" pink harbor freight tires  and TWO 10" rubber tires  between the time-trials and endurance race. The cart faired surprisingly well, caused plenty of chaos on the track and was really well received. Below details the build log and competition 

It all started with an alternator
A scrapped alternator from Charles g's  mikuvan became the center of attention. After removing the rectification control block from the back of the alternator, exposing the three-phase stator windings and routing the spring-loaded brush assembly for the rotor, we were able to spool it up from a simple sensorless motor controller. Ben K donated a very nice, sensored, current controlled kelly controller to the build, so the next move was to push for adding a hardware position sensor to the alternator.

Given that the alternator only has an output on the pulley side, and that there is only a relatively short, stumpy drive, there is little room for a sensor mount. A hole on the opposing side was bored out, leaving the casing lip in place so the alternator bearing would remain in place. This is where the awesome happened. Did you know a cd rom drive has a 12 pole motor (and an associated 12 pole magnetic rotor housing? Ben magically did. after gluing down three hall-effect sensors and designing a really fancy support jig (including flexures) the sensored alternator was born.
Look at this beast.

Ben's design featured a movable sensor mount, to compensate for starting rotor position. Shown (right) is the sensored alternator clamped to the labbench.
The three short feeds were tied to the three phase stator, the speaker-wire 2 conductor cable was tied to the rotor field and the kelly controller was tied directly to the sensor board for the first flight of the sensored alternator

Sensored Alternator Spoolup
Thats quite a bit of KV / RPM.

While we didnt get a chance to measure the kv/rpm at different field control currents, it varied substantially. For a field current we used everything from 5A (saturated) to 0A (induction drive). What would have been interesting is plotting a 3d torque vs rpm vs field current curve, unfortunately that heavily requires a dyno.

Frame & Steering
The frame started as a tack-welded together pile of 1/16th wall 1" square steel tube, measured and modeled using the finest in cad software, chalk drawn around the sawzalled powerwheels frame.
After rough tack-welding the frame was tig-welded together at the ever-awesome DLAB welding shop. The start of the frame was suprisingly lightweight at this stage, and we opted to hold it up with the only 3 stools in miters that were the same height.
Pillow blocks
The first thing that was tied to the frame were the two 5/8 ball-bearing pillow blocks. The frame wasnt super-symetrical as there was some incorrect polarity welding wich punched a hole through one of the mating surfaces, so the bearing-shaft assembly lined up with the rear part of the kart. The hardened shaft is shown shining brilliantly.
The steering of the beast begins

A comrade had gifted some 1/8" thick steel plate, way back in battlebot season, during the atomic-bumble-prime build. Having access to a 30 minute block of time on the hobbyshop waterjet, I cut out some haistly made steering bearing holsters. 
Rough placing of the steering points

Using steel plate as shims, the steering bearing blocks were positoned and migged in place. A slaughter-fest of mig-fueled gloriousness occured, with the bearings in place, clamps keeping everything tight and an allignment bolt holding everything true.
Welding ahoy

Yeah that was a lot of weld. As these were flanged bearings, with the center barstock on the center preventing lateral motion of the bearings. 
Hexagonal steel barstock to the rescue

A scrap hexagonal barstock was used as the nating between the steering pivot point and the soon to be steering axle. This worked excellently as, later-on, boring the press-fit mount for the steering axle became easy, as the hexagonal rod fit nicely in the lathes' three jaw chuck.
Toying with steering geometery
There was some back and forth (PUN INTENDED) for the placement and angle of the steering.
Steering geometry

There was a narrow mishap here, as the kart almost had reveresed steering, fortunatley, with the a-arms tigged in place and the ball-mounted steering linkages, everything lined up. OH! we also inherited a seat from Shane colton's Cap-Kart. THANKS SHANE!
Steering rod and upright

After some rough placement the steering column and seat, we tack welded everythig in place and began setting up the steering support hardware.
Pressing in a bearing and more Hex stock

With the position of the steering column mostly finalized, a hex-stock holster and bearing were added to the upright steering arm. As there's a significant potential for torsion here, copious amounts of weld were added.
The base of the steering
I thought this was brilliant, its a happy scrap block of delrin acting as a friction fit for the base of the steering column. Its constrained above by a shaft collar, and tacked in place with some angle iron.
I purchased a hardened 5/8 shaft for the rear axle, without a built-in keyway. Adding a keyway to a hardened shaft is, at best, time consuming. After burning through a 3/16th carbide bit, Nick Kirkby saved the day with a 1/4" carbide endmill. After a couple dozen passes, a 1/8" deep keyway was born.
Keyway was required for both tires, and the working end was clamped in the mill-vice. I ended up adding a small fan to blow away coolant vapor and sprayed compressed aire in to keep the tiny chips from being re-ground.
A most excellent disc brake

Ben fabricated the most excellent of disc brake mounts, which fit snug-ly on the newly key'd shaft. We eventiually opted to 'flip' the pillow block bearing mount to allow for increased ground clearance. More on the wheels and brake mount below.
Seat was added, steering was completed and welding was had by all

and like that it was becoming a kart, bowden-brake cable in place, steering assmebly solid. The seat grew tack-welded braces. All we needed were some gears.
A quick battery box was born

Polycarb and a heatgun did wonders. It was snug enough for the initial 10S2P battery packs that it almost didnt require an external constraint.

Mighty Helical Gears Image / Media Image / Media
The hobbyshop waterjet saved the day here
Each of the helical gears was made from a 1/2" thick steel round. The larger of which were roughly 5" in diameter and the smaller being 3" in diameter before machining. the tiny cuts were actually for a potential hexagonal alternator-drive coupling.
Gear planning
We opted for a roughly 1:10 reduction, Rob went to work doing gear maths and consulting gear manuals from the early 70s. After a bit of back and forth, helical gear plans were set and all sorts of attachments for the bridgeport were bolted in place.
A rob reeve approacheth

Preperation of gear making, using the miters-bridgeport, chinese gearcutters an indexing head and, variable gearbox contraption.
A closeup of the gear making contraption
At this point most of miters is staring at the setup in awe

Turning the whole setup required an extra lever arm on the bridge port axis, this is shown (below) in the video of the gear hobbing.
The gear cutter cuts, starting from the sacrifical portion of the spindle, through the new gearstock and into the sacrifical nut
Rob makes noises and cranks things for an hour or so

and then gears appear
look at these things. Video below shows the process for a smaller helical gear.

Rob Reeve 'reeving' a gear Image / Media
The following is three seperate shots of the making of one of the smaller helical gears. It was a sight to behold, full of manual cranking with a huge wrench.

After the gears are complete, they are haistly welded to 5/8 shaft, and mated to an extremely sexy, haistly waterjet housing. Ben worked his magic, which is why the thing came out frigging gorgeous. Unfortunatley there werent many pictures of the fabrication as it happened day-before heading to detroit.
The side plates of the gearbox were haistly waterjet from 1/4" thick 6061 aluminum plate. Spacing for the actual gears was very critical, the holes are for the 1.25" OD , 5/8" ID ball bearings. The bearings we used, link, were actually fairly horrible, get proper abec bearings people

Wheels and Brakes
There was some concern that the harbor-freight rims for the rear tires would shear under load, so inserts and bolt mounts were made to suport the tire. This started with a short round aluminum stock, cut in half on a horizontal bandsaw.
Next the wheel braces were faced on a lathe and center tapped. A large hole was required to fit around the existing harbor freight bearing assembly. The bolt plate and aluminum spacer are designed to prevent any axial loading on the wheel bearing, it mostly serves to keep the wheel centered about the assembly.
A large hole is bored to fit the harbor freight tire bearing assembly and the edge of the brace was beveled to fit against the surface of the wheel.
Finally the mill was used to bore four holes to match up with the wheels split-bolt pattern as well as the welded hub.

The same process was used to make the a nearly identical part for the opposing wheel.
welded lug plate + aluminum brace + wheel

Once the plate and brace assembly was finished a large keyed stock was welded on, this transfered torque from the motor + gearbox into the rear shaft and out onto the wheels.

Things shift about thermally when welding, as a result the spacer needed a bit of extra boring out to ensure it fit well.
With the welded, keyed part of the setup in place, this is just about what the final brake + wheel holster looks like.

The shell Image / Media Image / Media
A tiny powerwheels jeep wrangler frame was gifted to us from the EVT team, it seemed like the tiniest of the potential powerwheels frames, but ended up working out really well. The rear tire assembly is removed. We decided to take very little from the stock frame, as the driver seat would occupy a manjority of the rear area. With a rough size of the frame, we proceed to tear it apart.
Matching up the frame and chassis
As a rough estimate the frame was lined up to the available shell. Chalk was used to estimate spacing.
Cutting and rough spacing continues

A sawzall and a japanese saw made easy work of hollowing out the shell. Shown right is the rough placement, it was a bit... short
ignore the shell, build everything else.
So we went on and built the kart, steering seating, battery holster, and then placed the shell on, to find out, we needed to stretch it a bit. We contemplating adding suspenders to the shell and wearing it as a skirt. This was vetoed by a small margin. 
The shell is quickly cut in half (to add room for the doors) and test-fit with seat. THE KART FITS HUMANS. Right after this, red plastic-bonding paint was added and a bondo'd side-door was quickly thrown together thanks to Ciaran.

Moving on from alternators
After bench-testing everything related to the alternator-drive, we transfered over to battery power. unfortunatley, the Dc/dc converter's caps couldnt take the voltage transient. This was approaching the deadline and the midnight hour, we switched from sensored alternator to a backup-motor. We scrambled and mounted on a brushless 3phase 180kv/rpm motor.
This is the same motor used in the submersible thruster project 
Retrofitting, tweaking and the sunrise happend.
THE FIRST TEST, morning of the trip to Detroit.

The First-ish launch  Image / Media
Morning of the trip to detroit we fired up the correct gearing and the backup motor, on a 10S (30v) battery pack. This and two other runs were the only pre-race vehicle testing that were completed.


Transporting and Racing
Using an MIT discount, we rented a 5 seater and gently strapped the atomic chibi jeep to the roof atop some happy pink foam. The shell was removed for transport as it was feared to fly off sporadically. We piled in a bunch of other gear, a mig welder, backpacks power supplies, spare batteries, chargers, laptops and a pile of extra harbor freight pink tires. We departed at 10:30AM, took a run to a harbor freight, acquired extra batteries and pushed westward.
Wubbed Beef

Around 8 hours later, 'wait people need food right' happened and we stopped off at Alex's place, somewhere near buffalo ny. We were tired from driving and integrated sleep loss from the previous week.For whatever reason the carniverous food there was delicious. We hypostulated why it was so good and determined it was because vacuum chambers. Hypostulating continued for an hour, as we were waiting for the check, only to find out it was there the whole time. Also there was some hot-rolled steel.
We arrived in dearborn at~3 or 4 AM

And promptly fell asleep. Time trials began 5 hours later, after some dazed and confused IHOP, somewhere near detroit, the troops were ready for action. Charles G and comrades had already mostly setup, and we joined them in a very cozy tent.
Chaos and planning
the chibi atomic jeep side by side with the chibi mikuvan. Months of planning vs 3 hours of testing before racetime. 
The couse and the competitior.

The course was huge in comparison to the NY makerfaire of the previous year. They also had a neat tire + rope setup which kept the course together.
The  beast
armed with a gopro, googly eyes and a 3d printed doge. Chibi-jeep was not taking any prisoners. 
Pre-race prep
The kart was filled with batteires, compressed air and eventiually a driver
Ciaran kicking ass
Atomic, Mikuvan and Nimby all in one shot

This is my favorite shot, for nearly a whole lap, atomic-chibi held in front of both nimby and mikuvan, thanks again to magical Ciaran driving skills.
Rounding corners, attempting not to detroy the tires.


We seemed to burn through tires feverently, we swapped in 2x 10" rear tires for increased speed, only to have them shredded within 4 laps.
rage-charging battery packs (note, same battery pack as the ice-scooter)

Hyperion 1420I shown in the background, pumping 500W into the recently discharged pack.
After-race putzing around
an Audrey and becca enjoying the glorious chibi-atomic-jeep

Pit Crew Action Image / Media
Tire changing and battery swapping were hyper-quick.

The scooter packs (12S4P, 26650) with 50A anderson connectors proved to be super-useful, as they could be swapped in / out within 10 seconds. Tires were quickly swapped by leaving the previous rim in place and re-inserting a new tire / tube as they fell apart.

Endurance Race Image / Media


Loosing a Tire Image / Media
We shredded the tire-bearing assembly on a harbor freight rim.

Fortunatley this was caught on film.

Bulletproof Image / Media

while racing you dont notice pileups behind you, but the atomic jeep can be seen

  • 4 seconds in: [tire change]
  • 24 seconds in [making a hard turn]
  • 26 seconds in [Running over someone]
  • 1min09 seconds in [after flipping someone]
  • 1min29 seconds in [taking abuse and pummeling through]

Bill of materials Price Ea Price Total
Pillow blocks 11.18 22.36
Fuse Holder 8.95 8.95
Hardened rear axle 19.92 19.92
Gearbox Bearings 8.53 51.18
5/8 Shaft Collars 1.97 7.88
1/4" Keyway stock (1') 3.7 3.7
Steering rod-ends 3.26 6.52
1" square steel tube - -
Pink 8" Harbor freight tires 7.29 72.90
Kelly KBS 48V sensored 3phase controller 149.00 149.00
Blackmellon 3phase motor - -
Super swanky seat - -
12S4P battery module ~240 240





(There's other photos in the photo gallery)
Concluding Remarks:

Overall things went well, we had a routine issue with the kelly controller resetting during racing, however this issue cleared itself when we upgraded to 'fullpower' after racing. Still undetermined issue at the moment. The bearings in the gearbox were sub-optimal and needed replacement with something ABEC rated, the tires is an area of improvement, but quick-changing was good practice. Gearbox noise should dissapate after the new-bearing swapover.

If you have questions or comments, ask below or send over an email.
HTML Comment Box is loading comments...

(be careful, im not responsible for your incredible acceleration)

Rensselaer Polytechnic Institute 
Electrical & Electrical Power