GX “Green Experiment” – Hybrid Kart

[snowball]

In 2014 I built a small vehicle for a school project. It was powered by 2 robotics (CIM) motors and a custom made water-jet gearbox. It used an 18Ah 12V SLA battery and produced about 1hp. It also used a joystick to control the forwards, backwards, tank-steering movements. It was a lot of fun but I never really had reason to keep it around; plus it was so slow.

In 2015, I decided to build a go-kart using a scrap frame I had purchased from school. Making a 4-stroke go-kart seemed pretty cool but I was never really motivated to put in all the work. I do love cars, driving, racing and all that, but I was hoping I could make something more exciting and worthwhile. I decided that I would scrap the robot vehicle (which had many problems of its own, especially with the gearbox) and add the robotics components to my gas powered kart for a little torque boost. My inspiration included the 2014 F1 cars, McLaren P1, La Ferrari, and the Porsche 918.

This project has taken almost 3 years and is still not entirely complete. I may submit it for the 2017 build-off but at the moment I will just make a journal for it as a lot of work has already been done.

I am aware it would be faster if I used a bigger engine. It would be more economical to have no electric component. It would accelerate better with a torque converter. But these are not the reasons I am building this. I find all of the hybrid-technology out there these days super cool and I want to see if I could develop my own hybrid-concept. So this is the story of the Green eXperiment.

Here are some pictures of the original electric robotic vehicle:

10/07/2014
I sourced out a frame from a local technology institution, it was $120 and includes the frame, four 6” wheels and tires (two tubes need replacing) it has a nice steering wheel and shaft already installed. Someone else has already installed some custom spindle brackets and spindles. I’ll be fabricating everything else from scratch. This was a good deal.

12/14/2015
With the frame sitting and gathering dust for over a year I’ve decided to build the vehicle powered by a typical 4-stroke engine. I want electrical motors to assist in acceleration between 0-7kph and then we want the motors to turn off at speeds higher.

To avoid killing the CIM motors I should program a ball shifter transmission to shift out of gear after 7kph before turning off the motors. The first gear would be 20:1? 30:1? to the wheels and then the second gear would be 5:1 or 4:1 engaged after the vehicle is already in motion. Drive wheels are 11.5" diameter, if geared at 5:1 we’re looking at ~35-40kph.

If both electric and gas motor are engaged, the CIM motors will get toasted by the much stronger gas motor (unintentionally spinning at 12000RPM. The CIMs freely spin @ 5000RPM. Another issue that might occur is that the CIM motor will slow the vehicle down by providing resistance as it is geared much higher (unintentional alternator)

I’ve created a document with the outlined inputs and outputs I’m trying to configure. This is only for the electric portion.

I'm looking at the 3 CIM ball shifter on VEXpro, we have a few hundred dollars to play with.

Computer
Arduino

Inputs
Pedal with potentiometer
Speed sensor

Output
Gas motor position
CIM motors (victor motor controllers)
Ball shifter transmission 2-3 CIM motors

Gas motor position
No throttle
Full throttle

Transmission position
P1 = acceleration mode (gear reduction)
P2 = coasting mode (free gear, disconnected from CIM motors)

CIM motor mode
Throttle control (P1,<7 km/h )
Disconnect (P2, >7 km/h)

Control
INPUT Potentiometer = ON (voltage TBD)
AND Speed Sensor = < 7 kph
OUTPUT gas power = FULL
OUTPUT cim motors = FULL
OUTPUT trans position 1 (acceleration mode)

INPUT Potentiometer = ON
AND Speed Sensor = > 7 kph
OUTPUT gas power = FULL
OUTPUT cim motors = FULL
OUTPUT trans position 2 (coast mode)

INPUT Potentiometer = ON
AND Speed Sensor = > 8 kph
OUTPUT gas power = FULL
OUTPUT cim motors = OFF
OUTPUT trans position 2 (coast mode)

INPUT Potentiometer = OFF
OUTPUT gas power = OFF
OUTPUT cim motors = OFF
OUTPUT trans position 1 (acceleration mode)

1/26/2016
Pulled apart all the robotics components from the previous robot build. This build used an Arduino, 2 motor controllers, 2 CIM motors, 2 40A breakers, a PlayStation joystick, 12V 18 Ah SLA battery, 120A breaker.

The robot with one person on board was capable of 10kph theoretical top speed… I think we really only hit 8kph in testing though. It was really peppy with all that torque, the weight of the vehicle was only ~240lb with the driver. The new kart was 325lb with the driver.

2/2/2016
I added a Victor SP motor controller (already have 2 Victor 888s) to complete the 3 CIM gearbox, with the extra CIM the kart should produce ~1.5hp

For some reason the new controller keeps spinning the motor at idle and will not stop. The old Victor 888 does not spin under the same code. I have moved cables around between the two models but the Victor SP still keeps spinning during idle.

FIXED 04/2016: Finally fixed this issue after 2 months, the new Victor SP uses pulse lengths of 0.5ms to 2.4ms. I added this line to the code for the Victor SP to remap the output to motor:

servo.attach(pin, 1000,2000);​

I’ve also decided to stop being lazy and officially started using proper wire connectors instead of just soldering stuff together. This actually turned out to be easier.

More to come...

 

[chancer]

Your in Kali fornia aren't you? Nice Prius dude!

 

[Denny]

Denny

 

[ultralight01]

If I ever build an electric vehicle it will be 48v 1000w. I've thought of using a moped frame with a high power motor and a ton of SLA brick batteries wired in series and parallel with a balance charger. At 48v 1000w any 2-3 wheeled vehicle is street legal with no registration, insurance, or license here. A friend of mine is planning on building a high power electric drift trike with lithium batteries and about 40 miles of range. It would count as street legal.

Look into building 18650 cell packs. They make great batteries, and you can spot weld them with a homemade welder made from a car battery.

 

[snowball]

1/20/2017

Since the beginning of the project I have tried to figure out different ways to have the electric motors and the gas motor power the same axle without fighting one another.

The gas motor will be geared at 5:1 with a theoretical top speed of 39.6 kph
The electric motors will be geared 24:1 (5.4:1 in gearbox, 4.5:1 in chain) with a theoretical top speed of 14 kph

Originally the idea was to use a ball-shifter that would have 2 gears; A low-gear electric, high-gear gas. However, there would be a lag between the shifts because as centrifugal clutch needs time to catch. Also it would be too complicated to have two different input shafts generating two different types of energy.

The new plan was to have both the gas and electric engage at the same time through two different chains to the live axle. This way the electric and gas work together during acceleration.

The electric will kick the vehicle up to ~14kph while the gas clutch is engaging. This required adding freewheel to the robot gearbox output shaft so the robot motors will not slow down the gas engine.

I purchased a ½” hex shaft and I put aside the original aluminum one that came with the robotics gearbox. I took it to the lathe and turned it down a bit so it would fit the one-way clutch sprocket I found on ebay. I milled a slot for the keyway, it is so incredibly small, but I managed to make it work. It’s ugly but it will do. I had to “custom” make this shaft about 5 times to make everything fit properly. One day I will make a clean version of the key, I don’t want to commit to making a perfect part until I can prove this kart-concept will even work.

The inside was the easy part, getting the 16T sprocket to fit the outside of the bearing was difficult. I made an adapter to accept the 16T sprocket. Oddly, the outside of the clutch bearing had only a tiny slot. I do not know how I would be able to key this part in. I ended up welding the bearing directly to the sprocket, blew it off with compressed air after each tack weld to ensure the lubrication did not catch on fire (some of it still did).

I threaded the end of the shaft and put a nylon lock-nut on. I wanted to try to on the lathe, ended up doing it by hand, turns out it wasn't that difficult, just needed lots of cutting oil.

NOTE 04/2017: There are companies that make one-way clutch sprockets that would have been perfect for this concept. At the time I started working on the kart, I did not know this. At the moment, my custom free-wheeling sprocket is holding up just fine after extensive testing.

02/12/2017
I built a one-off pedal assembly to send signals to the Arduino and possible the gas as well depending on what I want to implement.

The original kart used a joystick controller for one motor on each side. This was necessary to enable tank-style steering. Now that I’m running 3-CIM motors onto one shaft, I don’t need the joystick controller. The pedal assembly uses a potentiometer to send a signal to the Arduino. The Arduino accepts the analog input and translates it into a digital motor output signal through the Victor 888 and Victor SP.

02/28/2017

Added a seat cut from an old office chair. If you’re wondering, office chairs make great makeshift go-kart seats!

Also, I built a left “wing” out of steel and polycarbonate for the electronic controllers.

Arduino with custom motor shield
12V 18aH SLA battery
120A breaker
2 x Power Distribution panels (pos and neg)
3 x 40A snap action breakers
2 x Victor 888 Motor controllers (with fans)
1 x Victor SP Motor controller
Lots of wires heading to the CIM motors attached behind the seat

 

[snowball]

06/2017

I had the kart running today under electric mode, tested each motor individually to make sure it was spinning the right way, good to go.

Then when it came to driving I tried both slow acceleration and hard launch, the motors random stutter under acceleration (not only hard accel, but also when adding power during cruising) and it causes the chain to pop off because the gearbox moves up and down due to the torque.

Otherwise I was able to get it up to 14kmh with gentle acceleration in a straight line. The one way sprocket adaptor works well.

Launching: the vehicle launches for half a second before the chain comes flying off violently. =(

To do:
1) drill more holes to move the gearbox so there's less tension on chain
2) add a fixed chain tensioner on slack side to help the chain fall back into place when it wants to "skip" because of the misalignment
3) somehow add more support for the gearbox so it doesn't move at all when it stutters

PROGRESS 07/2017

1) I actually milled a bunch of slots so I can slide the gearbox back and forth for tension. The free-wheel sprocket works better now after welding to a one-way bearing.
2) I made a fixed-chain tensioner which allows the chain to stay on track on the slack side as it likes to jump during initial acceleration (just so much torque to be moving the 170lb kart + 160 lb driver) There have been no issues with the chain jumping when the axle is being driven by the gas motor.
3) Added a ¼” aluminum plate under the gearbox to give it more leverage. It is much more ridged now and the gearbox doesn’t move.

08/2017

After two years of work and a lot of trial and error almost everything works. I made a quick video to demonstrate the torque-fill action of the vehicle. The only thing I don't have working is the hall-effect motor switch off. I plan to fix this some time this year.

Cole's notes:
3.5 hp Briggs
1000W combined 12V electric motors
160 lb driver + 170 lb vehicle
Top speed = 40 km/h (25mph)
0-100' = 6.5 seconds no electric
0-100' = 5.5 seconds with electric

https://www.youtube.com/watch?v=lH8cexIHdNw&feature=youtu.be

 

[snowball]

This project was just to show the torque-fill benefits of electric combined with gas

but don't worry! We're currently working on a light weight aluminum concept running off 48V haha

 

[landuse]

Nicely done!

 

[garolittle]

Just saw this. Very nice!