I started this project in 2013 while still in college and I've been tweaking the design ever since. It started out as a means to quickly get across campus and evolved into a never-ending desire to make the fastest, most rugged, and overkilled electric skateboard imaginable.
As it sits now the skateboard is about 6ft long with 10.5" pneumatic wheels, has 2 3HP electric motors, and a massive 350 cell lithium battery pack.
Making of the Custom Battery Pack
The battery pack consists of 350 Samsung ICR18650-22F 2200mAh cells in a 7s50p configuration. The pack is broken down into 5 smaller 7s10p packs, each with their own BMS rated for 20A. Upgrading the BMS for something with a higher current rating or building a custom BMS is something I would like to explore in the future. The current BMS was chosen because of the low cost and performance that should be sufficient for the current motors and controller configuration. The lithium cells were welder together using a car battery, solenoid, and Arduino. The Arduino script allowed control over how long the solenoid was active thus controlling how long the battery was shorted to perform the weld. The electrodes used to perform the welds was a sharpened piece of solid 10 gauge copper wire. All the welds seem to be holding together well with no failures since its initial construction.
After all the cells were welded together and balance / power leads were attached, the pack was wrapped in several layers of Kapton tape. This was done to prevent any accidental shorts in case debris is able to enter the pack.
The power capacity of the pack can be calculated as:
Vnom*AmpHours*Cells = 3.6V*2.2Ah*350 = 2772 Watt Hours
Because the cells used in the custom pack are purchased second hand, the actual capacity is going to be less than this ideal figure. Further testing is required to evaluate the actual capacity.
A custom enclosure was fabricated to enclose the pack and protect the cells. The longer sides of the enclosure was made from 1/4" 8020 aluminum plate while the other 4 sides were made from 1/8" aluminum. A measurement mistake while fabricating the enclosure caused the longer cuts to be several inches shorted than intended. I was able to get around this by welding additional material to the miss-cuts and re-machining the the pieces to the correct size. Rubber grommets where then added to all the holes used for the balance and power leads to prevent the edges of the holes from cutting into the wire jackets. Foam rubber was attached to the inner walls of the enclosure as well to all for tolerance in the construction of the enclosure to still enable a tight fit of the batteries.
Chassis and Suspension
The chassis of the skateboard was constructed using 1.5"x1.5" 8020 extrusions. The original design was done without any welds. All joints were connected only with fasteners. This worked well but was not as sturdy as I would have liked. There was some flex issues as slip concerns in some of the joints. Since then, I've acquired an AC tig welder (AlphaTIG 200X). This has allowed me to weld together all the joints of the chassis and results in a much more sturdy construction.
The battery is secured in the chassis with the use of tabs on the bottom and blocks toward the front. The top on the board the rider stands on compresses the battery into the tabs and the black secure the battery from sliding length-wise in the chassis.
Springs are used on the front and rear truck assemblies to stabilize the board so that the resting position allows the board to drive straight. Finding a good balance of spring stiffness to allow for ease of turning while also preventing speed wobbles has proven to be difficult.
The Electronics
In a rush to get the board running again, there was a mistake made in the wiring of the board that has since been corrected but is worth mentioning. As seen below, the ground wires from the battery pack are connected to the solenoid. This means that when power is cut, the ground connection was broken instead of the positive rail. During some early testing, the fuses on the distribution block blew while going at a high speed. This caused the solenoid to open and several components to break. I'm assuming this is because the ground path was interrupted and cause a voltage spike on the electronics side from the back EMF of the motors.
During that incident described above the voltage monitor and motor-controller both broke. The voltage monitor was completely destroyed but I was able to repair the motor controller. The image below show the switching regulator in the motor-controller that failed. The switching regulator used is a LT1171 and regulates the rails voltage down to 5V. This was the first time I've seen a failure like this. The case of the regulator was broke off from the package as seen in the right half of the chip. After removing the regulator, reversing engineering the motor-controller, and verifying that it was still operational, I replaced the LT1171 with a new chip and the controller was working again. I was pretty lucky this was the only thing that broke and the repair was a simple as replacing the chip.
Since the failure described above, the wiring has changed slightly. The output from the fused distribution block now goes to the power solenoid and all the ground wires are commoned together on a un-switched post. Power to the controller is switched via a safety kill-switch and key-switch. The key-switch engages power to the kill-switch and the kill-switch engages the solenoid only when the lanyard is attached. This configuration kills power to the motor-controller when the rider is separated from the board and thus the kill-switch is disengaged.
Future of the project:
Some of the next features I would like to add include a GPS to track distance, a more advanced BMS to monitor power usage and cell balancing, and headlights / brake lights for safer night riding.
Comments