Prototype Version 5 Wins Gold at Regional Science Fair!

Following the end of the construction of the fifth prototype electric skateboard, I decided to enter my project into the Greater Vancouver Regional Science Fair. The experience was absolutely astonishing! Not only did I meet hard-working, dedicated students and their innovative projects, but I also met experts in the field of engineering, who offered creative criticism. And all of this couldn’t be possible without the committed judges and voulenteers, who took time out of their daily lives to set up and monitor this event. Way to go!

In order to make my project more scientific, I decided to run an experimental trial to see how the gear ratio effects the power consumption on the fourth prototype of the electric skateboard. I did this with several gear ratios and weights, and I used the results to improve the skateboard, by choosing the best gear ratio for my weight. I also determined the speed at which the skateboard has the maximum range, which turned out to be 19km at 9km/h. This is just an estimate; I still need to test that.

Bluetooth Logging System:


I also decided to run an experimental trial on the maximum current and power consumption under different loads. Both of these trials were recorded with a custom-made, Bluetooth Logging Module that I designed, programmed and built. It uses a GPS to track location and velocity, a hall effect current sensor (for current), and the standard voltage divider/reference diode for voltage detection. More on the module in a further post.

Introducing Prototype Version 5:


Prototype version 5 is the best, most complex prototype I have built so far. I had to learn how to work with carbon-fiber for this board, as I wanted to reduce the weight of the board even further. The deck has a foam-core, and I used wooden supports around the trucks. There were a couple of differences with the gear drive as well; I 3D printed two PLA spacers so that they could more easily hold the gear in place. Additionally, I used a 25 tooth gear that I modified. It used the least amount of power out of all the gear ratios tested.

The electronics side was a bit different as well. The largest change was the addition of a lithium-ion BMS board. I found one on eBay for about $15 for the 6 cell configuration I was using. The major advantage: safe and fast balance charging. I can charge from dead to full in under 60 minutes! I also picked up an inexpensive 24v power supply, which I modified to become a constant current/constant voltage source through the “33R Mod”. More on that in a later post.


These were some of the major changes that I changed from the fourth prototype skateboard. Currently, I am working on alternative methods of controlling the board; everything from Wii-Nunchucks to Bluetooth Low Energy and Gamepad Controllers.

The development of this board, as well as the write-up and experiments that I did helped me win gold at the Greater Vancouver Regional Science Fair! I also recieved an award from the Canadian Institute of Energy, and I will be travelling to Montreal for the nationals in mid-May!

Read more


Bluetooth Skateboard Controller Update!

I have successfully produced 4 prototype electric skateboard controllers. Here are the specs:

  • Modes: Car/Boat and Electric Skateboard
  • Sensitivity: 1000 degrees
  • Input Voltage: 4.5~6.5 VDC
  • Max Voltage Measurement: 55 volts
  • Failsafe: Slowly turns off the motor
  • Dimensions: 50.0 x 38 x 12mm

These controllers are almost ready to ship. All that’s left is to waterproof each circuit using epoxy resin, and ensure compatibility across all Android devices.

Read more

Weather Station: Solar Power System


One of the goals for the weather station project is to make it solar powered. This, combined with a medium range transmitter (100 m), should make it environmentally friendly as well.

Choosing a battery and solar panel:

To figure out what type of battery/solar panel I need, I started by researching the climate of Vancouver. I also needed the amp draw of the circuit, which I calculated to be about 10mA.


Now, we get an average of 5.2 hours of sun per day, per year. However, during December we get only an average of 1.82 hours of sun per day. We want this weather station to survive the winter unaided, so we need to adjust the calculations for December.

To figure out the recommended current of the solar panel (in mA) we need to go (10mA * 24 h) / 1.82 h. We get about 133mA.


The solar panel I went with is a 9v, 350mA, 3.2 watt solar panel. It’s a bit overkill, but like all solar panels, it can only work in sunlight. Cloudy weather will not do.

On average we go about 2 weeks without a purely sunny day, so I used that to figure out the capacity of the battery. 10 mA * 336 hours = 3360 mAh


To provide power to the circuit on cloudy days and at night, I got a 6v, 5Ah sealed lead-acid battery. I’m curious to see how it performs in the wintertime.

Charging the battery:


It was a huge struggle to automate the charging of this battery. The current circuit I am using is similar to this one from I modified it so it can be controlled with the Arduino chip. Basically, this is how it works:

  • The Arduino checks the voltage of a 2-resistor voltmeter connected to the terminals of the battery. I used a 47k and a 10k resistor for this.
  • If the voltage is more than a fully charged battery, the Arduino brings the pin of the transistor low. This activates a rudimentary float charger, which limits the charging current to 20mA.
  • If the voltage is less than a predefined point, the Arduino brings the base of the transistor to high. This bumps the voltage to ~7.2v, and the LM317 charges the battery at full speed.

The actual current output is internally limited by the LM317, which limits current the hotter it gets. Since I was building the circuit and constantly testing it, the battery would often overcharge.


Lowering power consumption:

Initially, the weather station circuit drew ~30mA when not doing anything, I aimed to get this down to ~10mA:

I started by ordering low quiescent current regulators, specifically 5v and 3.3v regulators. Back when I constructed the circuit I used two resistors to provide the 3.3v needed for the barometer to run, I even calculated the power consumption to be 0.083 watts, or 16mA at 5v! What was I thinking? Replacing this voltage spliter halved my power consumption.


In my Arduino code I implemented the sleep function of the Low Power Library found at Rocket Scream. Since the max sleep time is 8 seconds (without external wake-up), this little loop makes the Arduino sleep for 10 minutes.

void sleep() {
for (int i = 0; i < 75; i++) {
LowPower.powerDown(SLEEP_8S, ADC_OFF, BOD_OFF);

Testing the circuit:


I think now is the best time to test the circuit, during the dark-dark days of October. Here is the entire assembly on two perfboards, with the solar panel facing West. Notice the plant watering prototype on the right!

The red multi-meter shows the amp use. A positive number indicates the battery is being discharged; a negative number indicates the battery is getting charged. Again, the solar panel only charges the battery when it’s in direct sunlight.

The voltmeter which reads 6.4v is powered separately from the circuit and doesn’t increase power consumption. My plan is to leave this setup for several weeks and in the meanwhile, assemble things like the new transmitter (the old one I got for $3 sucks), and begin designing the other sensors.

Read more