The American Rocketry challenge is a rocketry competition where accuracy is critical. However, it is difficult to keep accuracy in a constantly changing environment. So, I set out to develop active control electronics.
A Printed Circuit Board (PCB) fits all your components compactly onto one small plane. Saving mass and volume in model rockets is critical for success, so a custom circuit board helps tremendously with fitting into the strict requirements of ARC. To start out with my first circuit board, I followed BPS.Space’s excellent YouTube tutorial which you can see below!
After a couple iterations, I had ROCK V1.2, which is what we flew throughout the season while I developed our more complex circuit boards. We named it ROCK because our team was called “DWAYNEE THE ROCKETRY JOHNSON”!
What’s next?
ROCK was power-hungry, large, difficult to automate, ugly, and expensive. The solution? Surface mount technology. Essentially, instead of soldering pins through holes, tiny integrated circuits are soldered directly to the surface of the board. These are cheap, extremely small (some require a microscope to see!), look much better, and can be assembled through Pick-and-Place Machines automatically!
On the left you can see the surface mount design of Tektite-A, and on the right you can see the through-hole design of ROCK. While through-hole designs are simpler, surface mount designs are compact, cheaper, and can have automated assembly. ROCK cost over $60, while Tektite-A has a better microcontroller, sensors, and peripherals for just around $20! Tektite-A completely assembled is cheaper than just the Arduino Nano 33 BLE that ROCK used!
Why didn’t ROCK work at ARC Nationals?
The biggest problem with ROCK and why it didn’t work at ARC Nationals was that it used the Bosch BNO055 IMU in its Fusion mode. This maximum acceleration the sensor could read was just 4G. However, the rocket accelerates at up to 12G! The control algorithms must know the velocity, so the computer adds up the acceleration over time to get this.
Since the BNO055 couldn’t add up the accelerations during the motor burn since they exceeded 4G, I had to manually set the velocity after motor burnout. This had worked well during testing, however, the motor at nationals underperformed since it was from a different supplier. Because of this, the rocket thought it was moving much faster than it actually was, so it slowed itself down too much and ended up going 45 feet below the target altitude of 850ft.
Tektite-A has the Bosch BMI088 IMU, which can handle up to 24G of acceleration. So, this shouldn’t be a problem anymore!
Great Post, love to see the work. Ordered for my team!