Stabilized Model Rocket using Arduino: Project Advice

Hey everyone,

I’m an undergraduate at UCSD who has built a few home made model rockets and is currently in a rocketry club on campus. I’m aware of the risks of this project, and it’s not my intention to build a guided missile; just a stabilized rocket. (SEE EDIT) Disclaimer aside, I’ve decided to attempt a project over winter break where I will design and 3D print a rocket through AutoDesk Inventor.

I would like to stabilize the rocket through rotatable fins controlled through servos mounted on either the back end, where they will point roughly 10% towards the erroneous rotation, being in line with the body when the rocket is stabilized (flying directly upwards); or a second, smaller set of fins mounted on the front end (probably will help with center of mass given weight of servos) that rotate roughly 10% against the erroneous rotation, and be in line with the body when the rocket is stabilized. This will take a considerable amount of precision and rapidity of response.

I have limited exposure to microcontrollers, and would like to know the following: 1.) What controller is best suited for the required calculations, which must be both rapid and precise? I would need to integrate gyroscope and accelerometer data, use this to control 4 seperate servos, and would like to record and locally store altitude data. Is RPi better suited for this? 2.) What sensors would be best suited for this task? I have had a hard time finding a sensor which reliably provides high precision measurements of rotation. Would a seperate gyroscope / accelerometer be best, or an IMU? 3.) What servo motor would be best suited to rapidly respond to the input, while being lightweight?

I understand that with regards to sensors / motors, quality of measurement, weight, and price point will likely be a trade off. I’m willing to spend as much as necessary to get the minimum accuracy in calculations.

Edit: Just to clarify, there will be no telemetry or GPS to guide the rockets flight. I've done a fair amount of searching, and to the best of my knowledge, and based on other documented public projects, this is perfectly legal. I will either be launching and purchasing motors directly through my university club, or obtaining certification and documentation required to do so individually.

"I'm aware of the risks of this project, and it's not my intention to build a guided missile" It appears that you may not be totally aware of the risks, or just do not care about your future. Once you put guidance control hardware/software in a rocket, it becomes a missile, and categorized as a Guided Missile. Construction and/or possession of same is a federal felony in the United States, which will get you an all expenses paid (by you in the form of day labor) vacation for up to 15 years at a federally operated detention facility. you don't have to believe me, just your local branch of the ATFE.

If you want to stabilize your rocket, just set the fins to make the rocket spiral...

I’ve browsed the Internet and been unable to find anything specifically citing that a rocket designed to be stabilized, NOT guided, is strictly illegal. I have, however, found a handful of projects doing exactly what I have described. There is going to be no telemetry or systems that would enable this rocket to be guided at any target, nor is that my intention. I am planning to launch at a certified launch site, and I will fill out the necessary forms to do so.

1) A small one. Arduino Micro, Pro Mini or Teensy 3.2

2) You cannot get "high accuracy" in consumer-grade equipment. The real military stuff is expensive and you must sign stuff to say you're not going to export it.

However, the common consumer-grade stuff, like the sensors in a mobile phone, are quite adequate for this job. The MPU6050 or MPU9250 and their brothers are really very good for this.

3) A common RC hobby servo will be quite adequate. This is exactly the type of job that they are designed for. Just make sure that you put it on the aerodynamic balance point of the fin, to minimise the torque required.

The main problem is going to be vibration-hardening all the components. You can't rely on a breadboard. The initial acceleration of the rocket off the pad will pull chips out of their sockets. Get it all soldered together on proto-board and never let a wire just hang off a solder joint - always clamp it down to the board with a cable tie or something.

Don't throw away simple analog control. You can get extremely good results with 4 light-dependent-resistors (LDRs) aimed at the horizon. When one sees more bright sky than its counterpart on the other side, it can move the fins to make the rocket perfectly vertical.