Log No.1 of the 25/26 Hungarian National Engineering Olympiad

OTIO 2025/26 

I actually started this back in like June (It's mid-October now) but I just now thought about starting a blog. I'm honestly not sure who I'm writing this to, but why not? In any case, I'm 14 right now so excuse my unprofessional tone.

Well, for this year (since I also took part in this OTIO last year with PingPal) I first got the idea of a portable X-ray for which I would've learnt glassblowing to make the X-ray cathode tube (I believe that's what it's called), and while I myself would've been reasonably comfortable working with radiation, my parents weren't exactly thrilled about the idea (Jee, I wonder why) so I came up with something that I've actually been looking to do for quite a few years now; a rocket guidance system.

Since it's an innovation olympiad, I also decided to make it open-source (GitHub and Maker.io account pending lol)

First, I obviously had to select something to act as the main processor of the system, for which I chose a Raspberry Pi CM5 module. I thought about non-premade solutions like some TI microprocessors, but the CM5 has some insane processing power and RAM for the size, and already runs on Linux so I wouldnt've had to do assembly for every single thing, which would've made this project practically impossible for me.

For sensors, I currently have an MS561101BA03-50 as a barometer/altimeter and also a temperature sensor, though I don't really need one- This once can measure down to 10mbar, which is about 30k meters of altitude according to this random website, which seems plenty to me. And it draws current in the uA range, so fairly efficient. 

In addition, I have a BNO085 9-axis IC that does 3-axis acceleration, has a gyroscope and a magnetometer (compass). The good thing about this IC is that it checks all of the motion sensing parts I need with those 3 sensors i mentioned, but the accelerometer clips above 8G-forces. And since a high-speed rocket launch can get well above 8Gs (probably), I also included an ADXL375BCCZ, which can do up to 200g-forces. Plenty.

Then also a GPS module, pre-built because I'm nowhere near competent to work with RF, let alone with such high frequencies.

Handling all of these sensors (besides the GPS) will be handled by an STM32F042K6T6TR with I2C. The reason for this is to offload work from the CM5 by getting the signals processed externally in the STM, perhaps Kalman filter the accelerometer data and then send a parsed and easy-to-interpret data packet through UART to the CM5.

A second STM of the same model will act as a Flight Termination System [FTS] that acts as a watchdog. If the CM5 or the sensor STM stops responding or any flight-critical systems go down (again, this can be coded by the user, so it will engage whenever they want it to), it will trigger a programmable sequence of events such as opening MOSFETs and triggering parachute charges, cutting power to certain systems, disabling thrust, etc... Since this is one of the most crucial parts of the redundancy/safety systems, it will have its own 3.3V voltage regulator, but in case the regulator goes down (highly unlikely. How would it even?), the regulator of the Sensor STM can pick up the slack, but it doesn't backfeed the other way around due to a diode (the voltage drop is manageable, it drops to ~3V, which is still in range for the STM). 

But since all of these rely on having an actual voltage input, the FTS won't engage if power is cut to the whole PCB. Which is why it will also have its own connector for a (or multiple in parallel) 3.7V Li-Pol battery, with a diode to avoid it getting charged when main power is actually supplied, and it will also give enough voltage drop for it get into V-range for the STM. 

A PCA9685PW will drive 12 servo motor channels. Each channel will have a 2W 100mR THT resistor acting as a low-side shunt for current measurement. A PIC18F46K22 MCU (40-PDIP package) will use the ADC channels with the built-in 1.024V reference selected, and since the ADC channels are 10-bit, it will have ~1mV resolution steps (though in practice it will probably be closer to 10mV or even 100mV. Still good enough.)

I'Il say more about the telemetry in later posts, because I'm still not completely sure what to use. I currently have a 433MHz LoRa selected, which has some insane range (~10km), but it needs a very low duty cycle for transmission with it to be legal here in Hungary. Then you have your common HC-12 433MHz Transciever, but it only has a 1km range, though can transmit more. So still not sure.

I'm pretty sure that's all for now, I actually have most of the circuit done already (again, started in June), so there won't really be a lot of updates, only retroactively.

Tamás Zétény Tóth

TTZ

(P.S.: I have no idea how to write blog posts)