{"id":434,"date":"2026-05-31T00:43:42","date_gmt":"2026-05-30T21:43:42","guid":{"rendered":"https:\/\/keremgurevin.com\/?p=434"},"modified":"2026-05-31T00:43:42","modified_gmt":"2026-05-30T21:43:42","slug":"houston-how-did-i-get-here-rocketry-log-1","status":"publish","type":"post","link":"https:\/\/keremgurevin.com\/?p=434","title":{"rendered":"Houston, how did I get here? | Rocketry Log #1"},"content":{"rendered":"\n

Hallo zusammen! In the last few years, my life has gone through dramatic changes, including but not limited to dipping my toes into the utterly complex yet intriguing world of high school, speedrunning<\/s> learning German, and joining my school’s rocketry club.<\/p>\n\n\n\n

The latter was, in my honest opinion, one of the best coding opportunities I have ever come across.<\/p>\n\n\n\n

For those who don’t know, the Turkish government actively encourages schools to get students involved in project-based learning to solve real-world problems. The biggest, most prestigious government-backed platform for this is TEKNOFEST<\/a>. Founded by the T3 Foundation<\/a> (the same organization behind the Deneyap Technology Workshops<\/a>, from which I graduated with an award<\/a>), TEKNOFEST is less of a rigid science fair and more of a massive technology festival. It hosts high-stakes competitions across a variety of cutting-edge fields, including AI, UAVs, agricultural tech, and, most notably, rocketry.<\/p>\n\n\n\n

So, how did I start coding rockets? Well, I am still in awe of the whole experience, so I might be exaggerating a tad. Nevertheless, here is how it began.<\/p>\n\n\n\n

Meeting “Mr. Hypermach”<\/h3>\n\n\n\n

Around February of 2026, I had the chance to step into my school’s TBA, which roughly translates to a makerspace. Select schools, like the one I attend, have a room dedicated to all things maker, ranging from 3D printers and laser engravers to professional-grade music recording equipment. In this makerspace, I happened to meet the president of Hypermach<\/a>, the school’s rocketry team.<\/p>\n\n\n

\n
\"\"
Part of aforementioned TBA. Photo from publicly available sources<\/em><\/figcaption><\/figure>\n<\/div>\n\n\n

At the time, I knew the school had a rocketry team, that they had won an award somewhere<\/a>, and that they had a lengthy admission process for newcomers consisting of a giant education playlist followed by multiple interviews and presentations. Although I am well aware that a Google search or two would reveal who he is, I’ll call him Mr. Hypermach for the entirety of this blog post.<\/p>\n\n\n\n

Mr. Hypermach and I were having a chat about additive manufacturing and how Hypermach used different types of filament around the rocket to cut down on weight and improve structural integrity. All of a sudden\u2014I’m still not sure how it came to this\u2014he offered me a spot on the team as an Avionics Technician, essentially coding the rocket.<\/p>\n\n\n\n

At first, I was a bit bamboozled. It caught me absolutely off-guard. Me, a coding beginner whose portfolio consisted of nothing more than a simple calculator made in 2020 as a first Lua project<\/a> and a modular smartwatch using a Deneyap Kart<\/a>, was being offered the chance to code the avionics system for a rocket.<\/p>\n\n\n\n

I hesitated initially. Usually, whenever I get involved in projects like this, I end up having to undertake the entire workload because others simply stop caring. But as he showed me photos of previous rockets and the launch site, I developed a sense of trust in the team. They had done this multiple times before, there was an existing avionics team, I could get all the help I needed, and I got to bypass the lengthy admission process. So, after blankly staring into Mr. Hypermach’s eyes for about five seconds, I calmly said, “Sure!”<\/p>\n\n\n\n

Thus marked the beginning of my Hypermach journey.<\/p>\n\n\n\n

Trial by Fire: The Arduino Mega and TEKNOFEST Protocols<\/h3>\n\n\n\n

Combined with my excitement and utter surprise, the following hour resulted in one of the funniest situations I’ve ever found myself in.<\/p>\n\n\n\n

After I agreed to join, Mr. Hypermach literally pushed me toward the preexisting avionics team. They sat me down on one of their many stools in front of a laptop. Connected to that laptop was an Arduino Mega<\/a>. That was our custom flight controller.<\/p>\n\n\n

\n
\"\"
Arduino Mega 2560 Rev3, Arduino S.r l.<\/a><\/figcaption><\/figure>\n<\/div>\n\n\n

Yes, you heard me right: a custom<\/em> flight controller.<\/p>\n\n\n\n

You might be thinking, Well, what’s the big deal about a few sensors hooked up to a microcontroller? You’re just supposed to make it read and send data.<\/em><\/p>\n\n\n\n

WRONG.<\/strong><\/p>\n\n\n\n

TEKNOFEST operates on a bit of a different level. To pass their rigorous requirements, you must master, and I don’t use the term “master” lightly, two critical parts of their testing protocol: the Sensor Monitoring Test (SMT)<\/strong> and the Synthetic Flight Test (SFT)<\/strong>. These are not mere suggestions or nice-to-haves; they are the literal gatekeepers that determine whether your flight computer is cleared for flight or eliminated from the competition.<\/p>\n\n\n\n

1. Sensor Monitoring Test (SMT)<\/h4>\n\n\n\n

The SMT is essentially a “sanity check” for your data acquisition pipeline. Once the ROKETSAN test software<\/a> triggers this mode, your flight controller must ignore normal idle processes and focus entirely on streaming live sensor data back to the test bench at a stable 10 Hz. This test is designed to expose noise, filtering issues, or improper calibration in your hardware. You are validating that your raw inputs, altitude, acceleration, and orientation, are being correctly processed and that your communication link is rock-solid. If there is jitter or data dropout here, your team won’t be cleared for launch.<\/p>\n\n\n\n

2. Synthetic Flight Test (SFT)<\/h4>\n\n\n\n

While the SMT checks if your sensors work, the SFT is a trial by fire for your flight algorithm. In this scenario, the ROKETSAN test software stops listening to your physical sensors. Instead, it injects “synthetic” data directly into your flight controller via the RS232 port. Your flight computer is expected to ignore its own physical sensors and treat this artificial, high-pressure data as if it were a real, unfolding launch.<\/p>\n\n\n\n

The ROKETSAN software simulates both normal flight profiles and abnormal scenarios, such as sudden, erratic spikes in acceleration or impossible pressure drops, to see how your flight computer handles chaos. During this test, you are required to send back status information packets at 10 Hz, updating the committee on which stage of the flight algorithm has been triggered (e.g., engine burnout, apogee detection, or parachute deployment). If your algorithm cannot maintain stability or fails to trigger the recovery system under these synthetic constraints, your mission ends right there on the testing bench.<\/p>\n\n\n\n

Reverse Engineering the Misinformation<\/h3>\n\n\n\n

These are fair requirements, and I have to give the committee credit: they are thoroughly explained throughout the specsheet, complete with implementation examples. The problem? The documentation was riddled with errors.<\/p>\n\n\n\n

If you look at pages 7 or 9 of their official specsheet<\/a>, specifically bytes 32 and 33, you will notice duplicate byte assignments. Furthermore, the code snippets tell you to use Little Endian when the system actually requires Big Endian (or vice versa).<\/p>\n\n\n\n

This misinformation problem forced me to write copious amounts of debug scripts and COM port sniffers just to reverse engineer how the ROKETSAN test software actually reads and sends data. I’m not complaining, though! It forced me to learn how COM ports and serial connections work, and how to sniff them behind Windows’ back.<\/p>\n\n\n\n

Ultimately, I wrote the necessary code to run on an Arduino Mega. Despite the hurdle of navigating misleading documentation and encrypted SFT data files, making this part work was incredibly rewarding. I probably wouldn’t have learned any of this from a standard coding course or a YouTube video.<\/p>\n\n\n\n

Upgrading to the Raspberry Pi Pico 2<\/h3>\n\n\n\n

After I completed the SMT-SFT phase, the head of avionics, let’s call him Mr. Avionics, asked me if it was feasible for the team to move from the Arduino Mega to the Raspberry Pi Pico 2<\/a>, as it is smaller, more efficient, and significantly faster. After a lot of research and a few fairly sleepless nights, I put together a comprehensive document detailing how the transition would affect the team, alongside a successful port of the SMT-SFT code for the Pico 2<\/a>.<\/p>\n\n\n\n

After a bit of back-and-forth, we officially decided to make the switch. Following this endeavor, the team decided it would be best if I also took over the development of the entire communication software.<\/p>\n\n\n

\n
\"\"
The Raspberry Pi Pico 2 Raspberry Pi Foundation<\/a><\/em><\/figcaption><\/figure>\n<\/div>\n\n\n

Diving into Long-Range Comms (LoRa)<\/h3>\n\n\n\n

Before I tell you about the comms system, it helps to understand how it works. LoRa<\/strong> (short for Long Range) is a low-power, wireless communication technology designed for remote telemetry tracking. Instead of relying on cellular networks, Wi-Fi, or Bluetooth, LoRa operates on unlicensed radio frequency bands (ISM bands, like 868 MHz in Europe) to transmit small packets of data over massive distances.<\/p>\n\n\n\n

The secret behind LoRa\u2019s impressive range and noise immunity is its underlying modulation technique: Chirp Spread Spectrum (CSS)<\/strong>. Traditional wireless communications transmit data on a fixed, narrow frequency, making them highly susceptible to interference. LoRa, however, encodes data using “chirps”, radio signals whose frequency increases or decreases over time, spreading the signal across a wider bandwidth. Because the receiver knows the exact characteristics of these sweeping chirps, it can decode signals that are incredibly weak, even when they drop well below the ambient radio noise floor.<\/p>\n\n\n

\n
\"\"
A linear frequency modulated upchirp in the time domain. Wikipedia<\/a><\/em><\/figcaption><\/figure>\n<\/div>\n\n\n

Luckily, the communication implementation was much more straightforward than the SMT-SFT reverse engineering. I hooked up E22-900T30D LoRa modules to two separate Picos: one transmitter and one receiver. I wired the flight controller’s sensors to the transmitter Pico and started coding. TEKNOFEST requires telemetry data to be sent in strict 78-byte packets, but this time the specsheets were accurate, so I carried on without any issues.<\/p>\n\n\n\n

I had the Picos transmitting data exactly how TEKNOFEST wanted<\/a> in almost no time. I also added serial compatibility for the official Referee Ground Station program<\/a>, which displays the live telemetry on a Windows interface.<\/p>\n\n\n\n

Everything was going completely smoothly. Almost like there was going to be a fairly big issue coming up just around now.<\/a><\/p>\n\n\n\n

The First Comms Test and the Birth of HypermachHYI<\/h3>\n\n\n\n

On May 7th, 2026, Hypermach went out for its first long-range communication test. The target distance was roughly 6km, spanning from B\u00fcy\u00fckada<\/a> to coastal Pendik<\/a>. Theoretically, LoRa can easily handle a 6km line-of-sight<\/a> transmission provided it can draw enough current. Our setup was getting enough power, yet the connection abruptly cut out near the 1km mark. We tried everything we possibly could, including switching channels, but to no avail. The test was a semi-failure.<\/p>\n\n\n\n

After evaluating our findings, we determined that the Yagi<\/a> antenna we used on the receiver side wasn’t powerful enough for the distance, so we swapped it out for a higher-gain model. This fixed the problem; we successfully achieved a 1.5km link even through a crowded, high-interference urban area during a later test.<\/p>\n\n\n

\n
\"\"
A modern high-gain UHF Yagi television antenna with 17 directors, and one reflector (made of four rods) shaped as a corner reflector<\/a>. Wikipeda<\/a><\/em><\/figcaption><\/figure>\n<\/div>\n\n\n

However, another glaring problem emerged during field testing: carrying a giant laptop around. It required 2 to 3 teammates crowded around a dim, 300-nit screen that lasted 90 minutes on battery power at best.<\/p>\n\n\n\n

To solve this, I built a custom Android app called HypermachHYI<\/a><\/strong>.<\/p>\n\n\n\n

The goal was to replace our bulky, power-hungry laptop setup with a lightweight, bright smartphone or tablet screen that fits right in your palm and runs for hours. The core idea was simple: pipe the telemetry data directly from our LoRa receiver into an Android device. The app reads the incoming data packets exactly as the ROKETSAN Referee Ground Station software would, meaning I didn’t have to modify my existing flight computer code.<\/p>\n\n\n\n

To ensure maximum reliability without relying on unstable wireless connections in crowded testing zones, I implemented a direct wired link using a USB-OTG (On-The-Go) cable. By connecting the receiver Pico directly to the phone via a USB-to-TTL converter, the app taps into a raw serial data stream the moment it is plugged in.<\/p>\n\n\n\n

Behind the scenes, the app relies on a dedicated background service to manage the serial connection. Since the E22 LoRa module outputs data continuously, a custom UsbSerialManager<\/a><\/code> handles the low-level byte streams, filtering out data corruption and ensuring seamless hot-plugging if a cable accidentally gets tugged during a test.<\/p>\n\n\n\n

That is everything that has happened so far with Hypermach. Stay tuned for Part 2!<\/p>\n\n\n\n

Thank you for reading!<\/p>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

Hallo zusammen! In the last few years, my life has gone through dramatic changes, including but not limited to dipping my toes into the utterly complex yet intriguing world of high school, speedrunning learning German, and joining my school’s rocketry club. The latter was, in my honest opinion, one of the best coding opportunities I […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"iawp_total_views":8,"footnotes":""},"categories":[47,46],"tags":[41,63,70,65,66,57,59,60,58,61,68,67,69,64,62],"class_list":["post-434","post","type-post","status-publish","format-standard","hentry","category-hardware","category-software","tag-code","tag-deneyap","tag-deneyap-kart","tag-gurevin","tag-hardware","tag-hypermach","tag-hypermach-rocket","tag-hypermach-rocketry","tag-hypermach-roket-takimi","tag-kerem-gurevin","tag-pico","tag-raspberry-pi","tag-raspberry-pi-pico","tag-sensaat","tag-teknofest"],"_links":{"self":[{"href":"https:\/\/keremgurevin.com\/index.php?rest_route=\/wp\/v2\/posts\/434","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/keremgurevin.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/keremgurevin.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/keremgurevin.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/keremgurevin.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=434"}],"version-history":[{"count":10,"href":"https:\/\/keremgurevin.com\/index.php?rest_route=\/wp\/v2\/posts\/434\/revisions"}],"predecessor-version":[{"id":456,"href":"https:\/\/keremgurevin.com\/index.php?rest_route=\/wp\/v2\/posts\/434\/revisions\/456"}],"wp:attachment":[{"href":"https:\/\/keremgurevin.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=434"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/keremgurevin.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=434"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/keremgurevin.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=434"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}