Flight computer
The chip that flies inside N-1.
Our own flight computer measures height and acceleration through the whole flight, logs every sample and radios live telemetry to the ground. It only measures; it never fires anything.
Status
Designed, compiled, not yet flown.
The firmware compiles for both boards, and Mission Control has been tested end to end in Simulation mode, but none of it has been tested on real hardware yet. Everything here is the design, not a flight result.
The board
One small board does the measuring, logging and radio.
- Board
- Heltec WiFi LoRa 32 V3ESP32-S3 with an SX1262 LoRa radio and 8 MB of flash on one board.
- Barometer
- BMP280Air pressure, turned into height above the pad.
- Motion
- MPU6050Acceleration and rotation on three axes.
- Storage
- On-board flashEvery sample is written to a log file (LittleFS) that is downloaded over USB after the flight.
- Ground station
- A second Heltec V3Receives telemetry and passes it to Mission Control on the laptop over USB.
- Why not a Raspberry Pi
- ESP32A Pi is heavier, needs more power and is slow to boot.
Flight phases
It works out where it is in the flight by itself.
The thresholds below come straight from the firmware, so this page changes when the firmware does.
- PADOn the rail. Averages the pressure to zero the height, beeps 3 short for all OK, 2 long for don't fly.
- BOOSTMotor burning. Starts on a hard push (or a real climb if the motion sensor is missing).
- COASTMotor out, still climbing to the highest point.
- DESCENTPast the peak. The motor's own ejection charge has put the parachute out.
- LANDEDStill for 5 seconds. Chirps so it can be found; LOCATE works again.
Thresholds
The numbers it decides with.
| Setting | Value | What it means |
|---|---|---|
| PAD_LOG_MS | 100 ms | Sample at 10 Hz on the pad |
| PAD_FLASH_MS | 1,000 ms | ...but only write one sample a second to flash while nothing is moving |
| FLIGHT_LOG_MS | 20 ms | 50 Hz in flight |
| TELEMETRY_MS | 500 ms | 2 packets a second |
| LAUNCH_G | 2.5 g | Total acceleration that counts as launch (3 samples in a row) |
| LAUNCH_ALT_M | 20 m | Or climbing this far above the pad... |
| LAUNCH_VEL_MS | 5 m/s | ...while rising at least this fast (baro-only launch, used if the IMU is missing) |
| APOGEE_DROP_M | 2 m | Filtered altitude this far below the peak = past apogee |
| LANDED_STILL_MS | 5,000 ms | Vertical speed under 1 m/s for this long = landed |
| BOOST_MAX_MS | 5,000 ms | N-1's motor burns for about 1.6 s; after 5 s it is coasting whatever the IMU says |
| FALSE_LAUNCH_MS | 10,000 ms | 10 s after "launch" with less than FALSE_LAUNCH_M of climb = it was a knock |
| FLIGHT_MAX_MS | 300,000 ms | 5 minutes after launch, call it landed (an N-1 flight is about a minute) |
| GROUND_TAU_S | 600 s | On the pad, the ground pressure follows the weather with a 10 minute lag |
| LOG_RESERVE_BYTES | 524,288 bytes | 512 KB free = about 5 minutes at 50 Hz; less than that is a NO-GO |
| BATT_MIN_MV | 2,500 mV | A reading below this means the divider is off or no battery is fitted |
What it logs
30 bytes per sample, 50 times a second in flight.
On the pad it samples 10 times a second but only writes once a second while nothing moves, so a long wait at a club launch can't fill the flash.
Radio
Live telemetry to the ground, twice a second.
- Frequency
- 869.525 MHzUK licence-free band 869.40 to 869.65 MHz, 10% duty cycle limit.
- Spreading factor
- SF7Short packets, about 25 ms each.
- Bandwidth
- 250 kHz
- Power
- 14 dBm
- Rate
- 2 packets a secondAbout 5% duty, inside the 10% limit.
- Sync word
- 0x4EThe letter N, so other LoRa users are ignored.
Each packet is 25 bytes. The rocket and the ground station share these settings from one file, so they can't drift apart.
Measure only
Nothing on this board can fire anything.
- No pyro channel, no firing output, no deployment code. The certified motor's own ejection charge puts the parachute out.
- The only commands from the ground are PING and LOCATE. PING checks the radio link; LOCATE sounds the buzzer for 10 seconds. The rocket only acts on them on the pad or after landing.
- Electronic parachute deployment waits until the board has proven itself over several flights.
- Bad barometer readings are thrown away rather than turned into impossible heights, and a knock on the pad can't leave it stuck in flight mode.