Formula Student / RF telemetry and embedded data

In progress

N7 Racing Team

Formula Student EV - RF telemetry link and the STM32 / CAN data chain behind it

Identification

Team
N7 Racing Team, ENSEEIHT / INP Toulouse
Competition
Formula Student, electric vehicle class
Role
Data pole, then head of the Data pole (since May 2026)
Period
Sep 2025 - present
Rulebook
Formula Student rulebook, checked against every design decision

Scope

Radio
LoRa link between the car and the pit, and its integration on the vehicle
Embedded
STM32 acquisition, analogue conditioning and ADC on the sensor side
Bus
CAN and CAN-FD, with the DBC as the single source of truth for frame layout
Ground
Logging, offline decoding and validation of what was recorded
01

LoRa telemetry link

The link that carries vehicle state from the car to the pit during a run.

  • LoRa linkChirp spread spectrum in a licence-free ISM band, chosen for range and robustness at low data rate. Spreading factor, bandwidth and coding rate trade range against update rate. The operating band is not published until the hardware settles it.
  • Car to pit linkOne end on the moving vehicle, one in the pit. The link has to survive a metal and composite structure, a high-current traction system, and an antenna position constrained by the chassis.
  • Telemetry architectureVehicle health at a low rate over the radio; full-rate data logged on board and recovered afterwards. The radio is not the system of record.
  • Radio integrationAntenna placement and grounding on the car, separation from the traction system, and a pit receiver feeding a time-series store and a live display.
  • Frame designA packed binary telemetry frame, defined once and mirrored on both ends. A mismatched layout produces plausible wrong numbers rather than an error.

RelatedEmbedded and wireless skillsTelemetry on the avionics side

02

STM32, CAN and ECU data chain

The acquisition chain from a physical quantity to a decoded, validated record.

  • Sensor acquisitionWheel speed, steering, inertial measurement, suspension travel, temperatures and position, each with its own sampling requirement.
  • Analogue conditioningScaling into the converter range, filtering before sampling, and references that do not move with the supply.
  • ADCSampling rate set by the signal, acquisition time set by the source impedance.
  • STM32The acquisition microcontrollers: peripherals, timing, and getting samples off the converter and onto the bus without the processor becoming the bottleneck.
  • CAN and CAN-FDCAN for the classical traffic, CAN-FD where the payload no longer fits in eight bytes. Bus load and identifier priority are design parameters.
  • Inverter dataDemanded and actual torque, speed, motor and inverter temperatures and error states, read from the traction inverter over CAN.
  • DBCThe database defining every frame, signal, scaling and unit on the bus, kept as the single source of truth with one owner.
  • LoggingRaw frames logged on board with timestamp, identifier and payload, uninterpreted. Raw survives a wrong decoder.
  • DecodingOffline decoding through the DBC, so a decoding mistake is a rerun rather than a lost session.
  • ValidationRecorded data checked against what the vehicle did, and against the rulebook where a channel is required.
03

Tools and bus

Embedded and bus

  • STM32
  • C / C++
  • ADC
  • Analogue conditioning
  • CAN
  • CAN-FD
  • DBC
  • CANoe
  • SPI
  • I2C
  • UART
  • SD logging

Radio and ground

  • LoRa
  • Link budget
  • Antenna integration
  • Packed binary frames
  • Python
  • cantools
  • Time-series storage
  • Live dashboard

No range, packet-loss, bus-load or sampling figure is published; the system is in development and the numbers belong to the team's records.