Project Details

1. The Motivation: Why Build a Custom ECU?

The commercial standalone Engine Management System (EMS) market is dominated by expensive, closed-source "black boxes". While they work well for standard automotive applications, they often fall short in high-revving, single-cylinder motorcycle engines used in extreme motorsport environments like Formula Student. I needed a system that offered absolute bare-metal control over timing, uncompromised EMI immunity, and custom tuning algorithms. As a solo engineer, I decided to build the entire stack from scratch: the hardware, the firmware, and the PC tuning software.

2. Hardware Architecture

The physical ECU was designed to survive the brutal conditions of a race track, focusing on signal integrity and vibration resistance.

• Microcontroller: The core is an STM32 MCU. By running bare-metal C++ (no RTOS overhead), the system achieves deterministic, microsecond-accurate execution of ignition and injection events.
• Signal Conditioning: Crankshaft position reading is arguably the most critical function of an ECU. I utilized a MAX9924 Variable Reluctance (VR) sensor interface. This IC provides adaptive peak thresholding, ensuring clean digital pulses even when the raw analog signal is heavily distorted.
• Enclosure: The PCB is housed in a custom-designed, IP67-rated 3D-printed enclosure, featuring lap joints and mechanical dampening to protect against extreme engine vibrations.

3. Firmware: Algorithmic Innovations

Writing the firmware was the most challenging and rewarding part of this project. Instead of relying on standard generic code, I implemented custom mathematical models to solve specific motorsport problems:

• Cam-less Phase Synchronization:

  In a 4-stroke engine, the crankshaft rotates twice per full cycle. Traditional sequential injection/ignition requires a camshaft sensor to determine whether the engine is on the compression or exhaust stroke. I eliminated this hardware requirement by calculating the angular acceleration of the crankshaft. By dynamically analyzing the micro-accelerations during the power stroke versus the intake stroke, the ECU mathematically determines engine phase with zero extra sensors.

• Extreme EMI Immunity:

  Racing ignition coils operating at 15,000+ RPM generate massive Electromagnetic Interference (EMI), which often causes "Loss of Sync" in aftermarket ECUs. I implemented a dynamic temporal window algorithm. The ECU calculates the expected arrival time of the next crank tooth based on the current derivative of the engine speed. Any pulse arriving outside this mathematical window is instantly rejected as noise.

4. NX-Tuner Pro: Custom PC Software

Most DIY ECUs piggyback on third-party tuning software like TunerStudio. I chose to develop a dedicated, ground-up desktop application using Python and PyQt5.

• Mathematical Auto-Tuning: The GUI features a strictly deterministic, closed-loop Auto-Tune system. It reads live wideband O2 sensor data, calculates the AFR error delta, and autonomously applies weighted corrections to the Volumetric Efficiency (VE) table in real-time.
• High-Performance UI: The software utilizes pyqtgraph.opengl to render the 16x16 VE and Ignition maps in a fully interactive, GPU-accelerated 60FPS 3D space. It also features anti-aliased analog gauges and seamless MSQ file compatibility.

5. Track-Proven Validation (Motorsport Ready)

A bench-tested ECU is only half the battle. NX Tuning Tech is the official engine management system for our university's Formula Student race car. It has been stress-tested under maximum load on the track, natively running advanced motorsport strategies:

• Launch Control: Hardware-interrupt driven, holding precise RPM targets.
• Quickshifter Logic: Closed-loop gear cut strategies.
• Anti-Lag: Deep ignition retard and fuel enrichment to maintain manifold pressure off-throttle.

This solo-developed ecosystem proves that a custom-engineered, algorithmically superior system can stand toe-to-toe with the biggest commercial names in the motorsport industry.