How to test your wheel and pedals
- Connect the wheel base and the pedals by USB, plug in the wheel's power supply and let it calibrate. Most wheels need their power supply for the axes to work at all. Then turn the wheel or press a button: browsers only reveal a device once it sends input.
- Follow the set-up: turn the wheel a quarter turn to the right, then press the throttle, the brake and the clutch one at a time. Wheels have no standard layout in browsers, so this is how the tester learns which axis is which, whether a pedal sits on its own axis or shares one with another ("combined pedals"), and which way each one moves.
- Set the rotation to the value in your wheel's software. The Logitech G29 turns 900° lock to lock (Logitech) and the Thrustmaster T150 and T300 up to 1080° (Thrustmaster). The browser sees the wheel as −1 at full left lock and +1 at full right lock, so the degrees shown are only correct when both settings match.
- Run the three health tests, press every button and shift through every gear, then read or share the wheel health report.
Reading the results
Rest and centre. With the wheel straight and nobody touching it, the reading should be within about 2° of centre and stay still. A wheel that points straight but reads several degrees off needs recalibrating. Released pedals should read 0% and jitter less than ±0.5%: a pedal that reads a few percent while untouched will make a game brake or accelerate slightly all the time.
Pedal linearity and spikes. Press one pedal slowly over two or three seconds. The plot shows its output over time next to a dashed line for a perfectly steady press. Your foot is never perfectly steady, so a gentle curve is normal. What matters is the shape of the line: a spike is a single reading that leaps and comes straight back, a jump is a sudden step that stays, a dip is the output falling while you keep pressing, and a flat spot is the output standing still mid-travel. Spikes and jumps are the classic signs of a worn or dirty potentiometer.
Steering sweep. The same analysis for the steering sensor, from full left lock to full right lock. Hall-effect and optical sensors (the G29 uses a Hall sensor) rarely spike, so a spiky steering trace points to a loose connection or interference as much as to the sensor.
Glitches during use are counted all the time, not only during a test: any reading that leaps by at least 20% of the travel and is back on the very next reading. A fast stomp on the brake moves just as far but doesn't snap back in one reading, so it isn't counted.
Fixing spiky or dead pedals
- Rule out settings. Check the wheel software's pedal calibration, deadzones and "combined pedals" option. On Thrustmaster bases the pedal mode must match the pedal set, otherwise pedals stop short of full travel (Thrustmaster support).
- Clean the potentiometers. With the pedals unplugged, open the pedal base and spray electrical contact cleaner into the potentiometer, then work the pedal through its full travel twenty or thirty times. iFixit has a step-by-step guide for Logitech G29 pedals.
- Replace or upgrade. Worn potentiometers are cheap to replace. Contactless upgrades (Hall-effect sensors, or a load cell for the brake) measure travel without a sliding contact and don't wear the same way.
Run the linearity test before and after: the report link records the numbers, so you can compare them later.
What a browser can and can't test
Browsers read wheels through the Gamepad API, which reports axes and buttons only. Its only output is gamepad rumble (W3C Gamepad extensions), so force feedback can't be tested in any browser, and neither can rev lights, displays or Logitech's TRUEFORCE. Chrome and Edge also report at most 32 buttons per device (Chromium source), and Safari passes on only six axes with no hat switch or slider (WebKit source), so use Chrome, Edge or Firefox for a wheel. Flight sticks and HOTAS setups have their own joystick tester, and gamepads their controller tester with stick drift and rumble tests. To see how quickly you react to a light, try the reaction time test.