How to Test a Racing Wheel and Pedals

Most sim racing problems people blame on skill are actually hardware settings. A brake pedal that hits 100 percent at half travel will lock the fronts every time, and no amount of practice fixes it. A wheel whose rotation range does not match the in-game setting means your hands and the car disagree about where the front wheels are pointing.

None of this is hard to check. It just requires knowing what the correct numbers look like, which nobody tells you when the box arrives.

Start with device enumeration

Press Windows + R, typejoy.cpl, press Enter. Your wheel should appear with status OK.

Note how many devices you see. This varies more than people expect:

• Logitech G29, G920 and G923: wheel and pedals appear as one device
• Thrustmaster T300 and T248: base and pedals as one device, unless you have added a separate USB pedal set
• Fanatec: base as one device, with pedals either through the base or as their own USB device depending on the model
• Load cell pedal sets from Heusinkveld, Simagic and similar: almost always a separate USB device
• Shifters and handbrakes: usually separate unless they connect through the base

Select the wheel, click Properties, and you get a live test page. Turn the wheel and watch the X axis. Press each pedal and watch which axis moves.

For actual numbers rather than moving bars, a browser-based joystick tester reads a wheel the same way it reads any DirectInput device, and numeric output is what you need for the checks below.

Racing wheel calibration test: the sequence that matters

Do this in order. Later steps depend on earlier ones being right.

1. Let the wheel finish its startup sweep. Every force feedback wheel calibrates its centre and endpoints by turning lock to lock at power on. If you grab the wheel during this, the centre point will be wrong for the whole session. Hands off until it settles.

2. Check centre. With the wheel released, the steering axis should read exactly 50 percent, or 0 on a signed scale. Anything else and the startup sweep was interrupted, the base needs a firmware update, or the wheel rim is mounted off centre on a quick release.

3. Check full lock both ways. Turn slowly to the left stop and hold. The axis should reach 0 percent. The right stop should reach 100. If either side saturates early, or you can keep turning past the point where the number stops changing, your rotation range setting and your actual hardware limit disagree.

4. Check linearity. Turn to what feels like a quarter lock and note the value. Repeat on the other side. They should be symmetric. Asymmetry usually means the centre is off rather than the sensor being faulty.

5. Do not run the Windows calibration wizard. This is the single most common mistake. The joy.cpl calibration wizard was designed for cheap analog gamepads. Running it on a force feedback wheel writes registry calibration data that fights the wheel's own firmware calibration, and the result is a wheel that feels fine at centre and wrong everywhere else. Use the manufacturer's control panel instead. Our gamepad calibration guide explains the general principle of calibrating at the layer that owns the data.

Wheel rotation range: match hardware to game

Rotation range is how many degrees the wheel turns from lock to lock. It has to be set in two places and the two have to agree.

Typical hardware maximums:

WheelMaximum rotationLogitech G29 / G920 / G923900 degreesThrustmaster T248900 degreesThrustmaster T300 / T-GT1080 degreesFanatec CSL DD / GT DD1080 degrees and upMost direct drive bases1080 to 2520 degrees

Set the base to its maximum in the manufacturer software, then let the game control the effective range per car. Modern sims send a soft lock signal that limits rotation to whatever the car actually uses: roughly 900 degrees for a road car, 360 to 540 for a GT3, sometimes under 300 for a formula car.

If your sim does not support soft lock, set both the base and the game to the same fixed number. A mismatch here is the cause of the classic complaint that the car feels twitchy in one title and vague in another with identical settings.

Where to change it:

• Logitech: G HUB, or the older Logitech Gaming Software, wheel settings, operating range
• Thrustmaster: Thrustmaster Control Panel, rotation angle
• Fanatec: Fanatec control panel or the wheel's own tuning menu, SEN setting
• Moza: Pit House
• Simucube, Simagic and similar: each has its own configuration app

Logitech G29 test: the combined pedals trap

The G29 and G920 deserve their own section because of one setting that catches nearly everyone.

By default, some titles read Logitech pedals in combined mode, where throttle and brake share a single axis. Press the throttle and the axis goes up. Press the brake, and it goes down. Press both, and they cancel out, which is why you cannot trail brake and why left foot braking does nothing.

You want separate axes. The fix lives in the game rather than in G HUB for most titles: look for a setting called combined pedals, separate axes, or split axis, and turn combining off. In joy.cpl, you can confirm you have it right by pressing throttle and brake at the same time. If both axes move independently, you are set. If one number goes up and down, you are still combined.

Other G29 specifics worth checking. The brake pedal has a rubber progressive stop rather than a load cell, so it hits 100 percent well before the pedal reaches the floor. That is normal, and it is why a brake mod is the most common G29 upgrade. The clutch on the G29 is a third axis and is often unassigned by default. The G920 for Xbox lacks the clutch and the rev light strip that the G29 has, which occasionally confuses people comparing settings guides.

TrueForce on the G923 sends audio-derived vibration through the wheel alongside conventional force feedback. If a title supports it and it feels absent, check the TrueForce slider in G HUB rather than assuming the wheel is faulty.

Test pedals pc: the checks that actually matter

Pedals fail more often than wheels and get tested less.

Independent axes. Press each pedal alone and confirm exactly one axis moves. Then press all of them together and confirm all three move independently.

Full travel. Each pedal should read 0 percent at rest and 100 percent at the floor. Throttle almost always does. Brake frequently does not, and the direction of the error matters. Reaching 100 percent early means you have less usable travel than you think. Never reaching 100 percent means you can never fully lock the brakes, which is worse.

Rest position. Take your feet off and watch for thirty seconds. Values should sit at zero and stay there. A throttle resting at 2 or 3 percent is why your car creeps. This is usually a calibration issue rather than a fault, and a deadzone test will show you exactly how much dead travel you are working with.

Load cell calibration. If you have a load cell brake, its calibration is a maximum force setting rather than a travel setting. Set it to a force you can apply repeatably under braking, not the hardest you can stamp once. Most people set it far too high, then wonder why they cannot modulate. Recalibrate when you change your seating position or pedal angle, because both change the force you can apply.

Brake curve. Linear is the correct default. Add curve only after you have practised with linear for a few sessions, and only in one direction: most drivers who need a curve want more resolution at low pressure, not less.

Pedals that connect through the wheel base rather than direct USB are also worth testing separately from the base, because a bad RJ12 connection produces intermittent dropouts that look exactly like a failing sensor.

Force feedback test: what good and bad feel like

Every major manufacturer's control panel includes a force feedback test. Use it, because it isolates the hardware from whatever the game is doing.

Logitech G HUB has a force feedback test that applies spring and damper forces. Thrustmaster Control Panel has individual sliders for constant, periodic, spring, and damper, each testable on its own. Fanatec exposes FFB testing in its control panel and lets you drive individual effect types. Moza Pit House has a similar diagnostic page.

Run each effect type on its own and pay attention to:

Notchiness. Feeling detents as the wheel turns under force means worn gears on a gear-driven wheel, or a belt problem on a belt-driven one. It gets worse, never better.

Oscillation at centre. The wheel shaking around straight ahead means damping is too low or the base's internal damping setting has been turned off. Not usually a fault.

Dead spot at centre. A small band where forces disappear is mechanical slack in a gear or belt drive. Normal in small amounts on entry-level wheels, a problem when it grows.

Clipping. When forces exceed what the base can output, detail disappears at exactly the moment you need it. Test by driving a corner with maximum force and seeing whether the wheel goes numb at peak load. Lower the in-game gain until the numbness stops. SimHub and several sims have built-in clipping meters.

No force at all. Check the power supply first. Force feedback wheels draw far more than USB provides, and a partly seated barrel connector produces a wheel that enumerates fine and does nothing.

Since a wheel's force feedback runs on the same virtual device layer as controller rumble, a general vibration test is a quick way to confirm the operating system can drive haptics on your device at all before you go digging in the manufacturer software.

Latency, and when to suspect it

Sim racing rewards low, consistent input delay. A wheel that feels disconnected without any obvious calibration fault is worth latency testing before you replace anything.

Check with an input lag test, then look at the usual culprits: USB hubs between the base and the PC, wireless peripherals sharing a controller, and any overlay software injecting itself into the input path.

Wheels should be plugged directly into a rear motherboard USB port. Front panel ports and unpowered hubs both introduce problems that show up as intermittent disconnects rather than steady lag, which makes them harder to diagnose.

A quick pre-race checklist

Before a league race or anything that matters:

1. Power the base and let the startup sweep complete without touching it
2. Confirm centre reads dead centre in joy.cpl
3. Confirm full lock reaches both extremes
4. Confirm rotation range in software matches what the game expects
5. Press each pedal and confirm independent axes and full travel
6. Run the manufacturer's force feedback test for notchiness or dead spots
7. Check pedal rest values sit at zero

Two minutes, and it catches the class of problem that only becomes obvious on lap one.