Sim Racing Force Feedback Settings: A Repeatable Method

Force feedback settings are not a matter of taste, they are a signal problem with a correct answer. The wheel has a maximum output, the sim sends it a stream of forces that vary by more than an order of magnitude, and your job is to fit that stream underneath the ceiling without squashing it flat. Everything below follows from that. Set gain by measurement, set the damping family to zero, and only add effects back to fix a named complaint.

What does each force feedback setting actually do?

Overall gain is the master multiplier applied to every force the sim generates before it reaches the motor. It does not add detail, it scales what already exists. Doubling gain doubles the small forces and the large ones equally, which is why it is the setting that causes clipping.

Minimum force, sometimes called minimum torque or FFB floor, adds a fixed offset so that any nonzero signal is delivered at least at that level. It exists because gear and belt drivetrains have static friction: below a certain torque the motor cannot move the rim at all, so the quietest signals are lost. On a Logitech G920 or a Thrustmaster T300RS GT at $399.99 a few percent of minimum force recovers real information. On a direct drive base there is no gear train to overcome, so it mostly adds a constant hum that masks the very signals it is meant to rescue.

Damping resists rotational speed. The faster the wheel is turning, the harder damping pushes back. It is a good simulator of hydraulic power steering and a good way to stop a wheel oscillating on its own, and it is also a low pass filter on your hands, because fast small movements are exactly what it suppresses.

Inertia resists a change in rotational speed. It makes the wheel feel like it has more mass, which some drivers like because a real steering system does have mass. The trade is response: added inertia rounds off the leading edge of a kerb strike and delays the moment you feel a slide begin.

Friction resists motion at any speed regardless of direction. It makes a wheel feel heavier and more planted at very low steering speeds, and it is the effect most likely to hide the tyre signal you are actually listening for. Treat it as a last resort.

Filtering, sometimes called smoothing or interpolation, averages the force signal over time to remove noise and physics-engine steps. A small amount can clean up an older title that outputs a jagged signal. A large amount removes the sharp transients that tell you the surface changed, and it also adds latency, so what you feel arrives later than what happened.

What is clipping, and why does it matter more than strength?

Clipping is several different forces saturating at the same value, so you lose the ability to tell them apart. That is the whole definition and it is worth reading twice, because most descriptions frame it as the wheel being too strong, which misses the point.

Picture a mid corner on a bumpy kerb. The sim is generating self-aligning torque from the front tyres, a lateral load signal, a surface texture signal and a large impulse from the kerb. Suppose the self-aligning torque is 4 Nm, the kerb impulse is 9 Nm, and your gain is set so both exceed the base ceiling of 9 Nm. All of it arrives at 9 Nm. The kerb and the tyre load feel identical, the difference between them is gone, and the moment the front axle starts to give up is invisible to you because it was already pinned against the ceiling before it changed.

This is why heavy settings feel convincing and drive badly. The heaviest, most muscular force feedback configuration is usually the one carrying the least information, because it spends most of a lap saturated. The car communicates through the differences between forces, not their absolute size.

Clipping also explains the entire case for a high torque base. A MOZA R12 V2 at $399.99 run at 40 percent gain never saturates on a kerb, so the tyre signal underneath stays distinct. The 12 Nm ceiling was never about pulling your arms harder, it was about the ceiling staying out of the way. That is also why the drive type discussion in direct drive vs belt vs gear lands on resolution rather than strength.

How do you set gain by finding the clipping point?

This is a repeatable procedure and it takes about fifteen minutes per sim. Do it once per title, write down the answer, and stop guessing.

1. Set the base driver to full scale. In the MOZA, Fanatec, Thrustmaster or Logitech driver, set the maximum torque output to 100 percent and set every filter, damping, inertia, friction and interpolation slider to zero. You want the driver transparent, so the sim is the only thing shaping the signal. The exception is the safety torque ceiling, covered below.
2. Turn off every in-game effect that is not the physics. Road detail, engine vibration, kerb effect and similar sliders are synthetic overlays added on top of the tyre model. Zero them for the test. You can add a little back later once gain is fixed.
3. Pick one reference corner. Choose a medium speed corner with a kerb you use every lap. Consistency of the test matters more than which corner it is.
4. Open a force feedback meter. Most sims have one built in, and iRacing, Assetto Corsa Competizione and Automobilista 2 all show a live output bar or a clipping indicator. If your title has none, use the base driver telemetry window.
5. Start high and walk down. Set in-game gain deliberately too high, say 80 percent, and run three laps. Watch where the meter pins. Reduce gain in steps of 5 percent, three laps each, until the meter no longer reaches its ceiling through your reference corner including the kerb.
6. Take another 5 percent off. The corner you tested is not the heaviest thing you will ever hit. A wall tap, a wet track, a heavier car or a different circuit will all produce more force. That last 5 percent is your margin.
7. Now add back one thing at a time. If the wheel oscillates when you let go on a straight, add damping in steps of 5 percent until it stops and no further. If a specific sim outputs an audibly rough signal, add filtering one step at a time and stop the moment kerbs lose their edge.

Why not just use someone else's numbers? Because gain interacts with your base ceiling, your rim diameter and your car choice. A setting that is perfect on a 5 Nm base clips badly on a 3.9 Nm one, and a heavier rim changes the leverage at your hands. Community settings files are a good starting point and a bad final answer. The per-title starting points in force feedback settings by game are written to be tested, not copied.

What should you start from, by drive type and torque band?

The figures below are starting points for the clipping test above, not final settings. In-game gain assumes the base driver is at full scale with its own filters off. Every number is a percentage.

Read the gain column as the pattern it is: the more torque the base can produce, the lower the multiplier you apply to it. A MOZA R21 at $947.16 at 35 percent and a MOZA R3 bundle at 70 percent can deliver a similar peak weight at the rim. The difference is that the R21 has fourteen newton metres of unused ceiling above that peak, so nothing it sends you ever gets flattened.

Safety, before you touch gain. Set the hardware torque ceiling in the base driver before the first power-up, and set it low while you learn the wheel. A direct drive base at 9 Nm or more can sprain a wrist or a thumb during a sudden wall impact, and the reaction happens faster than you can let go. Never put your thumbs through the rim spokes, stand clear during the calibration sweep when the wheel turns to full lock on its own, and keep the emergency stop within reach on bases that have one. Confirm the frame can hold the base before you raise the ceiling: above roughly 8 Nm a desk clamp is not adequate and above roughly 12 Nm a light folding cockpit will flex or shift under a kerb strike. The wheelbase torque calculator matches a base to a mount.

What goes wrong when people tune by feel?

Gain set to whatever felt strong on lap one. An unfamiliar wheel always feels light because you are not yet reading it, so the instinct is to add gain until it feels like effort. Two weeks later the same setting is exhausting and you have been clipping the entire time. Gain is the setting to measure, never to feel.

Minimum force left at a driver default on direct drive. Some drivers ship with a nonzero floor, and on a base with no stiction that floor becomes a permanent low level force that never goes away. The symptom is a wheel that feels alive in the pit lane. Set it to zero on direct drive and let the physics start from silence.

Filtering used to fix a mounting problem. If the wheel feels rough and rattly, check whether the base is actually bolted through all four mounting points and whether the deck is flexing, before you smooth the signal in software. Filtering hides the symptom and costs you kerb definition permanently. A base that moves in its mount is a hardware fault, not a settings fault.

Effect sliders confused with physics. Road effect, engine vibration and kerb effect sliders in most sims are synthetic overlays, not outputs of the tyre model. Turning them up makes the wheel busier while making the tyre signal proportionally harder to hear. Keep them low or off, and get your information from the physics.

Chasing wheel settings when the real limit is the brake. A large share of the inconsistency people try to solve with force feedback tuning is actually brake modulation. A potentiometer brake measures travel, a load cell measures pressure, and pressure is how a real brake works. If your pedals came in a wheel bundle, a set like the Logitech G RS load cell pedals at $159.99 will change your lap times more than another evening of slider adjustment.

How do you know when the settings are right?

Three checks, all of which are things you can feel rather than measure. First, you can tell the difference between a kerb and the front tyres beginning to slide without looking at anything. Second, the wheel does not go flat and dead in the heaviest part of your slowest corner. Third, you can drive an hour without your forearms complaining.

If all three are true, stop adjusting. The remaining gains are in the rest of the rig: a frame that does not flex under the base, a load cell brake, a seat that holds you in place. If you are still choosing hardware, the ranked list in the best direct drive wheelbases is sorted by usable resolution at a mount you can realistically build, and the published figures for individual bases such as the MOZA R9 V3 sit on their own spec pages. Console drivers on the Fanatec CSL DD family and PC drivers on the Simagic Alpha Mini should run the same clipping procedure, since neither driver package sets a sensible default for you.