Outboard vs inboard wheel speed sensing for motorsport ABS

Quick answer: A motorsport ABS is only as good as its wheel speed signal. Measure at the driveshaft and the driveline's compliance, backlash and a moving sensor gap all become part of the measurement; the classic result is a sensor fault lamp and no ABS mid-session, under the hardest braking of the day. Relocating the tone ring and sensor outboard puts the measurement at the wheel. 909's kits do it per platform and per hub package, and keep the parking brake.

The session is going well. Brakes are up to temperature, you're braking later each lap, and somewhere in the middle of a big stop the ABS warning lamp comes on and the system drops out. The pedal goes hard. Nothing is broken, nothing is leaking, and back in the paddock every sensor reads fine on a rolling wheel.

That failure has a common cause on converted cars, and it usually isn't the sensor. It's where the sensor is measuring from.

What the ABS is actually listening for

A motorsport ABS doesn't measure grip. It infers it, from wheel speed.

The system watches each wheel against a reference for how fast the car is travelling, and when a wheel starts decelerating faster than the car possibly can be, it recognises the onset of lock and releases pressure. Then it reapplies, and watches again, many times a second. The whole quality of that loop rests on one question: how faithfully does the signal describe what the wheel is doing?

Everything below is about what happens when the answer is "approximately".

What sits between an inboard sensor and the tyre

Put the sensor at the inboard end of the driveshaft and every component between it and the tyre becomes part of the measurement: the shaft, the joints, the splines, the differential.

Compliance. A driveshaft is a torsion spring. Under braking load it winds up; as pressure releases it unwinds. The two ends of that shaft are not doing the same thing at the same instant, and the discrepancy is largest precisely when the system most needs the truth: hard braking, with pressure modulating.

Backlash. Spline fit, joint articulation and differential lash all permit a small amount of free rotation. ABS pressure cycling drives the driveline back and forth across that lash. The sensor reports it faithfully, but it isn't wheel motion; it's the driveline taking up slack in one direction and then the other.

A sensor gap that won't sit still. This is the one that ends sessions. An inboard installation reads off a driveshaft that moves: the yokes deflect under load, wheelspin and surface changes shift things again, and the air gap between sensor and tone ring varies with all of it. Wheel speed sensing depends on that gap staying inside a narrow band. Vary it and the signal degrades, drops out, or returns a speed the system cannot reconcile against the other three wheels.

Measurement at the wheel: 909 outboard tone ring and sensor mount on an S-chassis rear hub, past the compliance, past the lash.

And then somebody fits a sequential. A dog-engagement gearbox doesn't ease torque in and out the way a synchro box does; every shift lands as a shock through the driveline, and it happens dozens of times a lap, including on downshifts in the middle of the braking zone where the ABS is working hardest. Each one of those hits deflects the shaft and moves the sensor with respect to its tone ring.

This is why inboard sensing so often appears to work and then stops. The car runs a season on a standard gearbox with nothing worse than the occasional flicker, the sequential goes in over the winter, and the ABS starts dropping out, with nobody connecting the two, because the gearbox isn't a brake component. It's the same sensor, the same bracket and the same tone ring; what changed is how violently the thing they're mounted to now moves.

Why that ends with the lamp on

A motorsport ABS is built to be suspicious of exactly that situation. When one wheel's signal stops making sense next to the others, the system doesn't guess and it doesn't average; it declares a sensor fault, lights the warning lamp and takes itself out of the loop.

Which means the failure doesn't present as slightly worse braking. It presents as no ABS at all, usually under the heaviest braking of the day, in a car whose driver has spent the session learning to brake as though the system is there. An ABS that drops out mid-session is worse than one you never fitted, because by then you've stopped driving around its absence.

The cost is timing, not noise

It's tempting to file all of this under signal quality: a wheel speed trace that looks a little noisier than it should, nothing serious. The consequence is worse than that.

Because compliance delays and smears the signal, the system's picture of the wheel lags the wheel itself. Release comes fractionally late. Reapply lands against a wheel that has already moved on. The careful phasing between what the system commands and what the tyre is doing is exactly what makes good ABS feel like nothing at all, and it's the first thing lost when the measurement point is wrong.

A system fed a late signal isn't a system with a small handicap. It's a system solving the wrong problem, slightly after the fact, at the limit.

"But the factory measured inboard on my car"

On some of the very platforms we convert, that's true, and it deserves an honest answer rather than a hand-wave. The R32 GT-R sensed its rear wheel speed inboard from the factory. A factory-ABS S14 or S15 carries a single tone ring on the differential input flange.

Those systems could measure there because of what they were. They were three-channel: each front wheel had its own channel, and the rear axle was treated as one. The system never tried to tell the left-rear from the right-rear; it measured the axle as a unit and controlled to that coarser picture, so the lash and wind-up between the two rear wheels was never a lie it had to reconcile. It all lived inside the one channel.

They also read the ring with variable-reluctance sensors, whose signal amplitude grows with speed. At the frequencies that matter on circuit, a VR signal is large enough to ride out a fair amount of gap movement; the amplitude covers what a modern defined-gap sensor would call a fault.

A modern four-channel motorsport ABS breaks both of those assumptions at once. It controls each rear wheel individually, so an axle-average measured at the diff flange cannot tell it what either tyre is doing. And it reads active sensors with a defined gap window; their output is the same strength at every speed, so there is no growing signal to mask a gap that has drifted too far.

So on these platforms, the outboard conversion isn't second-guessing the factory. It's updating a 1990s measuring point to the standard the new system was designed around: the same standard every current production car is built to, with encoder rings at the wheel bearing, the sensor fixed to the upright, and the gap set by the bearing for the life of the car.

Where inboard sensing shows up in modern conversions on four-channel systems, it usually survives for a simpler reason: it is dramatically easier to fabricate. A tone ring on a driveshaft flange and a bracket off the diff housing can be made in an afternoon; sensing at the hub means solving heat, debris, suspension movement and whatever bearing and upright combination the car is running. That's a fabrication convenience, not an engineering position, and it usually works right up until the moment it matters.

Measuring at the wheel

Relocating the tone ring and sensor outboard puts the measurement past the compliance and past the lash, with a gap that stays where it was set. The system sees the wheel, rather than the driveline's impression of the wheel.

The hardware that does it: 909 machined sensor mount, outboard tone ring and Bosch sensor for an R200 rear end.

It is the harder engineering, and that's the honest reason it isn't universal. The hardware has to survive heat and debris, tolerate full suspension movement, hold the sensor gap under load, and suit the specific hub, bearing and axle combination in front of it, which is why our kits are specified per platform and per hub package rather than sold as a universal bracket.

And it keeps the parking brake

A conversion that sacrifices the handbrake creates a different problem: a car that can't satisfy scrutineering where a working handbrake is required, can't be secured on a trailer, and in tarmac rally has lost a control the driver uses deliberately.

Our outboard sensing packages are engineered around the factory parking brake rather than in place of it, so it stays where it is and keeps working. For some builds that's a convenience. For a tarmac rally car it's the difference between eligible and not.

The short version

• ABS infers grip from wheel speed, so the signal's honesty sets the ceiling on how well the system can possibly work.
• Measuring inboard puts the shaft, joints, splines and diff between the sensor and the tyre: compliance, backlash and a moving sensor gap all end up in the measurement.
• The failure that ends sessions is the sensor gap: vary it enough and the system declares a fault, lights the lamp and switches itself off, usually under the hardest braking of the day.
• A sequential gearbox makes it markedly worse: dog engagement shocks the driveline dozens of times a lap, which is why inboard sensing often works fine until the year the sequential goes in.
• The real damage isn't a noisy signal, it's a late one: release and reapply arrive out of phase with what the tyre is doing.
• Some platforms measured inboard from the factory (the R32 GT-R and factory-ABS S-chassis among them) but as a single rear-axle channel read by forgiving VR sensors. A four-channel system with per-wheel control and defined-gap active sensors needs the measurement at the wheel: the conversion updates the old OEM measuring point to the standard current production cars are built to.
• Outboard relocation is harder to engineer and has to suit the specific hub, bearing and axle package, which is why it's specified per platform.
• Our outboard sensing retains the factory parking brake, which matters for scrutineering, trailering and tarmac rally.

Outboard sensing is part of our Nissan R-chassis GT-R kit and our S-chassis kit. More on the system itself in our Bosch Motorsport ABS guide: or talk to us about sensing for your platform.

909 Motorsport is an Authorized Bosch Motorsport Dealer in Newcastle NSW. We design and manufacture vehicle-specific wheel speed sensing for Bosch Motorsport ABS conversions (tone ring relocation, sensor mounting and hub-package-specific hardware) for circuit and tarmac rally builds.