The road, indoors, held to ±0.05 km/h.

They answer different questions. An engine dynamometer couples directly to the crankshaft or output shaft of an engine on a stand — no vehicle, no gearbox, no tyres — and is the right tool for developing or accepting the engine itself. A chassis dynamometer tests the whole vehicle, through its own driveline and tyres, exactly as the road does; it is the machine for emissions cycles, range determination, calibration work and anything where the vehicle’s complete behaviour matters. The rule of thumb: if the thing you are signing off has wheels, it belongs on a chassis dyno; if it arrives on a pallet, it belongs on an engine dyno. Neometrix covers both — this page is the chassis machine, and the engine test rig has its own page.

Because a tyre does not meet a small roller the way it meets a road. On a flat road the contact patch is flat; on a curved roll it wraps, and the smaller the roll the sharper the wrap — changing rolling resistance, contact pressure and tyre heating, and doing so differently for every tyre size. A 48-inch (1,219 mm) roll is large enough that the patch geometry approaches the flat-road case, which is why this diameter became the reference architecture for emissions-grade and high-speed work. It also gives the roll enough circumference to hold its chrome running surface within 0.25 mm of true — at 250 km/h the tyre reads every deviation, and balance to ISO 1940 G2.5 is what keeps a two-axle, thirty-tonne machine quiet at that speed.

Flywheels create inertia honestly but coarsely: each disc adds a fixed lump, so a flywheel dyno can only approximate a vehicle’s mass in steps, and changing configuration means mechanically engaging discs. Electrical simulation lets the drives create the force a given mass would exert during every acceleration and braking event — here across roughly 150–5,500 kg in 0.454 kg increments, so the simulated vehicle can match the real one to the kilogram, spare wheel included. The catch is honesty under transients: mass reacts instantly, control loops do not. That is why the <65 ms system response is a headline figure — fast enough that a gearshift, a launch or a regenerative braking event feels like mass, not like software catching up.

Because on the road, all four wheels roll on the same surface, and a modern vehicle continuously checks that they agree. On a 4WD dynamometer the front and rear axles stand on separate roll sets driven by separate machines — and if those two “roads” drift apart by more than a whisker, the vehicle’s wheel-speed comparison concludes something is slipping. Traction control intervenes, torque split changes, ABS may trigger, and the test is no longer measuring the vehicle — it is measuring the vehicle fighting the machine. Holding both axles to within ±0.05 km/h of each other keeps every on-board system convinced it is on one continuous road, which is the entire premise of the measurement.

Yes — and this is exactly where older dynamometers fall short. An EV under regenerative braking turns its motor into a generator and pushes torque back through the wheels; a dyno built only to absorb sees that as a disturbance, but a four-quadrant machine treats it as a normal operating point — the drives motor and absorb interchangeably, so the full cycle, launch to regen stop, is applied faithfully. Range determination then runs the standardised driving cycle repeatedly — at 0.454 kg inertia fidelity and ±1 kgf road-load accuracy, in a climatic cell if the standard demands temperature, with a driving robot removing driver-to-driver variation that would otherwise swamp the percent-level differences being measured.

We integrate them, and it is worth being precise about the split. AC flux-vector motors, IGBT drive units, precision load cells and encoders are proprietary products from established manufacturers — the right ones are selected, not reinvented. What Neometrix engineers and builds is the dynamometer as a working system: the roll sets and their balancing, the mechanical build in the pit, the road-load and inertia-simulation control scheme, the measurement chain and its calibration, the vehicle interface and restraint, the cell services around the machine, and the installation, commissioning and acceptance testing that turn all of it into a defensible test result. On upgradation work the split is the same — the existing pit and rolls stay, and the electronics, measurement and control brain are brought to current standard. And to be straightforward: this configuration is engineered to order and has not yet been built, so nothing on this page is offered as a delivered reference.