How Dyno Testing Works

A dyno gives a controlled, repeatable environment to measure power and torque while logging everything else the engine is doing. Without one, 'it feels faster' is the only evidence available — and feelings lie, especially after you've spent money. With one, you get before/after curves, A/B comparisons between maps, and proof that the result holds across repeated runs.

A dyno also lets you test safely at full load without breaking the law or the car. The workshop controls load, can stop the pull at any sign of trouble, and can hold a steady RPM for diagnostics that no road can provide.

An engine dyno bolts to the crank — the purest measurement of the engine alone, but it ignores the drivetrain and requires the engine to come out of the car. A chassis (rolling-road) dyno measures at the wheels, which is what actually reaches the tarmac; losses from tyres, drivetrain, and inertia are included. A hub dyno bolts directly to the wheel hubs, removing tyre losses for more repeatable measurements.

Different dynos, different numbers — a mainline, a Dynojet, and a Mustang will all report slightly different figures on the same car. What matters is consistency on the same machine, not the absolute number. That's why we quote 'before on our dyno, after on our dyno' and log everything alongside the pull.

Air density changes power. On a hot, humid day an engine simply makes less power than on a cool, dry one — nothing is wrong with the engine. To compare runs fairly, dynos apply correction factors (SAE, DIN, STD) that standardise measurements to a reference atmosphere: typically 25°C, 100 kPa, and a set humidity.

Correction factors are useful and honest when used consistently — and abused when shops chase big corrected numbers from engines that would make far less in the real world. Always ask whether a figure is corrected or 'as measured', and what the air conditions were.

Ignore the peak number first. Look at the shape: where does torque peak, how flat is it, does it fall off early? A car with a fat, flat torque curve is faster and more enjoyable than one with a spikey peak that fades by redline. Smoothing also hides truth — a heavily smoothed curve can turn a knock-fighting timing mess into a pretty line.

The power curve only tells half the story. The useful graphs are the overlays: boost, AFR, ignition advance, and knock count on the same RPM axis as the power run. That's the data that shows whether a tune is safe or just impressive.

On every validation run we log MAP, IAT before and after the intercooler, AFR, lambda, ignition timing, knock counts, coolant and oil temperature, and fuel pressure. Repeated pulls back-to-back reveal heat soak — a tune that looks great on pull one and falls apart on pull three is not finished.

This is the documentation you get with every build: the graph, the logs, and the session report. A dyno sheet without logs is a poster, not proof.