How it works
The absorber housing sits on bearings so it can rotate slightly.
As the engine turns the rotor, the housing tries to follow, and a load cell on a torque arm measures that reaction. Torque is force times arm length, and power is torque times rotational speed.39
A water brake throws water outward with an impeller and returns it through a stator. Load rises with the amount of water inside, set by inlet and outlet valves. Water brakes are small and cheap for their capacity, but Wikipedia notes they are slower to stabilize than electric absorbers.Medium73
Eddy-current absorbers apply load through an adjustable magnetic field and respond quickly, but they cannot load an engine at zero speed or motor it. AC motor-generator dynos can absorb, motor and feed power back to the grid, at the highest cost and complexity.Medium83
Measuring at the crank removes drivetrain and tire variables. EngineLabs notes an engine dyno still has inertia to compensate for during sweeps, including the engine's own rotating mass.Medium43
The cell must supply what the car normally would: fuel, coolant through a heat exchanger, exhaust extraction, combustion air and ventilation. Nearly all absorbed power ends up as heat in the absorber water or air, so cooling capacity sets the cell's practical limit.Medium789
Typical values
How an ECU uses it
On an engine dyno the ECU runs as it would in the car, but inputs the car normally provides must be replaced: vehicle speed, pedal signals and any CAN messages from other modules. Missing inputs can set faults or limp modes that cap output.Low2
The dyno controller and the ECU log in parallel. Cell data such as torque, speed, fuel flow, airflow and temperatures combine with ECU timing, lambda and knock to build brake specific fuel consumption and MBT timing maps.Medium19
Steady-state control is the engine dyno's strength. The absorber holds speed while the operator sets throttle, so each cell of a fuel or ignition table can be measured at stable temperature.Medium3
How to test it
5 checks, in order.
- Check the coupling between engine and absorber: driveshaft guard in place, alignment within the maker's limits, and an adapter rated for the engine's torque.Low3
- Warm coolant and oil to their target temperatures under light load before any full-load test, and log both through every run.Low1
- Repeat a reference test at the start and end of each day. If corrected torque drifts more than run-to-run noise, investigate the sensors and engine before trusting new data.Low4
- SafetyKeep an emergency stop and extinguishing equipment within reach, run exhaust extraction, and stand clear of the plane of the driveshaft while the engine is loaded.Low3
Symptoms and causes
Torque reading drifts with no engine change
Likely causes Load cell zero drift, temperature effects or an out-of-date calibration
Check Re-zero, then repeat the dead-weight check.
Load control hunts or overshoots
Likely causes Slow water brake response, unstable water supply pressure or sticking valves
Check Check supply pressure and valve response with the engine at a fixed hold.
Cavitation noise or scale in the absorber
Likely causes Outlet water running outside the recommended temperature window
Check Log outlet water temperature against the 49-71 °C (120-160 °F) range.
Medium7
Output falls through a test session
Likely causes Rising coolant, oil or intake air temperature
Check Compare temperatures at the start of each run and restore them before comparing.
Medium1
In tuning
Engine dynos suit development
Comparing cams, intakes and exhausts, mapping fuel consumption and MBT timing, and building a base calibration before installation. The car's own intake routing, exhaust and cooling are absent, so the in-car calibration still needs checking.Medium41
Engine and chassis results are not interchangeable. Crank torque excludes drivetrain loss, and Wikipedia cites wheel figures typically 15 to 20 percent lower. Compare like with like.Medium3
Builders also use the engine dyno for a controlled run-in of fresh engines, with load, oil pressure and temperatures watched closely before full-power work.Low3
Formula and worked example
P = T × ω, with ω = 2π × N / 60
- P
- powerW
- T
- torque (load cell force × torque arm length)N·m
- ω
- angular speedrad/s
- N
- engine speedrpm
Try it
Result
Planning estimate from the formula below, not a measurement.
Open the full dyno correction factor calculator (sae j1349 and std) →
Common mistakes
Mistake Comparing engine dyno crank figures with chassis dyno wheel figures.
Do instead Compare crank with crank and wheel with wheel; drivetrain loss is not a fixed number.Medium36
Mistake Skipping dead-weight calibration checks.
Do instead Check with known weights before a test program and whenever readings look wrong.Medium9
Mistake Testing with a different intake and exhaust than the car will use.
Do instead Use installed-style parts; Wikipedia notes old SAE gross ratings ran higher partly because of test headers and missing accessories.Medium6
Myths vs facts
Myth An engine dyno measures horsepower directly.
Fact It measures torque through a load cell and speed through a sensor. Power is calculated from the two, as EngineLabs and Wikipedia both point out.43
Frequently asked
How does an engine dyno work?
The engine drives an absorber, such as a water brake or eddy-current unit, that resists rotation. The absorber housing tries to turn with the rotor, and a load cell on a torque arm measures that reaction force. The computer multiplies force by arm length for torque and combines it with speed for power.Medium39
Is an engine dyno more accurate than a chassis dyno?
It measures closer to the source. An engine dyno reads torque at the crank, removing drivetrain, tire and strapping variables, and can hold steady load points. A chassis dyno measures the whole car as driven. Each is accurate for its own measuring point.Medium43
What types of engine dyno absorbers are there?
Common types are water brakes, which are compact and cheap but slower to settle; eddy-current absorbers, which respond quickly but cannot motor the engine; and AC motor-generator units, which can absorb, motor and regenerate at the highest cost. Hydraulic and friction designs also exist.Medium378
Do engine dyno numbers include accessories?
It depends on the test basis. Wikipedia notes SAE net ratings under J1349 are measured with standard accessories, air cleaner and exhaust fitted, while older SAE gross ratings ran without accessories and sometimes with test headers. Always check how the engine was set up.Medium65
Why use an engine dyno instead of tuning in the car?
An engine dyno isolates the engine, so cam, intake and exhaust changes are measured without drivetrain or tire variables, and it holds load points with full instrumentation. It suits development and base calibration; the final calibration still needs checking in the car.Medium43
Wiring, pinouts and parts
Sources
Numbered as cited above. Hover or tap a citation chip to preview it.
9 sources
- Internal Combustion Engine Fundamentals, 2nd edition (2018) · McGraw-Hill (John B. Heywood) · book · retrieved 2026-10-27
- Automotive Handbook, 10th edition (2018) · Robert Bosch GmbH / Wiley · book · retrieved 2026-10-27
- Dynamometer · Wikipedia · reference · retrieved 2026-10-27
- The Data And Accuracy Implications Of Different Dynamometer Designs (Bradley Iger, May 2018) · EngineLabs · press · retrieved 2026-10-27
- SAE Adopts New Engine-Power Test (Bill Visnic, April 2005) · WardsAuto · press · retrieved 2026-10-27
- Horsepower · Wikipedia · reference · retrieved 2026-10-27
- Water brake · Wikipedia · reference · retrieved 2026-10-27
- Eddy current brake · Wikipedia · reference · retrieved 2026-10-27
- How An Engine Dyno Works (April 2019) · PowerNation · press · retrieved 2026-10-27

