Verifying a Brake Tester: Reference Force, Drag and Zero
A brake tester can pass calibration and still read wrong. Calibration proves the sensor and electronics against a traceable reference at one moment; the free-roller drag and zero checks catch the mechanical wear that accumulates between calibrations. Both belong in the operating routine, and neither replaces the other.
Calibration and verification are different jobs
Calibration is a metrology exercise: apply a known force, compare it with the displayed value, adjust if the difference exceeds the permitted error, and issue a certificate traceable to a national standard. It is done by an accredited body at an interval the authority sets.
Verification is what the station does itself, more often, to confirm nothing has changed: zero with no vehicle on the rollers, a reference check against a known load, and a look at the free-running drag. A station that only calibrates twice a year and never verifies in between has two data points and no idea what happened in the other fifty weeks.
The reference force method
A roller brake tester is checked by applying a known force to the roller or to a calibration arm and comparing the reading. The reference device — weights, a lever with certified weights, or a force transducer — must itself be traceable, and its certificate must cover the range being checked.
The accuracy the tester has to hold is stated in ISO 21069-1: below 5 000 N the permitted error is within ±100 N, and above that it is within ±2 % of the measured value. That is a demanding band at the low end, which is why a machine that reads heavy trucks well can still be marginal on a light car axle — and why the verification should be done across the range the station actually tests, not only at one convenient point.
Zero and the free-roller drag check
The zero point must be settable with no vehicle on the rollers, because the tester has to distinguish brake force from the resistance of its own drive train. If zero drifts, every reading on the lane carries the same offset and every vehicle looks slightly worse or slightly better than it is.
The drag check exposes the mechanical side. With the machine in calibration mode and no vehicle present, the rollers are run and the displayed brake force is read: what appears is the friction of bearings, chain or belt and gearbox. Some specifications put a ceiling on this — the UK DVSA roller brake tester specification states that the displayed brake force must not exceed 50 kgf in that condition. A machine that creeps above its ceiling has a worn bearing or a tightening drive, which a static force calibration will not reveal because it never turns the rollers.
Interval, certificate and the out-of-calibration state
The interval is set by the authority, not by the supplier. Where the DVSA specification applies, a roller brake tester is calibrated every six months, to the end of the calendar month in which it expires. Other markets set their own period; the common thread is that a result produced by an out-of-calibration machine is not defensible.
That makes the out-of-calibration state a design requirement rather than an administrative one. The machine should tell the operator, on screen and on the printout, that calibration has lapsed — a warning the inspector cannot miss and cannot switch off. Keep the certificates per machine, with the serial number on them, because a certificate that does not identify the machine proves nothing about the reading it produced.
- Reference device certificate: traceable, covering the range tested.
- Calibration certificate: per machine, by serial number, within the authority interval.
- Zero check: with no vehicle on the rollers, before the first test of the day.
- Drag check: free-roller force compared with the ceiling in the applicable specification.
Common questions
How often does a roller brake tester need calibration?
The interval is set by the national authority. Where the UK DVSA roller brake tester specification applies it is every six months, to the end of the calendar month in which it expires. Other markets set their own period, and a result produced outside it is not defensible.
What accuracy must a brake tester hold?
ISO 21069-1 sets the measurement accuracy: within ±100 N below 5 000 N, and within ±2 % of the measured value above it. Because the low end is the tighter band in relative terms, verification should cover the light-vehicle range as well as the heavy end.
Why check free-roller drag if the machine is calibrated?
Calibration applies a static reference force and never turns the rollers, so it cannot detect bearing, chain or gearbox friction. Running the rollers with no vehicle shows that friction as a displayed force, and the applicable specification sets a ceiling for it.
Need the equipment side of this?
Tell us the vehicle classes and the volume you need to handle, and we will map these options onto a station list with the specifications for each unit.