A load cell datasheet is a short document, but almost every line on it decides something about the scale you are going to build. This is what each figure means and what it costs you to get it wrong.
Rated capacity
Unit: kgf or tf. The maximum load the cell is built to measure. Load it beyond that and the reading stops being accurate; do it repeatedly and the cell is damaged permanently.
Rated capacity is per cell, not per scale. A bridge on eight 30 tf cells has 240 tf of cell capacity but is still specified and used at a much lower figure — 120 tf is typical.
Accuracy class
Class II or class III — the grading of how accurately the instrument works. Assessed against international standards such as OIML or NTEP, and it is the class rather than the manufacturer’s own wording that matters when the scale has to be verified for trade.
Resolution
The number of divisions the cell can resolve across its range — typically n = 1,000 to 100,000, quoted as 1/6,000, 1/10,000 or 1/30,000. A 30 tf cell at 1/3,000 resolves to 10 kgf steps; the same cell at 1/30,000 resolves to 1 kgf.
The higher this figure, the more the cell costs.
Rated output
Unit: mV/V. How much signal the cell returns per volt of excitation — its sensitivity.
On an analog cell the output depends on the excitation supplied by the indicator. Take a 10 tf cell rated 2 mV/V with 10 tf applied: at 10 V excitation the output is 20 mV; at 8 V excitation the same load gives only 16 mV.
The matching figure on the indicator is input sensitivity in µV/d. The smaller that number, the finer the display can resolve.
The error figures
These are all quoted as a percentage of full scale, and they are the numbers that decide whether a cell will pass verification at the class you need.
| Specification | What it measures |
| Combined error | The algebraic sum of non-linearity and hysteresis — the single figure that best summarises the cell. |
| Repeatability | Spread between readings when the same load is applied repeatedly, under the same conditions, from the same direction. |
| Creep | How far the output drifts while a constant load is held, typically over 30 minutes, at constant temperature. This is what makes a tank scale read differently in the afternoon than it did at the start of the shift. |
| Zero return | The difference in the zero reading measured before loading and again after the load is removed. |
The two overload limits
Both are quoted as a percentage of rated capacity, and they mean very different things.
Safe overload
The load the cell can take without permanent change to its characteristics. Exceed the rated capacity but stay under this figure and the cell recovers — commonly 150%.
Ultimate overload
The point at which the cell is damaged — readings become unreliable or the cell fails mechanically. Often above 300%. Nothing above safe overload should be treated as usable.
A cell that has been through an overload event often still works — it simply no longer holds its zero, or its corner behaviour has shifted. That is why an unexplained corner error after a heavy vehicle is worth investigating rather than adjusting away.
Specification note
KALA Technical Notes
Datasheets are easy to compare line by line and hard to compare in practice, because a figure that matters on a tank scale may be irrelevant on a weighbridge. Send us the application and the division you need, and we will tell you which specifications actually constrain the choice.
● Load cells with full datasheets and OIML certificates on request.