WEIGHING SYSTEMS AND ERRORS: A CLOSER LOOK

As you may know, every type of measurement relies on a system of interconnected devices designed specifically to capture and record data. Each system has a defined level of measurement accuracy, which includes a total error, an aggregate of various individual errors resulting from both the devices themselves and the conditions under which they are installed. In simple terms, the smaller the total error, the more accurate the measurement. Whether you’re troubleshooting a weighing CAL error or ensuring precision in daily operations, understanding error types is crucial. Let’s take a closer look at the types of errors that can affect weighing systems.

TYPES OF ERRORS IN WEIGHING SYSTEMS

The first distinction to make is between systematic errors and random errors:

Systematic errors are predictable, quantifiable, and can be summed in
Random errors, on the other hand, are statistical in nature and vary unpredictably from one measurement to another.

To minimize overall error, it is essential to define two key performance indicators for the system:

Error at constant room temperature: this serves as a reference value for verification tests once the system is operational. It is calculated by combining the systematic error with the total random error. Any zeroed errors caused by system design or environmental conditions should be excluded from the calculation.
Error due to thermal drift: this long-term error emerges from temperature fluctuations and can be mitigated through periodic calibration. The frequency of calibration depends on the user’s operating conditions and requirements.

SYSTEMATIC ERRORS

As mentioned earlier, systematic errors are predictable and can be summed algebraically in absolute terms to form the total systematic error. In measurement devices, such errors are typically expressed as a percentage relative to the full scale. In multi-cell systems, these percentage values remain consistent across all cells.

The primary types of systematic errors include:

Cell rated output tolerance: this applies specifically to systems with multiple load cells and is influenced by shifts in the load’s center of gravity. it can be compensated through corner calibration. Without compensation, the error can be calculated using the following formula:

Error= (rated output tolerance / 100) * (barycentre shift % / 100) * full scale capacity

Linearity error: this refers to deviations from the ideal measurement curve and must be considered not only for increasing loads but also during decreasing load conditions (downward linearity).
Hysteresis error: this occurs when the measured value depends on whether the load is increasing or decreasing. It should be considered only when the load changes during the weighing process.
Drift under load: this type of error arises when there is a delay between the application of the load and the measurement. It’s relevant in systems where readings are taken some time after the load is applied.
Measurement resolution: this represents the smallest detectable change in weight and must be considered in all applications.

RANDOM ERRORS

Total random error is calculated as the square root of the sum of the squares of individual random errors. Due to their inherent unpredictability, random errors must always be taken into account in every application, as they can’t be compensated or corrected. The calculation is based on absolute values and refers to the system’s full scale.

Key types of random errors include:

Load cell repeatability error: defined as the standard deviation of 10 repeated measurements using a weight that varies approximately 50% of the rated capacity. This value is expressed as a percentage relative to the weight variation. When calculating this error, consider the actual weight variation that occurs in the system, as it may not correspond exactly to the full scale capacity.
Stability of load cell signal booster: this value is often not readily available and should ideally be separated into two components for clarity: stability with respect to supply voltage fluctuations; long-term stability over time.
Stability of the A/D converter (also referred to as “noise” by manufacturers): this can be expressed either as a standard deviation or as a peak-to-peak value. In most cases, it is negligible because the peak-to-peak variation is usually smaller than the system’s resolution.

ERRORS DUE TO THERMAL DRIFTS

Like systematic errors, thermal drifts errors are calculated by summing individual components algebraically. Since environmental operating conditions differ from one system to another, the error must be determined based on the actual temperature variation relative to the calibration reference temperature.

Over the long term, fluctuations in ambient temperature can be mitigated through seasonal recalibrations, thereby reducing the impact of thermal drift on measurement accuracy.

Thermal drift errors primarily result in:

Zero errors: these can be disregarded if the system features automatic zero adjustment prior to measurement.
Range errors: these cannot be eliminated and must therefore be accounted for in all applications.

The following devices and components can introduce thermal drift errors:

Load cells
Signal boosters (load cell amplifiers)
A/D converters
Load cell power cables
Structural thermal expansion (to be considered in large, heavy, or long structures, or in cases of significant temperature differences)
Power supply to load cells (negligible in modern systems with microprocessor control, where this error is virtually eliminated)

NEED EXPERT SUPPORT WITH YOUR WEIGHING SYSTEM?

Understanding and managing measurement errors, like a weighing CAL error, is essential for accurate, reliable performance. If you’re looking to optimize your weighing processes or need tailored solutions for industrial or commercial applications, Celmi is here to help. With decades of experience and a wide range of precision products, we’re ready to guide you toward the most effective solution.