A load cell is an electronic sensor or transducer for measuring force or weight.

It converts the force/weight applied into an electrical signal (milivolts, mV) that is proportional to the mechanical deformation caused by that force/weight.

It is mechatronic device that has an elastic body as a kind of mechanical spring and an internal electronic sensors of deformation or stress like strain gauges.

High accuracy class load cells are able to measure the force/weight with an extremely level of accuracy, typically 3000 parts of the full scale range (0.03% F.S.), which makes one of the best ways to mesure the quantity of materia or mass of the products. For example, we can find different load cells for measuring a mass up to 3kg in divisions of 1g, weighing a palet of 1500kg in fractions of 0.5kg or a truck up to 60.000kg in steps of 20kg or a silo of 150t in divisions of 50kg.

Typical applications:
- Scales
- Packaging Machinery
- Dosing and Filling Machinery
- Level/Inventory Control on Tanks and Silos
- Load limiting in Cranes and Lifting equipment
- Test Machinery
- Quality Control

Typical materials used to manufacture load cells are special alloys with high strength, high fatigue life, high repeatability, high linearity and low hysteresis. For example steels alloys, stainless steels, aluminium alloys, beryllium-copper and others.
A load cell basically consists of an elastic metal body (Element), an electrical circuit (Wheatstone Bridge) and a protective housing (Cover).

Strain gauges are stress sensors that are applied to the elastic metal body. The strain gauges each consist of a grid of thin metal wire (nickel-copper alloy called constantan) applied to a support of insulating material and carefully glued into specific areas inside the load cell that are designed to concentrate the mechanical stress and strain in that area.

As the force/weight is applied to the metal body, the strain gauges follow the deformations of the surface of the metal body to which they are bonded, increasing and decreasing in length as the metal body does. These dimensional changes create a variation in the electrical resistance (in Ohms or ?) of the strain gauge.

Inside the load cell is a "Wheatstone Bridge" electrical circuit, which magnifies the small changes in resistance of the strain gauges and generates an electrical signal. This is normally expressed in mV/V, millivolts per Volt of excitation supplied to the load cell.

Mass (kg) -> Force (N) -> Mechanical Deformation -> Electrical Resistance -> Wheatstone Bridge -> Electrical Signal (mV/V)

During load cell manufacture, the load cells are tested at different temperatures and then they receive small adjustments in resistance for calibration purposes and to compensate for thermal effects in the materials and components.
Below is a practical guide for the selection of a load cell. It has to be taken into account that there may be other technical circumstances or requirements, and it must take as an orientation that may be valid for most cases. This guide is only suitable for systems totally supported on load cells and systems with evenly distributed loads without great asymmetries and is not suitable for systems where the power is transmitted to cells by means of levers, systems with great asymmetries in the distribution of loads or systems with rolling loads. In order to choose or recommend a load cell, basically, the following questions should be answered:

1st: What load is going to be applied on the load cell.
2nd: What environment is it going to work in.
3rd: Other considerations.

The load to be applied on the cell will give an orientation about the Nominal Capacity of the cell necessary for the load cell. With this, we can restrict the number of possible models from which to choose. The working environment, together with other considerations and the Nominal Capacity will help us to choose the model.

Selection of the Nominal Capacity:

The aim is to estimate the real load on each supporting point in all operating circumstances and life of the system, including extreme situations, and choose a load cell with a suitable capacity and enough safety margins.

The capacity of a load cell is determined in the following way:
- Dead Load: Estimate the dead load of the structure, tank or silo, including all its elements: pipes, pumps, motors, agitators, insulators, heating fluids and accessories.
- Product Weight: The capacity and maximum range of the scales or the weight of the product must be known.
- Gross Weight: It is the addition of the Dead Load plus the Product Weight.
- Number of Supports N: It is the amount of supports on which the weighing structure, tank or scales is supported. It usually has from 3 to 6 supports.
- The theoretical load per support is the result of dividing the Gross Weight into the Number of Supports.
- Select a load cell with a nominal capacity higher than the theoretical load per support according to:

Cell Nominal Capacity = k x Gross Weight / N

Where k has a value between 1,25 and 2,2 , as safety coefficient to increase the capacity of the cells between 25% and 120% of the theoretical value, according to the presence of static or dynamic loads, vibrations, asymmetries, effect of the wind, impacts or rolling loads. A good choice for static loads in indoor tanks is to use k= 1,5 and round up a nominal capacity of a commercial cell.

Examples of common applications:
- 3 supports interior tank k = 1,3
- 4 supports interior tank k = 1,5
- Tank with agitation (moderate) k = 1,7
- 4 cell platform k = 1,8
- Bridge scales for weighing trucks of 6 or 8 cells k= 2

Note: When the Dead Load is over 50% of the Gross Weight, it is recommended to increase the safety margin to k=2, as it is usually due to large motors, accessories or heating systems and very probably, there exist non-centred and not uniform loads on the supporting points.
Note: Alter instalation it is important to check the load distribution for each bearing point. Generally, the load cells may be over-dimensioned up to over twice the weight of the product without any loss of accuracy. It is very common in scales and the only thing to bear in mind is that the sensitivity of the electronic indicator used or the micro-volts per division is enough.

Environmental issues:
It is very common that there exist various models of load cells of the same nominal capacity and so, the most suitable for the concrete environmental working conditions should be chosen:
- For corrosive environments or in presence of permanent humidity, it is recommended the stainless steel load cells, instead of aluminium or nickel plated steel.
- The degree of environmental protection increases with the choice of hermetically sealed load cells with a welded capsule.
- For potentially explosive environments, there also exist specific load cells.
- Verify the need of additional safety elements for those areas with special requirements against earthquakes or strong winds.

Final check:
Finally, try to answer the following questions and correct the nominal capacity of the cell if necessary:
- Is the value of the Dead Load exact?
- Can the load be distributed non?uniformly?
- Are there any agitations or impacts?
- Is it possible that the tank has a superior capacity and it may overflow exceeding thus the estimated Product Weight?
- Is there the possibility of strong winds or earthquakes in the area?
- Can a vehicle impact on or overload the system?
- Can you asure a good leveling for obtaining a good load distribution for each bearing point after the instalation?
There are load cells of the same shape appearance, but made of different materials, as steel or aluminium. Both may even have similar accuracy, repeatability and linearity characteristics but not the same mechanical resistance to overload, shocks or fatigue.

In order to reduce costs in manufacturing, different aluminium alloys may be used and they bear good results in regards to accuracy but they have the disadvantage that they are much weaker than the ones manufactured of steel alloy, in the sense that if certain stress levels are exceeded, the cells are more easily deformed and undergo displacements of the output signal. Therefore, they are also weaker in front of overloads and much more sensitive to shocks. Furthermore, they wear out more easily with dynamic loads and last less.

Therefore, in case the choice is an aluminium load cell, precautions should be increased in overload stoppers and choose load cell nominal capacities with a higher over?dimension in respect to the applied load than for a steel load cell.

The aluminium load cells are usually used in great consumption applications because they save costs in great series, where the scales designing team may have studied, project, over-dimension and test this solution properly in order to avoid problems.

For industrial weighing processes, of little production, it is safer and more reliable to directly use versions of cells made of high resistance steel alloys.

Also, to mention that there exist other materials with a very high resistance to fatigue, as the Beryllium-Copper, but very seldom used due to its high cost. Currently, they are only justified in high fatigue applications.
In order to define the metrologic characteristics of a load cell, in general, the standard developed by OIML (International Organization of Legal Metrology) is used, which in the case of a cell is the OIML R60 recommendation “Metrological regulation for load cells”.

According to those recommendations, the vmin parameter is the “Minimum load cell verification interval”, which is the data supplied by the manufacturer of the cell to indicate the recommended minimum size in order to define the size of each division or resolution of the load cell.

The vmin value is in units of mass (weight). Usually, vmin is a value between 6.000 and 10.000 fractions of the nominal capacity of the load cell.

This data can be used by the manufacturer of the scale to validate if the chosen division for the specific scale is compatible with the minimum division that the load cell can supply, according to the following formula:
vmin ?e / N or e? vmin * ? N

e = the verification scale interval or division of the scale
N = the amount of load cells in the scale
The precision that we may expect from a weighing system is the highest value obtained between the two following calculations (a) and (b):

(a) Limit by minimum division of the load cell (related to repeatability): emin(rep) = vmin * ? N -> (a)

emin(rep): minimum error that can be obtained by the minimum division of the cell
vmin: the minimum load cell verification interval
N: number of load cells

(b) Limit by range of use of the cell (related to linearity):
emin(lin) = Max / nlc -> (b)

emin(lin): minimum error that can be obtained by range of use of cell
Max: Net product weight
nlc: number of load cell verification intervals

Result: emin = el mayor de emin(rep) o emin(lin)

Recommendation: The precision is the error. The resolution or division of “display” is the fraction that is displayed. In certified scales, the resolution or division of display should not be finer than the error of the instrument itself. In certain industrial environments, an increased resolution is used, twice as fine than the real error or the precision of the system.

Example 1. Normal scale
Data of the weighing system:
Product Max= 600 kg
Dead Load DL = 120 kg
Total Load = 720 kg
Supports N = 3

Data of the Cells:
3 units Model 350i 500 kg
Emax = 500 kg
nlc= 3000 divisions
vmin = Emax / Y = 500 / 10.000 = 0,05 kg

(a) Limit by minimum division of cell (related to repeatability): emin(rep) = vmin * ? N = 0,05 * ? 3 = 0,086 kg
(b) Limit by range of use of cell (related to the linearity): emin(lin) = Max / nlc = 600 / 3000 = 0,2 kg

Result: emin = 0,2 kg

For these scales we shall choose a display resolution of d = 0,2 kg

Example 2. Scale with quite a lot of dead load in respect to the weight of the product
Data of the Weighing System:
Product Max= 400 kg
Dead Load DL = 320 kg
Total Load = 720 kg
Supports N = 3

Data of the Cells:
Model 350i 500 kg
Emax = 500 kg
nlc= 3000 divisions
vmin = Emax / Y = 500 / 10.000 = 0,05 kg


(a) Limit by minimum division of cell (related to the repeatability): emin(rep) = vmin * ? N = 0,05 * ? 3 = 0,0865 kg
(b) Limit by range of use of cell (related to linearity): emin(lin) = Max / nlc = 400 / 3000 = 0,133 kg

Result: emin = 0,133 kg
For this scale we shall choose a resolution of display of d = 0,2 kg in an environment of certified scales for commercial transactions or also the smaller d = 0,1 kg for an environment of industrial control.

Example 3. Scales with a great amount of dead load and very little product weight
Data of the Weighing System:
Product Max= 220 kg
Dead Load DL = 500 kg
Total Load = 720 kg
Supports N = 3

Data of the Cells:
Model 350i 500 kg
Emax = 500 kg
nlc = 3000 divisions
vmin = Emax / Y = 500 / 10.000 = 0,05 kg


(a) Limit by minimum division of cell (related to the repeatability): emin(rep) = vmin * ? N = 0,05 * ? 3 = 0,086 kg
(b) Limit by range of use of cell (related to the linearity): emin(lin) = Max / nlc = 220 / 3000 = 0,073 kg

Result: emin = 0,086 kg
For these scales we shall choose a display resolution of d = 0,1 kg.
The signal that one or several cells of a weighing system deliver for a specific increase of load, normally the display division, is:

?u = (C * 1000 * Uexc * e) / (N * Emax)

?u = Increase of signal in ?V/div (micro-volts/division)
C = Nominal Sensitivity of the cell in mV/V
Uexc = Excitation Voltage of the cells in V (Volts)
e = Size of the division in kg
N = Number of load cells
Emax = Nominal capacity of the cells

Typical values:
?u = 0,8 a 5 ?V/div (micro-volts/division)
C = 1 a 3 mV/V (milli-volts per volt)
Uexc = 3 a 12 V (Volts)
e = 0,001 kg a 100 kg
N = 1 a 10
Emax = 1 kg a 400.000 kg

Example 1. Scale of Max range=15 kg, division e=0,005 kg (5g)
N = 1 cell of Emax = 20 kg, C= 2mV/V
Load Cell Excitation Uexc = 10 Volts
?u = (2 * 1000 * 10 * 0,005) / (1 * 20 ) = 5 ?V/div

Example 2. Same example as Ex.. 1) but with Load Cell Excitation at Uexc = 5 Volts
?u = (2 * 1000 * 5 * 0,005) / (1 * 20 ) = 2,5 ?V/div

Example 3. Scale of Max range= 600 kg, division e=0,200 kg
N = 4 cells of Emax = 500 kg, C= 2mV/V
Load Cell Excitation Uexc = 5 Volts
?u = (2 * 1000 * 5 * 0,2) / (4 * 500 ) = 1 ?V/div

Example 4. Scale of Max range= 1500 kg, division e=0,5 kg
N = 4 cells of Emax = 750 kg, C= 2mV/V
Load Cell Excitation Uexc = 6 Volts
?u = (2 * 1000 * 6 * 0,5) / (4 * 750 ) = 2 ?V/div

Example 5. Same scale as Ex..4) with slightly bigger cells:
N = 4 cells of Emax= 1000 kg, C= 2mV/V
Load Cell Excitation Uexc = 6 Volts

?u = (2 * 1000 * 6 * 0,5) / (4 * 1000) = 1,5 ?V/div

Example 6. Truck weighing type scale of a Max Range of =60.000 kg, division e=20 kg
N = 6 cells of Emax = 20.000 kg, C= 2mV/V
Load Cell Excitation Uexc = 10 Volts

?u = (2 * 1000 * 10 * 20) / (6 * 20.000) = 3,3 ?V/div

Example 7. Truck weighing type scale of a maximum Range =60.000 kg, division e=20 kg
N = 8 cells of Emax = 30.000 kg, C= 2mV/V
Load Cell Excitation Uexc = 6 Volts

?u = (2 * 1000 * 6 * 20) / (8 * 30.000) = 1 ?V/div

Recommendation: As we have seen in the examples above, the signal values that the cells deliver per each “display” division are very small; between 1 and 2 ?V/div. Therefore, specific high sensitivity electronic instruments should be used for the load cells, that have super-stable power supply voltages, stable differential amplifiers, high resolution analog-digital converters of between 16 and 24 bits and filters and suitable protections.

The shielding of the conduit of the cell wires and a grounding of the complete system will help to protect these weak signals in environments with interferences, such as the industrial ones.
- Select the correct nominal capacity of the load cell, which should be higher than the maximum load operative in the installation. Do not load them over their nominal capacity.
- Select the Model of load cell suitable for the application and the environment.
- Be aware that a cell is a delicate sensor, both electrically and mechanically. Its choosing and installation should be carried out only by professionals of the sector. Take precautions for the safety of the system. Do not allow that the safety of people or things depends on the mechanical resistance of a cell or of the signals delivered by a cell. Properly overdimension and use the safety external elements that you consider necessary.
- Use the cell accessories designed by the manufacturer for the specific cell.
- Assemble the cells and/or accessories on a clean, flat, solid and strong surface.
- Design suitable protection elements against mechanical overloads, wiring protection, problems with rodents and any other risks.
- Avoid temperature gradients in the load cell. Temperature must be stable all through the body of the cell. If there is a heat source nearby, insulate it by means of insulating plates in order to reduce the transmission or radiation of heat towards any part of the cell.
- Do not hold a load cell by the wire or pull it. Properly protect the wires in the installation.
- Protect both the scales and the load cell from shocks.
- Returns must be properly packaged against shocks.
- Do not open the load cells or try to repair them.
- Do not carry out welding tasks near the load cells.
- Keep the area where the load cell is installed clean.
- Have a suitable drainage system in the installation in order to avoid floods for a long time. Both the load cells as the wires must never be submerged for long periods of time.
- Do not strain the cells or submit them to force torsion moments, different of the ones in the main direction of measuring.
- Use a stable and free of noise power supply. Do not feed it with a higher voltage than the recommended one and prevent the load cells from overloads and electrical discharges.
- Do not set the electronic equipment to a higher resolution than the logical one available for the cell or higher than the necessary for the user of the application.
- Do not exceed any specification limit of the load cell or its accessories, nor of the usual practices of the sector.
- The above recommendations are just an orientation as general information and they are not the only ones to take into account. The person in charge of the installation should analyze the needs of each concrete case.