How weighing scales are made

In a previous article, we discussed the different types of load cells. A load cell is a transducer that converts a mechanical force into a measurable electrical signal. Different load cells are designed to convert different mechanical forces into electrical voltage.

Load cells are typically classified by shape and form, with a unique configuration. Common types include canister load cells, s-beam load cells, single-ended beam load cells, double-ended shear beam load cells, planer load cells, single point load cells, button load cells, tension link load cells, disc load cells, load pin load cells, and multi-axial load cells.

Based on the principle of operation, all load cells fall into one of three categories: hydraulic load cells, pneumatic load cells, and strain gauge load cells.

Load cells are used in several applications, including weighing machines, dynamometers, force feedback in robotic systems, measurement of tension in cables and support beams at construction sites, performance evaluation of anchors and struts, measurement of forces in presses and assembly lines, industrial process control, measurement of pressure in wellbores, etc.

One of the most common applications is in weighing scales. All weighing scales — whether bench scales and conveyor belt scales used in industries, medical scales, hopper scales, or personal scales — are built using load cells.

This article covers the type of load cells used to construct digital weighing scales (medical/personal/platform). It also explains how a load cell’s internal circuit works and how they’re used in a weighing scale.

Load cells in weighing scales
The most common type of load cells used in weighing scales are strain gauge load cells. Pneumatic or hydraulic load cells are only rarely used, like when the scale must be explosion-proof.

As the name suggests, a strain gauge load cell consists of strain gauges. It’s a thin metallic foil bonded to a flexible substrate that acts as a variable resistor. The substrate is slightly deformed when force is applied to a strain gauge load cell. The body of a load cell is typically a metallic structure designed to deform predictably. The strain gauge is securely attached to the metallic structure of the load cell. The deformation stretches or compresses the strain gauge, changing its resistance. If the metallic foil is stretched, its resistance is increased. If the metallic foil is compressed, its resistance is decreased.

Depending on the mechanical force applied to it, a strain gauge may experience tension, compression, torque, bending, or shear strain. If a pulling force is applied, it undergoes tension. If a pushing force is applied, it undergoes compression.

The metallic foil that forms the strain gauge comprises a constant wire. This wire is arranged in a grid pattern to form the conductive foil, the active element of the strain gauge. The conductive foil is secured by a non-conductive polyamide backing foil that supports the conductive grid.

At the ends of the conductive grid are lead wires that allow the flow of current through the conductive grid. The strain gauge always has markings to aid in their precise placement in a load cell or onto a surface. The strain gauge, the load cell’s active element, is sensitive and the size of a thumbnail.

Using load cell technology, such a sensitive element is securely enclosed within the load cell body, ensuring the reliability of the technology.

The strain gauge load cells could be a shear beam, canister, bending beam, or single-point load cells. The shear beam load cells are commonly used in constructing platform weighing scales and bench scales as they can handle tension and compression. Canister load cells are preferred for heavy-duty applications because of their sealed design. Single-point load cells are used for single-point force measurements or minor scales. Bending beam load cells are used in low-capacity scales.

The Wheatstone bridge
The output voltage of a strain gauge is of minimal magnitude and typically ranges from millivolts. The sensitivity of a strain gauge is its maximum voltage output per excitation voltage. It’s expressed in mV/V. For example, if the sensitivity of a strain gauge is 0.5mV/V and an excitation voltage of 5V is applied to it, the maximum output voltage of the strain gauge could be 2.5 mV.

A strain gauge offers high sensitivity to small resistance changes. When a force is applied to it, there’s a minimal change in its resistance. This change in resistance could be so small that the strain gauge might not be detected or calibrated properly for a no-load condition. As a result, strain gauges are arranged in a Wheatstone bridge in most load cells.

A Wheatstone bridge is a unique circuit that measures unknown resistance precisely. It consists of four series resistors split in two parallel paths. The Wheatstone bridge circuit is shown in the below diagram.

When all four resistors are equal in resistance, the bridge is considered balanced. The current flowing through R1-R3 and R2-R4 are equal, so there’s no voltage difference between points A and B.

Ideally, in a balanced condition, there should be no voltage output between points A and B and no current flowing between them. However, if there’s a change in the resistance of any of the resistors, the bridge becomes unbalanced, and a potential difference is generated between points A and B. The voltage output between A and B is a function of resistance values (R1, R2, R3, and R4) and the excitation voltage (Vin).

The following equation determines the voltage output:
Vo = Vin((R1/(R1+R3))-(R2/(R2+R4)))

The strain gauge in a load cell is a variable resistor. A load cell is constructed by arranging at least one or more strain gauges in a Wheatstone bridge. A load cell may have all four resistors as strain gauges. Some designs use only two or one strain gauge and have other resistors as fixed resistors arranged in the Wheatstone bridge.

The Wheatstone bridge remains balanced until no force or weight is applied to the load cell. So, there may be a 0V output or a fixed output voltage from the Wheatstone bridge in the balanced condition. When a force or weight is applied to a load cell, the resistance of its strain gauges changes, and the Wheatstone bridge becomes unbalanced.

As a result, there’s a change in the output voltage of the load cell. Only a Wheatstone bridge can detect small changes in the resistance of strain gauges. Most load cells have four strain gauges arranged in it.

Temperature compensation is another reason for using the Wheatstone Bridge, as strain gauges are sensitive to temperature. By strategically arranging the strain gauges in the Wheatstone bridge, these arrangements can help adjust for temperature variations. The load cell’s non-load-bearing portion has one or two gauges installed. The temperature fluctuations the gauges go through are identical to those of the active gauges, but they typically don’t feel the additional force. Reducing the impact of temperature-induced resistance variations can enhance the precision of the weight measurement.

The Wheatstone bridge also eliminates unwanted strain forces that the load cell should not measure. This is done by pairing the strain gauges in the load cell so that opposite responses due to unwanted strain forces, which essentially cancel one another out. The arrangement allows for measuring only a specific type of force, such as tension or compression while increasing the reliability of the output signal.

Generally, the strain gauges in a load cell are arranged so that one pair of gauges experiences tension and the other pair experiences compression. Additionally, the Wheatstone bridge increases the resolution of the output signal. The output signal is maximized by using four gauges. If a single gauge might output 0.5mV/V, four gauges would output 2mV/V.

Load cell wiring
As a load cell typically has a Wheatstone bridge of strain gauges strategically arranged in the structure, there are four main wires in a load cell. A pair of these are excitation wires responsible for transmitting excitation voltage to the Wheatstone bridge of gauges in the load cell. The excitation voltage must be stable and precise for an accurate weight measurement. So, the power supply to load cells on a weighing scale must be stable and interruption-free.

One of the excitation wires is positive, and the other is marked negative. The other pair of wires are signal wires responsible for transmitting output signals due to mechanical strain on the load cell. The output signal is analog. While the excitation voltage is in volts, the load cell’s output signal is in the millivolts range. The signal wires are also designated as positive and negative. Specific color wires differentiate all four wires in a load cell cable.

Some load cells have six wires with two additional sense wires (excluding excitation and signal wires). The sense wires tap into the excitation voltage at a point within the load cell, accounting for the voltage drop caused by cable length. As the cable connecting the load cell to the weighing instrument gets longer, there’s a potential for some voltage to be lost due to cable resistance. This can affect the accuracy of the weight measurement, as the final signal received might be slightly weaker than the original signal produced by the load cell.

The sense wires in a six-wire load cell directly measure the voltage reaching the load cell, not the voltage that started at the source. This allows the weighing instrument to compensate for any voltage drop in the cable. It also provides a more accurate picture of the force being applied by measuring the voltage where it matters most. The six-wire load cells are slightly more costly than the four-wire load cells.

The load cell cable is usually shielded. The shield in a load cell cable is a defense against radio-frequency interference (RFI), which can skew the load cell’s analog signals. The shield provides a safe conduit for any detected interference to be deflected by grounding its end, often at the point of signal interpretation or data collecting. The interior wires are typically covered with a conductive covering made of aluminum foil or braided copper. By keeping noise from tainting the signal, this grounding is critical for preserving the precision and clarity of the measurements. In electrically loud industrial situations, the efficiency of this protective layer depends on the shield’s proper installation and grounding.

Load cell combinator
In most weighing scales, more than one load cell is arranged in a specific configuration to measure the weight. Usually, there are two or four load cells.

A device called a load cell combiner is an electronic circuit used to combine the signals from multiple load cells into a single output signal.

Each load cell generates its voltage signal based on the force it experiences. The combiner combines these individual signals from all four load cells into a single output signal. This combined signal represents the total force acting on the platform scale, the sum of the forces measured by each load cell.

Again, the most common method load cell combiners use is the Wheatstone bridge configuration. The combiner simplifies the wiring of the weighing scale and improves the overall resolution and accuracy of the output signal.

Signal conditioning
The output signal from the combinator is too weak to be read by a controller or analog-to-digital converter. So, the raw output signal from the load cell or combinator must be conditioned before transmission to a controller or ADC. Signal conditioning involves excitation voltage regulation, amplification, filtering, and linearization.

The signal conditioner often includes a regulated power supply that ensures the excitation voltage for load cells remains constant, even if the power supply feeding the system fluctuates. The signal conditioner amplifies a weak signal, so it’s stronger and suitable for further processing by the weighing instrument. A signal conditioner’s amplifier could be an operational amplifier or an instrumentation amplifier. If it is an operational amplifier, it has a closed feedback system.

Instrumentation amplifiers are more efficient in amplifying load cell signals. They offer low noise and drift at a low dc offset with a large common mode rejection ratio. The amplification gain must be carefully chosen to ensure the signal doesn’t become too strong and clip (distort) or introduce noise. In most signal conditioners, the amplification gain is programmable, though the gain value is still limited to a set of values, and usually only two.

Another critical step in signal conditioning is filtering. The raw signal from the load cell can be susceptible to various electrical noise sources, such as power line interference or environmental noise. These noises could appear as unwanted signals or frequencies in the output signal. The signal conditioner removes these noise signals using a low-pass filter.

The low pass filter lets the low-frequency analog output of the load cell/combinator pass while blocking the high-frequency noise. The signal conditioner also performs linearization to correct non-linearity, ensuring the output signal accurately reflects the weight across the entire measurement range. After signal conditioning, the output signal is passed to a controller or ADC to convert the analog output from load cells into a digital reading.

Junction box
The load cells, along with the combinator, signal conditioner, controller, and display module, can be assembled in a junction box to build the weighing scale. A junction box facilitates a safe and organized enclosure for connecting multiple load cell cables. It helps protect the wiring from damage and environmental factors, and may provide terminal strips for easy connection and disconnection of load cells.

The junction box can connect the load cell cables in a large industrial weighing system, which is how a weighing scale is constructed.

Calibration
The complete weighing system must be calibrated regularly to provide accurate weight readings. To ensure the weight displayed on the scale corresponds to the actual weight applied, calibration entails applying known weights to the scale and modifying the system’s parameters.