A load cell is loaded with 5,000 N. After the load has been completely removed, however, the evaluation system does not return exactly to 0 N, but instead shows, for example, +18 N. A few minutes later, the value slowly decreases to +7 N and remains there.
Does this mean that the load cell is defective? Not necessarily. A zero point that does not return completely can have several different causes that must be distinguished from one another from a measurement perspective. These include the normal time-dependent behavior of the sensor as well as mechanical stress, transverse forces, friction, cable forces, temperature changes, electrical drift or a previous overload.
The behavior over time is therefore particularly important. A signal that slowly returns toward its original zero point after a prolonged load behaves differently from a value that immediately remains at a new fixed level after unloading. Temperature-dependent zero drift must likewise be distinguished from hysteresis, which only becomes visible when increasing and decreasing loads are compared.
In real machines, the load cell itself is also frequently not the main cause. The sensor is part of a mechanical force path. Even slightly misaligned force application, friction in a guide, an overly stiff hose or a connection cable under tension can cause a small residual force to remain on the sensor after unloading.
Electrical zeroing can remove such an effect from the displayed value, but it does not eliminate the underlying cause. Particularly after a possible overload, simply re-taring the system can even be problematic because permanent deformation of the spring element or a changed characteristic could be concealed.
For reliable diagnostics, the complete signal and force path should therefore always be considered: from the actual force application through the load cell and connection cable to the measuring amplifier and PLC. A comparison between the installed condition and a defined reference setup is particularly informative.
The most important diagnostic question is therefore not only: “How large is the zero-point error?”, but rather: “When does the deviation occur, how does it change during loading, unloading and the subsequent waiting period – and does it remain when mechanical influences are eliminated?”
Why Does a Load Cell Normally Return to Zero?
A conventional strain-gauge load cell has an elastic spring element. When a force is applied, this element deforms slightly. Strain gauges bonded to the spring element detect this mechanical strain as a change in electrical resistance. A Wheatstone bridge converts this into an electrical output signal, typically specified in mV/V for passive load cells.
As long as the spring element remains within its permissible elastic range, it should largely return to its original mechanical condition after the force is removed. Accordingly, the output signal should also return close to its original zero value.
In a real sensor, however, this return is never mathematically perfect. The spring element, strain gauges, adhesive and bonding layers as well as temperature and mounting influences cause small deviations. This is precisely why high-quality load cells have specified parameters such as hysteresis, creep, repeatability and temperature effect on zero.
The magnitude of the deviation is important. A small deviation within specification is part of normal measurement behavior. A suddenly large or non-repeatable zero-return error, on the other hand, may indicate a problem with the sensor, installation or complete measuring chain.
Distinguishing Creep, Hysteresis and Zero Return
In practice, very different effects are often grouped together under the general term “zero drift.” For systematic troubleshooting, however, this is too imprecise. The decisive factors are when the error occurs and how the signal subsequently behaves.
| Effect | Typical Signal Behavior | When Does It Become Visible? | Typical Diagnostic Approach |
|---|---|---|---|
| Creep | signal changes slowly despite constant load | while the load remains applied | hold load constant and observe the signal over time |
| Creep recovery | signal slowly returns toward the original zero after unloading | after prolonged loading | record the zero point after defined waiting periods |
| Hysteresis | the same force produces different readings during loading and unloading | when comparing increasing and decreasing load sequences | compare identical intermediate load points |
| Mechanical stress | zero value depends on mounting position or mechanical condition | often immediately after a load change | check bolts, mounting surfaces and force application |
| Temperature drift | zero point changes with sensor or ambient temperature | during thermal changes | monitor temperature and zero point simultaneously |
| Overload damage | new permanent zero point or changed characteristic | after an exceptional load | check the complete sensor characteristic |
This distinction is important because the corrective actions are completely different. Mechanical stress cannot be corrected by performing a new electrical zero adjustment. Normal creep recovery, on the other hand, is not automatically an indication of permanent load-cell damage.
What Is Creep in a Load Cell?
Creep describes the time-dependent change in the output signal while the applied load remains constant. The external force does not change, but the indicated value slowly changes over time.
For example, a load cell is loaded with 10,000 N. Immediately after the load is applied, the evaluation system shows 10,002 N. The force is then held constant for 30 minutes. After this period, the system shows, for example, 10,009 N.
The mechanical load has not changed, but the electrical output signal has. This time-dependent difference is referred to as creep.
The effect results from the time-dependent behavior of the complete elastic measuring system, including the spring element, strain gauges and their connection to the base material. The permissible magnitude is normally specified in the data sheet for a defined load and observation period.
Creep is evaluated while a constant load is still applied. A residual value after complete unloading must therefore not automatically be described as creep.
What Does Creep Recovery Mean After Unloading?
After prolonged loading, the signal can continue to change over time even after the force has been completely removed. This effect is known as Creep Recovery.
Before loading, the load cell may indicate 0 N. Immediately after a prolonged loading phase and complete unloading, the display initially shows +20 N. After several minutes, the value falls to +9 N and later to +3 N.
From a measurement perspective, this behavior must be assessed very differently from a zero point that immediately jumps to +20 N after unloading and then remains practically constant over a long period.
Residual values after unloading should therefore always be documented together with the corresponding time. The statement “zero-point error 20 N” is far less useful than the information “20 N immediately after unloading, 9 N after 5 minutes and 3 N after 30 minutes.”
What Is Hysteresis?
Hysteresis describes the difference in output signal at the same force depending on whether that load point is approached from a lower or a higher load.
For a load cell with a measuring range of 0...10 kN, for example, a value of 5.001 kN may be indicated at 5 kN during the increasing load sequence. The load is then increased to 10 kN and subsequently reduced again. When the 5 kN point is reached for the second time, the system now indicates 5.008 kN.
The difference between these two readings is part of the hysteresis behavior of the complete measuring point. However, it must be determined whether the difference actually originates from the sensor or whether it is caused by friction, mechanical stress or non-repeatable force application.
A meaningful hysteresis test therefore always requires an increasing and a decreasing load sequence with several identical intermediate points. A single zero value after one loading cycle is not sufficient.
Mechanical Stress as a Common Cause
In machines and test rigs, the cause of poor zero return is often outside the actual load cell. The sensor is part of a mechanical force path. If this force path is not fully reversible during loading and unloading, a small force or moment may remain on the sensor after unloading.
Typical causes include uneven mounting surfaces, misaligned force introduction components, preloaded bolted connections, missing joints, deformation of the machine frame or lateral contact with stops and guides.
Targeted mechanical intervention can provide particularly useful diagnostic information. If the measured value changes significantly when a mounting bolt is slightly loosened or the force application assembly is moved, this strongly indicates a mechanical cause.
Performing a new electrical zero adjustment in this situation would merely compensate for the currently acting residual force. During the next loading cycle, the mechanics may settle differently again and create a new zero-point error.
Detecting Transverse Forces and Off-Center Loading
Force transducers are designed for a defined primary measuring direction. Additional transverse forces, side loads, bending and torsional moments or off-center force application can influence both the output signal and the mechanical return behavior.
A design in which the mechanical assembly shifts slightly under load is particularly critical. During unloading, friction or mechanical play may prevent it from returning to exactly the same position. A small side force or residual moment can then continue to act on the force transducer.
On the display, this effect may appear similar to sensor hysteresis or poor zero return even though the actual cause lies in the force application system.
In precision test rigs, unavoidable minor alignment errors should therefore be compensated using suitable force introduction components, rod ends, load buttons or other components intended for the sensor.
Eliminating Friction and Force Shunts
In force or weight measurement, the force to be measured should pass as completely as possible through the intended sensor. An unintended additional force path is referred to as a force shunt.
In a vessel weighing system, for example, a rigid pipe can carry part of the vessel weight. On a weighing platform, a side guide may rub against the frame. In a test rig, a stop may remain in slight contact after the load has been removed.
Friction is particularly critical. A rubbing connection can move in one direction during loading and remain in a slightly different position during unloading because of static friction. The force distribution after the cycle is then no longer identical to the initial condition.
The result can be poor zero return, apparent hysteresis and insufficient repeatability. The higher the required measurement accuracy, the more important a low-friction and reproducible mechanical force path becomes.
Why Even the Connection Cable Can Influence the Zero Point
An electrical connection cable also has mechanical stiffness. In simplified terms, it behaves like a small spring that can transfer an additional force to the sensor or moving structure.
With large measuring ranges, this force is often negligible. With small load cells or high-resolution force measurements, however, it can become clearly visible. Cables under tension, a sharp bend immediately at the sensor outlet or cable routing that moves during the loading cycle are particularly unfavorable.
A simple test is to completely unload the sensor and then carefully change the cable position. If the measured value changes reproducibly with the cable movement, mechanical feedback is present.
The cable should therefore be routed without tension, with a suitable bending radius and in such a way that its position remains as constant as possible during repeated loading cycles.
Temperature Change or Mechanical Fault?
Temperature changes can also influence the zero point of a strain-gauge load cell. Force transducers therefore have data-sheet specifications for the temperature effect on zero and often also on sensitivity.
A typical situation occurs when a cold sensor is brought into a warm test laboratory and operated immediately. Motors, hydraulic power units or solar radiation can also change the temperature of the transducer and its mechanical surroundings.
A certain amount of thermal stabilization can also take place after the bridge excitation voltage is switched on. For precise measurements, the sensor and electronics should therefore be operated under stable conditions for a sufficient period before the zero point is evaluated.
A typical indication of temperature drift is that the zero point slowly changes in parallel with the sensor or ambient temperature even without a preceding loading cycle.
Permanent Zero Shift After Overload
A much more critical cause is mechanical overload. If the permissible load range of the spring element is exceeded, it can leave the purely elastic range and undergo permanent plastic deformation.
After unloading, the sensor then no longer returns fully to its original mechanical condition. Typical indications can include a suddenly and significantly changed zero point, altered sensitivity, increased hysteresis or poorer linearity.
Overload does not only mean excessive axial force. Impact loads, transverse forces, bending moments or misalignment during installation can also place excessive stress on the spring element.
After such an event, simply performing a new electrical zero adjustment is not an adequate test. Although the offset may disappear from the display, changes in sensitivity, linearity and structural integrity can remain.
Checking the Electrical Measuring Chain Separately
Before removing or replacing a load cell, it should be determined whether the deviation actually originates from the sensor. The complete signal path may consist, for example, of load cell → connection cable → junction box → measuring amplifier → PLC / display.
Unstable bridge excitation, contact resistance, moisture in junction boxes, damaged cables, poor connectors, EMC interference or drift in the measuring amplifier can also cause a drifting zero point.
Comparing different signal levels is particularly useful. If the raw signal at the measuring amplifier input remains stable while the value displayed by the PLC changes, the cause is probably located downstream of the sensor.
For passive strain-gauge load cells, a suitable mV/V simulator can also be used. This allows the measuring amplifier and evaluation chain to be tested independently of the mechanical load cell.
Testing the Load Cell Outside the Machine
One of the most informative methods is to remove the sensor from the problematic mechanical assembly and test it separately under defined conditions.
This largely eliminates influences such as frame stress, pipe forces, guide friction, mechanical stops, machine movement or insufficiently flexible process connections.
If the load cell shows good zero return, repeatable load values and a plausible increasing and decreasing characteristic in the reference setup, this strongly suggests that the cause lies in the original mechanical installation.
If the deviation remains under controlled conditions, however, a sensor fault, overload damage or changed calibration becomes more likely.
A Simple Loading and Unloading Test
For a systematic assessment, the sensor should first be thermally stabilized and then completely mechanically unloaded. The unloaded output value is documented as the reference. The force is then increased in several defined steps, for example 0 → 25 → 50 → 75 → 100 %.
At the upper load point, the force is held constant for a defined period. This makes it possible to determine whether the signal changes slowly despite an unchanged load. This part of the test is used to assess creep behavior.
The load is then reduced through the same intermediate points, for example 100 → 75 → 50 → 25 → 0 %. Comparing identical force points between the increasing and decreasing sequences provides information about hysteresis and mechanical repeatability.
After complete unloading, the zero point is recorded immediately and then again after defined waiting periods, for example after 1, 5, 15 and 30 minutes. This shows whether creep recovery is taking place or whether a largely constant residual offset remains.
The complete cycle should then be repeated. Only repeated measurements reveal whether the behavior is reproducible or whether the zero point changes differently after each cycle.
Interpreting the Signal Over Time Correctly
| Observation | Likely Direction for Troubleshooting | Recommended Next Check |
|---|---|---|
| Signal drifts under constant load | creep, temperature or electrical stabilization | keep load and temperature constant |
| Signal slowly returns toward zero after unloading | creep recovery | document waiting times and residual values |
| Increasing and decreasing values differ | hysteresis, friction or mechanical stress | inspect mechanics and force application |
| Zero point changes when the cable is moved | mechanical cable force | optimize cable routing |
| Zero point jumps when a bolt is loosened | mechanical stress | check mounting surfaces and installation |
| Zero point changes with temperature without a load change | thermal influence | record sensor and ambient temperature |
| Large permanent shift immediately after an exceptional load | possible overload or plastic deformation | test or calibrate sensor separately |
| Sensor operates correctly outside the machine | fault in the mechanical installation | check force application, friction and force shunts |
This classification does not replace a complete diagnosis. However, it provides considerable help in selecting the next test steps and avoiding unnecessary sensor replacement.
Practical Example: Force Sensor Remains at +120 N After a 20 kN Load
In an assembly test rig, a compressive force is measured using a force transducer. The measuring range is 0...25 kN. After a loading cycle up to 20 kN, the system still shows +120 N after complete unloading.
First, the signal is observed over time. The value remains largely constant for ten minutes. Creep recovery alone is therefore less likely to be the main cause.
The connection cable is then moved carefully. The indicated zero point changes by only a few newtons. The mechanical influence of the cable therefore cannot explain the complete offset.
In the next step, the upper compression plate of the test rig is loosened slightly. The zero point immediately changes from approximately +118 N to only +14 N. This significant difference strongly indicates mechanical stress.
Further inspection shows that the upper mounting component is not completely parallel to the support surface of the force transducer. Tightening the bolts therefore introduces an additional bending moment into the sensor.
The support surface is corrected and a suitable force introduction component is installed. The same loading cycle is then repeated. After unloading, only a small residual offset remains, which continues to decrease during the subsequent stabilization period.
The force transducer itself was therefore not the main cause. This example shows why poor zero return should always be evaluated together with the complete mechanical force application system.
When Is Calibration Useful?
Calibration or reference testing is particularly useful when a zero-point offset remains even in a controlled test setup, when overload is suspected or when the increasing and decreasing load characteristic is abnormal.
A change in sensitivity or poor repeatability can also make a new metrological test necessary. Defined calibration intervals may of course apply independently of the current fault condition.
A complete force calibration can evaluate zero deviation, sensitivity, linearity, hysteresis and repeatability, among other parameters. It therefore answers the question of whether the sensor still operates within the required tolerances under defined conditions.
However, calibration does not replace mechanical diagnosis of the installed system. A force transducer can perform perfectly in a calibration laboratory and show errors again after installation if the force application is unsuitable.
Systematic Troubleshooting
- Allow the load cell and measuring electronics to stabilize thermally.
- Ensure that no force is actually acting along the measuring axis.
- Document the unloaded zero point before each test without immediately re-taring it.
- Observe the signal under constant load for a defined period.
- Increase the load through several reproducible intermediate points.
- Reduce the load through the same intermediate points.
- Compare increasing and decreasing readings at identical force points.
- Document the zero point immediately after complete unloading.
- Continue recording the zero point after defined waiting periods.
- Check cables, hoses and other flexible connections for mechanical influence.
- Inspect mounting surfaces, bolted connections and force application.
- Eliminate transverse forces, bending moments and off-center loading.
- Check guides, stops and frame components for friction or contact.
- For vessel weighing systems, inspect pipes and hoses as possible force shunts.
- Compare the temperature trend with the development of the zero point.
- Check bridge excitation, measuring amplifier and PLC input electrically.
- If necessary, test the evaluation electronics using a suitable mV/V simulator.
- If possible, test the load cell outside the machine under defined conditions.
- After a possible overload event, do not simply re-zero the system; check the complete characteristic.
- Document results and signal trends for future comparison measurements.
Common Diagnostic Mistakes
- Calling every residual value after unloading creep: Creep is evaluated while a constant load remains applied.
- Confusing creep recovery with hysteresis: Hysteresis only becomes visible by comparing increasing and decreasing load sequences.
- Suspecting only the load cell: Mechanical stress and friction are very common causes.
- Re-taring before every test: This hides the actual zero-point trend.
- Testing only 0 and 100%: Intermediate points provide considerably more information about hysteresis and characteristic behavior.
- Failing to document waiting times: Without a time reference, creep and creep recovery can hardly be evaluated meaningfully.
- Underestimating cable forces: With small measuring ranges, even a stiff connection cable can have a visible influence.
- Ignoring hoses and pipes: In vessel weighing systems, they can create significant force shunts.
- Mounting the load cell misaligned or under stress: Transverse forces and moments reduce zero return and repeatability.
- Simply re-zeroing after an offset: A mechanical cause remains unchanged.
- Considering overload only as excessive axial force: Impact, transverse forces and bending moments can also damage a sensor.
- Ignoring temperature: Zero point and output signal can change during thermal stabilization.
- Never testing the sensor separately: This makes it difficult to determine whether the deviation originates from the sensor or the mechanical installation.
Suitable Force Transducer
For force measurements in test rigs, material testing machines and industrial test equipment, the WIKA Type F2222 Tension/Compression Force Transducer, for example, is suitable.
| Feature | Importance for the Application |
|---|---|
| Tension and compression force measurement | suitable for a wide range of test and machine applications |
| Wide measuring range selection | versions from small forces up to the MN range |
| Low installation height | suitable for compact mechanical designs |
| High long-term stability | suitable for repeated industrial measurement tasks |
| Static and dynamic measurements | suitable for test rigs and machine applications |
| IP66 protection | suitable for industrial environments |
| Optional force introduction components | support suitable mechanical installation |
| Optional measuring amplifier | simplifies integration into downstream evaluation systems |
For reliable measurement, the force must be applied centrally and as free as possible from transverse forces and unwanted moments. A flat mounting surface and reproducible mechanical installation are therefore just as important as the electrical characteristics of the sensor.
The faults described in this article in particular show that even a highly accurate force transducer can only achieve its specified measurement performance if the mechanical force path is designed reproducibly.
Further load cells, tension/compression force transducers, shear beams and weighing electronics can be found under Force, Weighing, Speed and Torque Sensors at ICS Schneider.
Further information on mechanical installation can be found in the article Installing a Load Cell Correctly: Avoiding Force Introduction Errors, Mounting Problems and Force Shunts.
For electrical installation, the article Extending Load Cell Cables: Correctly Considering 4- and 6-Wire Technology, Sense Lines and Voltage Drop is also helpful.
Conclusion
If a load cell does not return exactly to its previous zero value after complete unloading, this alone does not provide a clear indication of a sensor defect. Different physical and mechanical effects can produce very similar symptoms.
Creep describes the time-dependent change in the output signal while a constant load remains applied. After unloading, creep recovery may occur, during which the measured value slowly returns toward the original zero point.
Hysteresis is a different effect. It occurs when the same force produces different output signals depending on whether the force is approached during increasing or decreasing load. Complete loading and unloading sequences are therefore required for its evaluation.
Additional influences occur in the installed condition. Mechanical stress, transverse forces, friction, force shunts, cables, hoses and deformation of the machine frame can cause a small force to remain on the sensor after unloading.
Temperature and electronics must also be considered separately. A zero point that slowly changes with ambient temperature requires a different diagnostic approach from an offset that appears suddenly after an overload event.
A large permanent zero shift after an exceptional mechanical load is particularly critical. Performing a new zero adjustment must not be confused with a successful repair. Sensitivity, linearity, hysteresis and repeatability must also be checked.
The most reliable diagnosis therefore results from a controlled loading cycle with a documented time axis, followed by comparison between the installed condition and a defined reference setup.
The decisive question is not only: “Does the load cell return to zero after unloading?”, but rather: “When does the deviation occur, how does it develop over time and does it disappear once force application, friction, cables, temperature and the electrical measuring chain have been excluded as influencing factors?”
FAQ: Load Cell Does Not Return to Zero
Why does my load cell still show a value after unloading?
Possible causes include creep recovery, hysteresis, mechanical stress, friction, cable forces, temperature changes, electrical drift or a previous overload. It is particularly important to observe how the residual value changes over time and whether it remains under defined mechanical conditions.
Is a residual value after unloading automatically creep?
No. Creep describes the time-dependent signal change while a constant load remains applied. The time-dependent change after the load has been removed is referred to as creep recovery.
What is hysteresis in a load cell?
Hysteresis describes the difference in output signal at the same force when the load point is reached once during increasing load and once during decreasing load. At least one increasing and one decreasing load sequence are therefore required for evaluation.
How can I detect mechanical stress?
A typical indication is a significant change in the zero point when mounting bolts are loosened, the force application system is moved or the sensor is removed. Poor repeatability can also indicate friction or mechanical stress.
Can the connection cable really influence the measured value?
Yes. Particularly with small measuring ranges, a cable under tension or routed too rigidly can apply an additional mechanical force to the sensor. If the zero value changes when the cable is moved, the cable routing should be optimized.
Can a vessel connection influence the zero point?
Yes. Pipes, hoses, cables and other connections between the weighed and fixed parts of a system act as additional force paths. If these forces change during filling or emptying, zero and repeatability errors can occur.
How can I distinguish creep from temperature drift?
Creep is evaluated under constant load and, as far as possible, stable temperature conditions. If the zero point instead changes in parallel with sensor or ambient temperature even without a load change, a thermal influence is more likely.
What indicates an overloaded load cell?
A suddenly and significantly changed zero point that remains after an exceptional loading event can be an indication. Sensitivity, linearity, hysteresis or repeatability may also deteriorate.
Can I simply re-zero an overloaded load cell?
A new zero adjustment only removes the displayed offset. It does not reverse possible plastic deformation of the spring element. After suspected overload, the complete measurement characteristic should therefore be checked.
How should zero return be tested correctly?
First, document the thermally stabilized unloaded zero value. Then apply a defined load and remove it completely. Record the zero point immediately after unloading and again after defined waiting periods.
Why should I measure increasing and decreasing load points?
Only this makes it possible to determine whether the same force is indicated differently depending on the direction of loading. This difference is important for evaluating hysteresis and mechanical repeatability.
When should the load cell be tested separately?
If the cause cannot be clearly identified in the installed condition or if sensor damage or overload is suspected. A defined reference setup separates the load cell from influences such as frame stress, friction, pipe forces and mechanical stops.
When is calibration useful?
Calibration is particularly useful when the zero point remains permanently shifted, sensitivity or hysteresis is abnormal, or overload is suspected. It should complement the mechanical examination of the installation rather than replace it.