Chinese
Spanish
Russian
Portuguese
French
German
Indonesian
Thai
Vietnamese
Arabic
Italian
Turkish
Polish
Japanese
Load cell failure symptoms often originate outside the load cell. A drifting zero can come from a pipe force, a missing weight value can come from the instrument, and a wrong PLC value can come from scaling. Start by finding where the expected reading first becomes wrong. Replacing sensors before locating that point can leave the fault intact and erase useful evidence.
Record the symptom at more than one point
Before resetting zero or changing gain, record the local display, controller value, alarms and process state at the same time. Save any accessible raw signal or diagnostic count without altering the configuration. Note recent cleaning, maintenance, overload or software changes. Photographs of the mounted sensor and terminals can be more useful than a description such as "weight unstable".
Confirm that the load is actually unchanged. Filling, draining, vibration, moving cables and attached hoses can all alter the force being measured. For a vessel, a valve operation may change pipe force even when no product enters or leaves. For a test fixture, a guide can stick and release as the actuator moves.
Use load cell failure symptoms to choose a test
| Symptom | Compare first | Next area to investigate |
|---|---|---|
| PLC wrong; local weight correct | Time-matched values, units and output configuration. | Analog scaling, register interpretation or stale communication data. |
| Different result at one load position | The same reference load at controlled positions. | Platform contact, support seating and channel balance. |
| Zero moves after cleaning | Before/after readings under the same empty condition. | Retained liquid, moved components, cable entries and moisture. |
| Stable but proportional error | Several known loads and the saved calibration settings. | Span, sensitivity configuration or a changed signal path. |
| No usable indication | Instrument power, diagnostics and the documented input type. | Supply, excitation, wiring, input saturation or instrument failure. |
The table identifies starting points, not confirmed causes. A sensor fault and a mechanical fault can coexist. For position-dependent behaviour, use the dedicated corner-error procedure rather than recalibrating the centre point until it looks correct.
Compare the instrument output with the controller
If the indicator reads correctly, investigate the downstream path before disturbing its sensor connections. Confirm the selected output option, range and unit. For 4-20 mA, compare the actual output and receiving configuration using an approved test method. For a digital connection, verify the device's own register map, data representation and update or fault status.
RS485 describes an electrical interface; it does not by itself identify a weight register or prove that two devices use the same application protocol. A value that remains numerically steady may simply be old data. The PLC integration guide covers the checks between bridge input, transmitter and controller.
Check the input configuration against the exact sensor
The METRAVON MV-LC0190 lists 0.5 mV/V nominal sensitivity at 1 kg, but 1 mV/V at its 5, 20 and 50 kg capacities. Configuring a replacement from the family name alone can therefore give the wrong scaling. Read the capacity and supplied data rather than assuming that every version has the same signal.
The MV-LC0030 is a different bridge configuration, with nominal 2 mV/V output and 10-12 VDC excitation. These values are not permission to apply the same excitation to an unidentified sensor. Verify the exact unit, wiring and instrument input before electrical testing.
A raw mV/V sensor is not a current transmitter. Connecting it to a standard PLC current input will not establish a useful weight reading. Conversely, a correctly operating conditioned output can be misread if the receiving channel is configured for the wrong electrical range.
Narrow a channel fault without creating another one
For a multi-cell system, use available individual-channel diagnostics to identify where behaviour differs. Equal readings are not required if support loads differ. A meaningful comparison uses the expected load distribution, sensitivity and actual excitation. If only a summed signal is available, follow the system's documented isolation procedure; removing one electrical input changes the summed result.
Inspect terminations, cable strain and visible moisture before attributing a suspect channel to the sensing element. Resistance tests require de-energized, appropriately disconnected circuits. Insulation tests need the permitted test voltage and isolated electronics. Do not disconnect or loosen load-bearing parts on an operating structure to obtain a zero reading.
A suitable simulator can help separate the instrument input from the upstream sensor circuit. A successful simulator check supports the tested electronics, not the mechanical installation or the disconnected sensor. Record exactly where it was connected and which sections of cable or junction box it bypassed.
Request help with measurements, not only a replacement part
Send the sensor and indicator identification, a wiring diagram, the time-matched readings and a description of the operating condition to METRAVON technical enquiries. Include what was changed and which tests restored or failed to restore the reading. State whether the system must display weight locally, transmit it to a PLC or do both.
If the evidence now points to physical sensor damage, continue with the damage and replacement assessment. If it points to a controller configuration, do not order a set of new cells to correct a software scaling error.
Questions when the fault is intermittent
Q1. What if cycling the power restores the reading?
Keep the event in the fault record. A restart may reset a communication state, instrument fault or temporary condition without removing its cause. Capture diagnostics before the next restart when safe, and compare power quality, process state and configuration. Recovery after reboot does not identify the sensor as good or bad.
Q2. Should stronger filtering be used to stop an unstable display?
First determine whether the fluctuation represents real force, electrical noise or a failing connection. Filtering can reduce visible variation while adding delay. Verify the process decision at the required time; a quiet display that responds too late can still cause a filling error.
Q3. Can a connector be unplugged while the scale is controlling a process?
Do not assume this is safe. Signal loss can trigger an unintended control action or leave a stale value in use. Put the process into its approved safe state and follow the equipment's isolation and test procedure before changing connections.
Summary
Locate the discrepancy along the force and signal path, preserve the original readings and change one test condition at a time. Confirm the repair under the condition that produced the fault. An empty display that looks stable after intervention is not enough to prove the system is ready for production.
