Quick answer: You can test a strain gauge load cell with an ordinary digital multimeter in three stages. First, with power off and the cell disconnected, measure bridge resistance across the excitation pair and the signal pair. Second, apply a steady excitation voltage and read the no-load output in millivolts. Third, add a known weight and confirm the output rises by the amount the datasheet predicts. Each reading is compared against the cell’s own datasheet, never against a generic number.
A multimeter will not calibrate a scale, and it will not tell you a cell’s accuracy class. What it will do, in about fifteen minutes and with a tool most technicians already carry, is separate the three things that account for the great majority of weighing complaints: a damaged cell, a damaged cable or junction box, and a perfectly healthy cell that is being fed bad information by the rest of the system. This guide walks through exactly how to do that, in the order you would do it on site.
Why Test a Load Cell with a Multimeter at All?
Anyone who has worked on industrial weighing for long enough knows the call. A hopper scale reads three kilograms high on Monday and eleven kilograms low on Thursday. A truck scale gives a different weight each time the same truck rolls on. A tank that was accurate for two years suddenly will not return to zero. The first instinct, almost always, is to blame the load cell, and the second instinct is to order a new one.
That is expensive and, more often than people admit, wrong. Load cells are among the most durable components in a weighing system. The parts that fail far more often are the ones around them: a cable crushed under a moved pallet, a junction box that let in condensation, a connector that has corroded, an indicator setting someone changed, or a mounting arrangement that has started to bind. A multimeter lets you test the cell on its own, in place, and rule it in or out before anybody spends money or shuts down a line.
There is a practical argument too. A multimeter is cheap, portable, and understood by every electrician and instrument technician. A dedicated load cell tester or simulator is a better instrument for some jobs, and we will cover where those fit, but a field fault at a remote site is usually diagnosed with whatever is in the toolbag. Knowing how to squeeze a trustworthy answer out of that tool is a skill worth having.
What Can a Multimeter Tell You About a Load Cell, and What Can’t It?
Before touching a single wire, it helps to be honest about what this method can and cannot prove. Overselling a multimeter test is how people end up signing off a cell that is quietly out of tolerance.
| Question | Can a multimeter answer it? | How |
| Is the strain gauge bridge intact (no open or shorted arm)? | Yes | Bridge resistance check |
| Is the cell leaking to the body, or has water got in? | Partly | Resistance to ground; a megohmmeter is far better |
| Has the zero balance shifted, for example after an overload? | Yes | No-load output in mV with excitation applied |
| Is the cell responding to load at roughly the right sensitivity? | Yes, to within a percent or two | Loaded output check against calculated mV |
| Is the cell within its accuracy class (OIML, NTEP and similar)? | No | Needs reference weights or a reference standard |
| Is there creep, hysteresis or poor linearity? | No | Needs a controlled load sequence and fine resolution |
| Is a digital or CAN or RS485 load cell healthy? | No | Needs the manufacturer’s software or a configurator |
The limit is resolution and accuracy. A handheld meter that reads to 0.01 mV on its lowest millivolt range is looking at a 20 mV full-scale signal in steps of about 0.05% of full scale, and the meter’s own accuracy specification, often something like half a percent of reading plus a couple of counts, sits on top of that. That is plenty to catch a dead bridge, a drifted zero, or a cell that has lost a chunk of its sensitivity. It is not enough to certify a legal-for-trade installation.
One more boundary: a multimeter test tells you what the cell is doing at the moment you test it. Intermittent faults, the kind that appear only when it rains or when the pump starts, need patience, repeated measurements, and sometimes a data logger left in place overnight.
How a Load Cell Makes Its Signal (The Five-Minute Primer)
Every number you will measure makes more sense once you picture what is inside the cell. A strain gauge load cell contains four strain gauges, thin foil grids bonded to the metal spring element, wired as a Wheatstone bridge. When load deforms the metal, two gauges stretch and two compress, which unbalances the bridge by a tiny amount. That unbalance appears as a small voltage across the signal wires.
The bridge has four connection points, and they come in two pairs:
- Excitation wires (EXC+ and EXC−): the supply voltage goes in here. Most indicators supply somewhere between 5 and 12 volts DC.
- Signal wires (SIG+ and SIG−): the millivolt output comes out here.
The output is ratiometric, meaning it scales directly with the excitation voltage. That is why datasheets quote sensitivity in millivolts per volt (mV/V) rather than in plain millivolts. A cell rated at 2 mV/V and fed 10 V will give 20 mV at full capacity. Fed 5 V, the same cell gives 10 mV at full capacity. Same cell, same load, different millivolts. Keep this in mind the first time you read a number and wonder why it does not match the datasheet: you probably used a different excitation.
The working formula you will use throughout this guide is simple:
Expected output (mV) = (applied load ÷ rated capacity) × rated output (mV/V) × excitation voltage (V)
For example, a 1,000 kg cell rated at 2 mV/V, excited at 10 V, with 250 kg on it, should give (250 ÷ 1000) × 2 × 10 = 5.00 mV. Every loaded check in this article is that arithmetic and nothing more.
Some cells have six wires instead of four. The extra pair, the sense wires (SENSE+ and SENSE−), connect to the excitation leads at the cell end and let a suitable indicator measure the voltage that actually arrives at the cell, compensating for cable losses on long runs. For a multimeter test they are mostly a continuity check, and we will treat them that way.
What You Need Before You Start
The Meter
Almost any decent digital multimeter will do, provided it can measure DC millivolts on a low range. Look for a range that resolves 0.01 mV, so a 60 mV range on a 6,000-count meter, or better still a 5½-digit bench meter for workshop testing. Cheap meters with a lowest range of 200 mV and one decimal place will show a 20 mV signal as “20.0”, which is too coarse to see small zero shifts.
The meter’s DC voltage input impedance is typically 10 MΩ or higher. Next to a bridge of a few hundred ohms, that is effectively no load, so connecting the meter across the signal pair does not disturb the signal. That matters, because it means you can measure in place without disconnecting the indicator.
Use fresh batteries. A weak battery can make low-range readings wander.
The Paperwork
Get the datasheet or the calibration certificate for the exact cell model, ideally with its serial number. You need five numbers from it:
- Rated capacity.
- Rated output (sensitivity) in mV/V.
- Input resistance and output resistance, in ohms.
- Zero balance tolerance.
- Recommended and maximum excitation voltage.
Without these you are measuring without a reference. A reading of 352 Ω means nothing until you know the cell is meant to read 350 Ω, or 1,000 Ω.
The Supporting Kit
- A stable DC supply for excitation, either the indicator itself or a bench supply set to a known voltage (5 V or 10 V works well). Avoid unregulated wall adapters.
- A second meter, if you have one, to monitor excitation while the first reads the signal.
- Small flat and Phillips screwdrivers for the junction box, and a torch.
- Terminal markers or tape and a pen. You will be disconnecting wires, and you will want to put them back exactly where they were.
- A notebook or phone, to record readings next to the cell’s identity and the date.
- Known weights, if you plan to do the loaded check, and a way to apply them safely.
Safety, Briefly But Seriously
Load cell excitation is low voltage, but the cabinet it lives in often is not. Isolate and lock out mains supplies before opening an indicator or junction enclosure. In a classified hazardous area, do not use an ordinary multimeter at all unless the area has been made safe under a permit or the meter is certified for the zone. Intrinsically safe installations exist precisely so that nobody introduces an energy source into the wiring.
Never put a weight on a cell without confirming it is within capacity, and never work under a suspended or partly supported vessel.
Identify the Wires Correctly
Most load cell cables follow a loose colour convention, such as red and black for excitation, green and white for signal, and a bare or yellow wire for shielding. The word “loose” is doing real work in that sentence. Manufacturers differ, and cables get re-terminated. Treat colours as a hint and the datasheet as the authority.
Work from the wiring diagram on the datasheet. If there is none, or if you suspect the cable has been modified, you can find the two pairs electrically, using the resistance pattern of a bridge.
With the cell disconnected and the meter on a resistance range, measure between every pair of the four signal-carrying wires. In an ideal bridge of four equal arms of resistance R:
- The two opposite-corner pairs, which are the excitation pair and the signal pair, each read about R.
- The four adjacent pairs each read about 75 × R.
For a 350 Ω bridge, that is roughly 350 Ω across the two true pairs and about 262 Ω for the four cross-readings. The two wires that make up a “true pair” are the ones that give the higher readings, with the matching partner found by elimination.
Real cells have compensation resistors added in series with the excitation leads to trim temperature performance and sensitivity. This usually makes the input resistance a little higher than the output resistance, for example 410 Ω in against 350 Ω out. In that case the pair with the higher reading is the excitation pair. If the readings are too close to call, rely on the datasheet or ask the manufacturer.
To confirm polarity, apply excitation, connect the meter across the signal pair, and load the cell in its working direction. For most compression cells the +SIG wire goes positive with compression. For tension cells it goes positive under tension. If the output goes negative, you have either found the signal wires reversed or are loading the cell in the opposite sense. Neither damages the cell, but you must know before wiring it back.
For six-wire cells, measure between EXC+ and SENSE+, and between EXC− and SENSE−. Each should read close to a short circuit, a few ohms at most, because the sense wires attach to the excitation wires inside the cell. An open reading means a broken sense lead.
Do the Visual and Mechanical Check First
Before you pick up the meter, spend two minutes looking. It costs nothing and often ends the search.
Look at the cable from gland to junction box. Is there a nick, a crush mark, a kink, or a section where something has been dragged across it? Is the cable pulled tight, so that every movement of the vessel tugs at the cell’s cable gland? Is the gland itself cracked or loose?
Look at the cell body. Is there rust, a dent, a weld spatter mark, or signs of an impact? Is the cell sitting square on its mounting, and are the mounting bolts present and properly seated? A cell that has been side-loaded or lifted off its base will behave oddly no matter how healthy its bridge is.
Look at the junction box. Open it and check for moisture, white or green corrosion on terminals, a tide mark of dried water, or insects. Check that the gasket sits properly and that unused cable entries are plugged. Many a “failed load cell” turned out to be a junction box that had been quietly breathing humid air for a year.
Make a note of what you see. If a cable gland is cracked, you have a likely cause already, and you can use the electrical tests to confirm it rather than to hunt.
How Do You Measure Load Cell Bridge Resistance?
Bridge resistance is the quickest and most revealing test, because it needs no excitation, no weights, and no special setup. It tells you whether the strain gauge bridge is physically intact.
What you need: the load cell fully disconnected from the indicator and junction box, and the meter on a resistance range suitable for a few hundred ohms (the 600 Ω or 6 kΩ range, or auto-range).
Procedure:
- Power down and isolate the system. Disconnect the load cell cable from the indicator, amplifier or junction box. In a summing junction box, disconnect the individual cell’s wires from the board, and label them.
- Short the meter leads together and note the lead resistance, typically a fraction of an ohm. Some meters have a relative or zero function to cancel it. For a 350 Ω reading it is negligible, but it matters on very low ranges.
- Measure between EXC+ and EXC−. This is the input resistance.
- Measure between SIG+ and SIG−. This is the output resistance.
- Wait a few seconds on each reading and note whether it is steady. A reading that creeps or jumps points to a poor connection or a damaged conductor.
- Record both values, with the temperature if you can.
What to expect: the figures come from the datasheet, which typically states them with a tolerance. Common values for strain gauge cells are 350 Ω, 700 Ω, 1,000 Ω, and occasionally higher. The datasheet may say, for instance, “input resistance 410 ± 10 Ω, output resistance 350 ± 3 Ω”. Your reading should fall within those limits.
A few points that catch people out:
- Strain gauge foil has a small temperature coefficient. A cell measured in a hot boiler house and again in a cold store will read slightly differently. The shift is small compared with the tolerance, but if you are comparing against an old record, compare like with like.
- Cable resistance. If you measure at the far end of a long cable, the copper adds series resistance. A 4 mm² cable is negligible; a long run of thin 0.25 mm² cable can add a few ohms. Measure at the cell end where possible, or subtract the known cable resistance.
- Do not measure with the cell connected. An indicator’s input circuitry and the other cells in a summing box all sit in parallel with the cell you are testing and will give false, usually lower, readings.
Interpreting the result:
| Reading | Likely meaning |
| Both within datasheet tolerance and steady | Bridge intact. Move on to the next check. |
| Open circuit (OL) on one pair | Broken conductor in the cable, a failed solder joint, or a burned gauge. |
| Reading much lower than expected | A short between arms, often from moisture inside the cell or damaged insulation. |
| Reading higher than expected | A partially broken conductor or high-resistance joint. |
| Reading drifts or jumps while you touch the cable | Intermittent break in the cable or at a terminal. Flex the cable gently along its length to find the spot. |
| Input and output both off by a similar percentage | Check the meter range and lead resistance before suspecting the cell. |
A bridge that reads correctly proves a good deal but not everything. A cell can have a perfect bridge resistance and still carry a shifted zero from an old overload. That is what the next checks are for.
What Is a Good Insulation or Resistance-to-Ground Reading?
Moisture is the quiet killer of weighing systems. It enters through a damaged cable, a failed gland, or a cracked seal, and it creates leakage paths between the bridge and the metal body of the cell. The symptoms are unstable readings, slow drift, and a scale that behaves differently on humid days.
The multimeter version of the test: with the cell disconnected, set the meter to its highest resistance range (megohm range if it has one) and measure between each wire and the cell body, one at a time, and then between the shield and the body if the shield is not bonded to it. A healthy cell reads open circuit or an extremely high resistance. Anything that reads in kilohms, or that settles to a steadily falling value, points to leakage.
A typical multimeter on its top range can display up to around 40 or 60 MΩ. That is useful for catching gross moisture ingress, and it is a legitimate field screen.
The proper version: a dedicated insulation tester, often called a megohmmeter or “megger”, applies a higher voltage and reads in gigaohms. For load cells, the usual advice is to use a test voltage of 50 V DC or less, and to confirm the limit in the datasheet. Many datasheets quote an insulation resistance of 5,000 MΩ or more at 50 V DC. Testing at a higher voltage, such as the 500 V or 1,000 V commonly used on motor windings, can damage a strain gauge bridge. Do not use a 500 V megger on a load cell.
When using a megger, short all four bridge wires together first, so every conductor is at the same potential, and then test from that bundle to the body. This avoids stressing the gauges.
Interpreting the result: the pass criterion is the datasheet figure. In practice, a reading that is in the thousands of megohms is healthy, a reading in the tens of megohms is suspect, and anything below a megohm is a clear fault. A cell that fails here can sometimes be rescued if the moisture is only in the cable gland or the cable end, which can be dried, trimmed and re-sealed. If the water has reached the strain gauge area inside the cell, the repair is usually not practical.
If you do not have a megohmmeter, record the multimeter’s high-range reading anyway. A cell that read “OL” last year and reads 18 MΩ this year has changed, and the change is the information.
How Do You Check the Excitation Voltage Reaching the Cell?
A surprising proportion of weighing faults are excitation faults. The cell is perfectly healthy but is being fed too little voltage, or an unstable voltage, or a voltage that falls under load. Because the output is ratiometric, a drop in excitation shows up as a drop in the apparent weight.
Procedure:
- With the system powered, set the meter to DC volts (the normal range, such as 20 V).
- Measure across EXC+ and EXC− at the load cell end, that is, at the junction box terminals for the cell or at the cable end if you can reach it.
- Measure across the excitation terminals at the indicator or amplifier as well.
- Compare the two. Note the voltages.
What to expect: at the indicator you should see the rated excitation, commonly 5 V or 10 V, within a few tens of millivolts. At the cell the voltage should be close to that, slightly lower on a long run of a four-wire cable because of the voltage drop in the excitation conductors. A drop of a few percent on a long run is normal and is exactly what six-wire sensing corrects. A large drop points to a high-resistance joint, a corroded terminal, or a damaged conductor.
If the excitation is unstable, watch the meter for a minute. A flicker of more than a few millivolts, or a slow wander, suggests a failing supply board in the indicator, a poor connection, or electrical interference. Be careful to separate a noisy meter reading from a noisy supply: the same noise appearing on a second meter confirms the supply.
Always record the actual excitation, because you will need it to convert your millivolt readings to mV/V in the steps that follow. Do not assume it is exactly 10.000 V. A real supply may well be 9.87 V, and the difference is more than enough to make a healthy cell look 1.3% low.
How Do You Measure Load Cell Zero Balance with a Multimeter?
Zero balance (also called zero offset or zero output) is the signal the cell gives with no load. A perfect bridge gives exactly zero. Real cells are trimmed at the factory to a small tolerance, commonly quoted as a percentage of rated output, with figures around ±1% of rated output being a typical example, and tighter for precision cells. Always use your own datasheet figure.
This test is the best single indicator of overload damage. A cell that has been pushed beyond its elastic limit, even briefly, tends to take a permanent set. The spring element no longer returns to its original shape at zero load, and the bridge never rebalances. The zero balance shifts, often by far more than the tolerance.
Procedure:
- The cell must be completely unloaded. This is the catch. In an installed cell, the weight of the empty vessel, platform or hopper (the dead load) is already on it. A true zero-balance test means removing the cell, or at least relieving all load from it. If you cannot, see the note below.
- Reconnect the cell to the excitation source and let it warm up for a few minutes, so the bridge settles. Cells and meters both behave better once thermally stable.
- Measure the excitation voltage across EXC+ and EXC−.
- Set the meter to DC millivolts on its most sensitive suitable range, and measure across SIG+ and SIG−.
- Record the reading in mV and convert to mV/V by dividing by the excitation voltage.
Worked example. A 1,000 kg cell is rated at 2 mV/V with a zero balance tolerance of ±1% of rated output. Rated output at 10 V is 20 mV, so the allowed zero is ±0.2 mV. You measure 0.08 mV with 10.00 V of excitation. That is 0.008 mV/V, or 0.4% of rated output. Well within tolerance.
Now suppose a cell reads 1.35 mV at the same excitation. That is 6.75% of rated output, far beyond ±1%. The bridge has shifted, and the most likely cause is an old overload. The cell may still give a linear response, but it has been mechanically changed and should be treated with suspicion, usually replaced in any application where accuracy matters.
Testing an installed cell. When you cannot unload it, you cannot measure true zero balance, but you can still make a useful comparison. Measure the output with the structure in its normal empty condition and compare it with:
- The same measurement recorded at commissioning.
- The other cells in the same system, which carry similar dead loads in a symmetrical installation.
A cell whose empty-vessel reading has moved by a lot since commissioning, or that differs wildly from its neighbours when the loads should be similar, is the one to look at. This is why the log sheet at the end of this guide matters. A zero reading is far more valuable with a baseline beside it.
A word on stability. Watch the reading for at least a minute. A good cell with a good cable settles to within a digit or two. A reading that drifts steadily in one direction suggests a moisture problem or a temperature gradient across the cell, for example one side in direct sun. A reading that jumps in steps suggests a loose connection.
How Do You Check Load Cell Output Under Load?
Zero tells you the cell’s resting state. The loaded check tells you whether it responds to force the way it should. You apply a known weight and compare the change in output with the calculated change.
What you need: the cell connected to a stable excitation, the meter across the signal pair, and one or more known weights. For a cell in a working scale, you can use test weights or a known load such as a calibrated reference mass. For a cell on a bench, a certified weight set or a calibrated test frame works. The weight does not have to be full capacity. A load of 20% to 50% of capacity gives a large enough signal to measure meaningfully and is far safer.
Procedure:
- Measure and record the excitation voltage.
- Record the zero reading (the no-load output, with the dead load in place if installed).
- Apply the known load smoothly and allow it to settle for a minute. Never drop weights on a cell.
- Record the new reading.
- Subtract the zero reading to get the change in output caused by the added load.
- Calculate the expected change using the formula and compare.
- Remove the weight and confirm the reading returns to zero.
Worked example. A 500 kg cell is rated at 3 mV/V. You measure 9.95 V at the cell. Full-scale output would be 3 × 9.95 = 29.85 mV. You add a 100 kg reference mass, which is 20% of capacity. Expected change: 0.20 × 29.85 = 5.97 mV. You measure the zero as 0.12 mV and the loaded reading as 6.05 mV. The change is 5.93 mV. That is 0.7% below the calculated value, which is acceptable for a multimeter-level check once you allow for meter accuracy, weight uncertainty, and any friction in the mounting. If the figure had been 5.2 mV, a deficit of 13%, something would be wrong. The cell may have lost sensitivity, the load may not be reaching it (binding hardware, a bumper touching, a stuck foot), or the excitation may be lower than you think.
A multi-point version gives you more information. Apply 25%, 50%, 75% of capacity in steps, recording the reading at each, and then reduce in steps. For a 1,000 kg, 2 mV/V cell at 10 V, the expected readings are:
| Applied load | Expected output (mV) |
| 0 kg | 0.00 (± zero balance) |
| 250 kg | 5.00 |
| 500 kg | 10.00 |
| 750 kg | 15.00 |
| 1,000 kg | 20.00 |
Plot your readings against the expected line. They should fall on a straight line. A cell that curves, or that reads differently on the way down than on the way up, has non-linearity or hysteresis, but the multimeter’s resolution means only gross errors will show. That is a prompt for proper calibration rather than a verdict.
What the numbers mean when they are wrong:
- Output proportional but consistently low or high by the same percentage: check the excitation reading first, then the weight value. If both are right, the cell sensitivity has changed.
- Output normal for the first load but not for the next: possible mechanical binding or a hardware problem that appears only under load.
- Output returns to a new zero after unloading: permanent set, a sign of overload or fatigue.
- Output unstable while load is steady: moisture, a cable fault, or electrical interference.
Testing a Multi-Cell System: Isolate, Compare, Conclude
Most real installations have three to six cells wired through a summing junction box into a single indicator. A reading from the indicator tells you about the sum, which hides which cell is at fault. Multimeter testing is at its best when you break the system apart.
Test one cell at a time. At the junction box, disconnect one cell’s signal and excitation wires from the board (labelling them first), and measure that cell’s bridge resistance, as before. If you can supply excitation to the individual cell from a bench supply or an indicator, you can also read its individual zero and loaded outputs.
Compare the cells with each other. In a symmetrical installation, such as a four-cell tank, the cells carry similar dead loads and respond similarly. Compare the individual readings. A cell whose output differs sharply from its neighbours points to either an uneven load distribution or a problem in that cell or its cable.
Do not touch the trim potentiometers. Many analogue summing boxes have small trimmer pots used for corner adjustment during commissioning. These set the relative sensitivity of each cell. Adjusting them while diagnosing a fault, and failing to record the original position, can make a bad situation much worse. Leave them alone unless you know what you are doing, and mark their positions with a photograph if you must open the box.
Remember the common cause. If every cell in the system shows the same fault, such as all outputs being 10% low, suspect the shared elements: the excitation supply, the indicator, or the junction box, rather than four simultaneous cell failures.
Reading the Results Together: A Diagnostic Guide
Individual measurements make sense only in combination. The table below links common symptoms to likely causes and the next step.
| Symptom | Likely cause | What to do next |
| Open circuit on a bridge pair | Broken cable conductor or terminal | Inspect the cable and terminals; test at the cell end of the cable |
| Bridge resistance low | Moisture inside the cell or cable; short between arms | Test resistance to ground; check the gland and cable |
| Resistance to ground in kilohms or low megohms | Water ingress | Dry and re-seal the cable end; if the cell interior is wet, replace |
| Zero far outside tolerance | Overload damage, permanent set | Compare with the commissioning record; replace if confirmed |
| Zero drifts slowly in one direction | Moisture, thermal gradient, or a supply problem | Check insulation, shade or insulate the cell, monitor excitation |
| Zero jumps in steps | Loose terminal or a failing connector | Re-terminate; check the gland and cable strain relief |
| Loaded output low by a fixed percentage | Wrong excitation assumed, mounting binding, or lost sensitivity | Verify excitation at the cell; inspect the mounting hardware |
| Excitation at the cell much lower than at the indicator | High-resistance joint or thin cable on a long run | Check the terminals; consider six-wire sensing |
| Excitation unstable | Failing supply or interference | Test with a known-good supply; check grounding and shielding |
| All cells wrong by the same factor | Shared element at fault | Check the indicator settings, the supply, and the junction box |
| Readings change when the cable is touched | Intermittent conductor break or damaged shield | Flex the cable along its length; replace the damaged section |
A Walkthrough: The Hopper That Drifted Overnight
The following scenario is illustrative, assembled from the kind of fault pattern that technicians meet regularly, to show how the steps combine in practice.
A batching hopper sits on four shear beam cells in an outdoor plant. The operators report that the empty hopper reads near zero in the afternoon but creeps up by 15 to 20 kg overnight, and returns towards normal by midday. Calibration was done six months ago and passed.
The technician starts with the visual check. The junction box on the hopper frame has a gasket that looks flattened, and one cable entry has a loose gland. There is a faint tide mark inside the lid. That is a hint, not a conclusion.
The meter goes in next. With the power off, each cell is disconnected at the junction box. Bridge resistance on all four cells reads within a couple of ohms of the datasheet, steady and clean, so the bridges are intact. Resistance to ground is then measured on each cell, with the meter on its highest range. Cells 1, 2 and 4 read open circuit. Cell 3 reads 2.4 MΩ and slowly falls as the leads are held on.
That is the finding. Cell 3 is leaking, and the leakage varies with humidity, which explains why the scale drifts overnight when condensation forms and recovers as the day warms. The technician traces the leak by testing the cable separately from the cell. The cable end, where it enters the junction box, is damp, and the cell body is clean. The cable is trimmed back past the damaged section, the conductors are re-terminated, the gland is replaced with a properly sealed one, and the junction box gasket and the loose entry are corrected.
After the repair, the resistance to ground on cell 3 reads open circuit, the zero is stable over a 24-hour check, and the system is returned to service. No cell was replaced. The multimeter turned a vague complaint into a specific, cheap repair in under an hour.
Ten Mistakes That Waste Hours (and Sometimes Cells)
Measuring resistance with the cell still connected
The indicator, the other cells and the junction box all sit in parallel and give false readings. Disconnect first.
Leaving a probe in the current jack
If the red lead is in the mA or A socket and you touch the excitation terminals with the meter on a current range, you create a near short circuit across the supply. Fuses blow, or supplies fail. Check the jack position every time you change function.
Using the wrong excitation in the calculation
Using 10 V because that is what the indicator is supposed to supply, when the actual voltage is 9.6 V, builds a 4% error into every conclusion. Always measure it.
Using a high-voltage megger
A 500 V or 1,000 V test can destroy a strain gauge bridge. Keep to 50 V DC unless the manufacturer says otherwise.
Trusting wire colours
A cable that has been re-terminated may not follow the original code. Confirm with the datasheet or the resistance pattern.
Testing zero on an installed cell and calling it “zero balance”
With dead load on the cell, you are measuring something different. Treat it as a comparison against past readings, not as a datasheet check.
Dropping weights on the cell
Shock loading can cause the very damage you are testing for. Apply loads smoothly.
Ignoring temperature
Comparing a reading from a cold morning with a baseline from a warm afternoon can mislead. Record conditions.
Adjusting trim pots during diagnosis
It changes the baseline and hides the original fault. Leave them, and photograph them if the box is open.
Not recording anything
A set of readings with no date, no cell identity and no excitation voltage is worth very little next year. Write it down.
What a Multimeter Cannot Replace
Knowing the limits of the tool is part of using it well. There are jobs that need something better.
| Tool | What it does that a multimeter can’t |
| Calibration with certified reference weights | Verifies accuracy, linearity and repeatability of the whole weighing system, and produces a certificate |
| Load cell simulator | Injects a precise, known mV/V signal to test the indicator and wiring without any cell attached, separating cell faults from instrument faults |
| Load cell tester | Combines excitation supply, precise millivolt readout and often resistance checks in one purpose-built unit |
| Megohmmeter (50 V) | Measures insulation resistance in gigaohms, finding moisture problems long before they affect the reading |
| Data logger | Captures intermittent faults over hours or days |
| Manufacturer’s software and interface | The only way to diagnose digital load cells, which present data over RS485, CAN or similar buses rather than a millivolt signal |
A simulator deserves a special mention. If you suspect the indicator, connect a simulator to the indicator’s input in place of the cell and dial in a known signal. If the indicator reads correctly, the instrument is fine and the fault lies in the cell, the cable or the box. If it does not, the indicator is your problem. Pairing a multimeter with a simulator splits a weighing system neatly into two halves.
For any scale used for trade, or any process where accuracy has a legal or financial consequence, a multimeter test is a screening step. The final proof is a calibration with certified weights, carried out by a competent person.
Keeping a Test Log (and Why It Pays for Itself)
The single best thing you can do to make future multimeter tests useful is to make this one repeatable. Record the same measurements, the same way, at commissioning and at every maintenance visit, and the numbers become a history rather than a snapshot.
A simple log sheet needs one row per cell per visit:
| Field | Example entry |
| Date and time | 21 Oct 2026, 10:40 |
| Site / scale ID and cell position | Hopper H2, cell 3 |
| Cell model and serial number | (from the nameplate) |
| Ambient temperature | 24 °C |
| Input resistance (EXC+ to EXC−) | 410.6 Ω |
| Output resistance (SIG+ to SIG−) | 350.2 Ω |
| Resistance to ground | > 40 MΩ (OL) |
| Excitation at cell | 9.94 V |
| Empty (dead load) output | 3.41 mV |
| Loaded output with known mass | 9.38 mV at 100 kg |
| Notes | Cable gland replaced; no visible corrosion |
| Technician | (initials) |
With a few visits recorded, a slow drift in input resistance, a creeping dead-load output, or a gradual fall in insulation resistance appears long before anyone notices an error on the scale. That is the difference between preventive maintenance and an emergency call.
Preventive Habits That Make Faults Rarer
Testing finds faults. A handful of installation habits prevent them.
Support the cable properly. Leave a drip loop so water runs off before it reaches the gland, and fix the cable so that movement of the vessel does not pull on the cell’s cable entry.
Seal the junction box and keep it sealed. Check gaskets, plug unused entries, and avoid opening it in rain. Desiccant packs help in humid climates, though they are a supplement to good sealing, not a substitute.
Protect against welding and lightning. Welding near a load cell without bypassing it can pass heavy current through the strain gauges. Always fit a heavy welding earth cable that bypasses the cells, or remove the cell and replace it with a dummy. In lightning-prone locations, surge protection on the signal and supply lines is a worthwhile addition. After a nearby strike or weld, test bridge resistance and insulation before returning the scale to service.
Avoid overload at the source. Mechanical stops, correct mounting hardware, and sensible procedures for loading and unloading do more for a cell’s life than any test regime.
Check regularly, not only when something breaks. A quick bridge resistance and zero check at routine maintenance visits takes minutes, and it is the cheapest form of reliability engineering available.
How Rudrra Sensors Supports Field Testing and Troubleshooting
Rudrra Sensors manufactures load cells, load cell amplifiers, digital indicators and related weighing assemblies in India, and supplies customers across industrial sectors. A good multimeter test starts with good documentation, and that is something we treat as part of the product.
When you buy a Rudrra load cell, you can ask for the datasheet figures that this guide keeps returning to: rated capacity, rated output in mV/V, input and output resistance, zero balance tolerance, insulation resistance and recommended excitation. Having those on file at installation turns a future fault-finding visit from guesswork into a ten-minute comparison.
If you are troubleshooting an existing installation and the numbers do not add up, our engineering team can help interpret what you are seeing. Send us the cell model or serial number, your measured bridge resistance, zero reading, excitation voltage and a description of the symptoms, and we will help work out whether the cell, the cable, the amplifier or the indicator is the likely cause. For installations where a cell has been confirmed faulty, we can recommend a suitable replacement and the mounting hardware to go with it, so the same problem does not return.
Frequently Asked Questions (FAQs)
Q1: Can you test a load cell with a multimeter?
Yes. A digital multimeter can check a strain gauge load cell’s bridge resistance, its resistance to ground, its zero output in millivolts and its change in output under a known load. These tests reveal open or shorted bridges, moisture leakage and overload damage. They do not prove accuracy class, linearity or hysteresis, which need calibration with reference weights.
Q2: What should the resistance of a load cell be?
It depends on the model. Strain gauge load cells commonly have bridge resistances of 350 Ω, 700 Ω or 1,000 Ω, and the input (excitation) resistance is often a little higher than the output (signal) resistance because of compensation resistors. Always compare your reading with the datasheet figures and tolerance for that specific cell.
Q3: What is a good zero balance reading for a load cell?
Zero balance is the output with no load applied, and datasheets usually give a tolerance as a percentage of rated output, with figures around ±1% being common and tighter for precision cells. For a 2 mV/V cell excited at 10 V, ±1% corresponds to about ±0.2 mV. A reading far outside the datasheet tolerance suggests overload damage or a wiring fault.
Q4: How do I know if a load cell is bad?
Typical signs are an open or very different bridge resistance, a low resistance to ground, a zero output far outside tolerance, a loaded output that does not match the calculation, or readings that drift or jump. Rule out the cable, junction box, excitation supply and indicator before condemning the cell, since those are more often the cause.
Q5: How do I calculate the expected millivolt output of a load cell?
Multiply the fraction of capacity applied by the rated output in mV/V and by the actual excitation voltage. For example, 250 kg on a 1,000 kg, 2 mV/V cell with 10 V excitation gives 0.25 × 2 × 10 = 5 mV. Measure the excitation rather than assuming it, because the output scales directly with it.
Q6: Can I measure a load cell while it is connected to the indicator?
You can measure the signal voltage across SIG+ and SIG− while the system is powered, because a multimeter’s high input impedance barely disturbs the signal. You cannot measure resistance while connected, since the indicator and other cells in parallel give false readings. Disconnect the cell first for any resistance test.
Q7: What voltage should I use to test a load cell?
Use the excitation voltage given in the datasheet, commonly 5 V or 10 V DC, from a stable supply, and never exceed the maximum stated. For insulation testing with a megohmmeter, keep to 50 V DC or lower unless the manufacturer specifies otherwise. A high-voltage insulation test can damage the strain gauges.
Q8: Can I test a digital load cell with a multimeter?
Only partly. You can check the supply voltage and wiring continuity, but digital load cells output data over a communication bus such as RS485 or CAN rather than a millivolt signal, so their health is checked with the manufacturer’s software or a suitable interface, not a multimeter’s millivolt range.
Q9: Why does my load cell read differently on different days?
Common causes are moisture in the cable or junction box, temperature gradients across the cell, an unstable excitation supply, or a loose connection. Measuring resistance to ground, excitation voltage and zero stability over time usually identifies which one it is. A test log makes the pattern easier to see.
Q10: How often should load cells be tested?
At minimum, check bridge resistance, insulation and zero at routine maintenance visits, commonly once or twice a year, and after any event that could stress the system, such as an overload, a nearby lightning strike, welding work, flooding or an impact. Scales used for trade also need periodic calibration and verification as required by local regulation.
A Glossary of Key Terms
Bridge resistance — The electrical resistance across the strain gauge bridge, measured between the excitation pair (input resistance) or the signal pair (output resistance).
Excitation voltage — The DC voltage supplied to the bridge, commonly 5 V or 10 V, to which the output signal is proportional.
Rated output (sensitivity) — The change in output at full capacity, expressed in millivolts per volt of excitation (mV/V), commonly 2 or 3 mV/V.
Zero balance — The output of the cell with no load applied, expressed as a percentage of rated output or in mV/V.
Insulation resistance — The resistance between the bridge circuit and the cell body, a measure of how well the cell is sealed against moisture, usually quoted in megohms or gigaohms.
Ratiometric output — An output that scales in direct proportion to the excitation voltage, so that the ratio of output to excitation stays constant.
Six-wire connection — A cable arrangement with two additional sense wires that let an indicator measure and compensate for voltage drop in the excitation leads.
Dead load — The weight of the empty structure (vessel, platform or hopper) carried by the load cells before any material is added.
Permanent set — A lasting deformation of the cell’s spring element after overload, causing a shifted zero that does not return to its original value.
Load cell simulator — A test instrument that outputs a precise, known signal in place of a load cell, used to verify an indicator or amplifier independently of the cell.
Closing Thoughts
A multimeter is not a glamorous instrument, and a load cell is not a glamorous component. Together, though, they solve a remarkable share of the problems that stop a scale from being trusted. The method is not complicated: check the bridge, check the insulation, check the supply, check the zero, check the response to a known load, and write it all down. Each step narrows the search, and by the end you usually know whether you are looking at a cell, a cable, a box or an instrument.
The habit that matters most is the least technical one. Measure the same things the same way every time, keep the numbers, and compare. A cell that has quietly changed over a year will tell you so in its resistance and its zero long before it tells you in its weighing accuracy.
For plant engineers, service technicians and procurement teams who want their weighing systems to be easy to diagnose, the starting point is documentation and a clear baseline at commissioning. Rudrra Sensors can supply load cells, amplifiers and indicators with the datasheet information this kind of testing depends on, and our team is happy to help interpret field readings when a scale is misbehaving. Reach out with your cell details and measurements, and we will help you find the cause.
