Rice Lake Load Cell Testing: Why the Load Cell Isn’t the Problem

Friday, 4:37 p.m. A filler line in a food plant starts underreading by 4%. The operator swaps in a “known good” load cell, the indicator still drifts, and now there’s a spare load cell on the bench, a $1,800 service bill, and a line that’s been down for six hours. I’ve been on that call more times than I can count.

In my role as a field service engineer for industrial weighing systems, I handle the emergencies. The “we need this running by Monday” calls. When a Rice Lake load cell starts misbehaving, the first instinct is to test the load cell—and that’s exactly where the real problem hides.

Rice Lake Load Cell Testing Usually Focuses on the Wrong Thing

Rice Lake makes solid load cells. Their documentation is good—for what it covers. But “rice lake load cell testing” usually focuses on the cell itself: resistance checks, zero balance, proof load. What it doesn’t cover is the stuff around the cell. The cable. The junction box. The grounding. The splice someone made three years ago and forgot to tell anyone about.

Over the last six years, I’ve handled more than 200 emergency field calls. 47 in 2024 alone. Based on our service logs, the majority were not sensor failures. I don’t say that lightly—it took me a while to accept it, because replacing a sensor feels like the responsible thing to do when a sensor is giving you bad numbers.

The Surface Problem: “The Sensor Is Bad”

When the numbers on the display jump, your brain blames the component that turns weight into a number. The load cell. It’s the same reason people blame a sanitary flow meter for a flow issue, or a hydrostatic level sensor for a tank level issue. The sensor is the only thing you can touch that speaks in engineering units. Everything else—the wiring, the junction box, the cable, the grounding—is invisible.

But a load cell’s full-scale output is tiny. A 2 mV/V load cell with 10 V excitation produces just 20 millivolts at full capacity. That’s roughly the signal strength of a cheap dynamic microphone. You can run a microphone cable beside a fluorescent light ballast and get noise. Now do the same with a load cell cable running beside a variable frequency drive, and guess what happens.

The deeper cause is almost never the cell itself. It’s the signal path. And the more I troubleshoot, the more I keep finding the same pattern in other instruments.

Sanitary Flow Meters and Hydrostatic Level Sensors Fail the Same Way

Take sanitary flow meters. I’ve seen a brand-new electromagnetic flow meter read 7% high because a gasket was protruding into the bore and causing turbulence. The meter was perfect. The installation wasn’t. Or a hydrostatic level sensor in a wastewater tank that drifted every morning—turns out the vent tube that serves as atmospheric reference had a spider’s nest in it. The sensor was trying to measure the difference between liquid pressure and atmospheric pressure, but the second half of that equation was wrong.

These are all examples of the same trap: we test the component, not the system. And the component passes, because the component never failed.

The same logic applies to smaller electrical troubleshooting. If you’ve ever looked up how to test a capacitor with a Fluke multimeter, the standard advice is straightforward: discharge it, switch to capacitance mode, and compare the reading to the printed rating. That’s useful, but it’s also a component-level test. The circuit-level test—checking voltage across the capacitor while it’s powered—is often the more useful one.

Why the Signal Path Is the Real Problem

Load cells aren’t fancy. They’re strain gauges arranged in a Wheatstone bridge. When the bridge is healthy, the output is proportional to load. But that output is small, and it sits right next to industrial sources of electrical noise: motor starters, VFDs, remote radios, lighting ballasts, even other cables in the same conduit. Without proper shielding and grounding, a “bad” signal can look exactly like a “bad” load cell.

Moisture is another classic. A junction box with a cracked seal, a cable with a pinched jacket, a loose gland—all can let water in. Water plus copper terminal strips equals a small battery, and that battery’s voltage gets added to the load cell signal. The result is a zero shift that drifts with humidity. If you only test the cell on the bench (where it’s dry), it’s fine. That doesn’t mean it’s fine in the field.

A few years ago, in August 2021, a dairy plant called with a similar issue. Their tank scale was reading 150 pounds high at 3:00 a.m. The operator did a live load test with a calibration weight, and the error was still there. He was ready to replace the load cell. I asked him to check the junction box before opening the new part. He called back 20 minutes later: there was water inside. The fix took 10 minutes. He later told me that if he’d followed his first instinct, the plant would have been down for another two days waiting for a replacement.

I once spent two hours at another plant chasing a Rice Lake load cell that read perfectly at night and gained 200 pounds during the afternoon. We replaced the cell, then the junction box, then the indicator. Still no luck. Finally, I noticed the cable crossed over an overhead steam line. The heat was changing the cable’s electrical characteristics enough to shift the bridge balance. We rerouted the cable 14 inches away, and the scale never moved again. The load cell wasn’t at fault. Neither was the junction box or the indicator. The installation was.

What I mean is that a “sensor failure” is usually a collection of small installation details that add up to a bad signal. A cold solder joint. A shield grounded at both ends, creating a ground loop. A cable too long for the excitation voltage. A splice in the sense wires. Each one looks like a sensor problem.

The Cost of a Wrong Diagnosis

Let’s put some real numbers on a misdiagnosis.

• New Rice Lake load cell: $450–$1,200 ballpark depending on capacity.
• Emergency field service call: $150–$250/hour plus travel.
• Line downtime: $500–$5,000 per hour in a food or beverage plant.
• Re-calibration after a replacement: $200–$600.

In one case, a plant replaced three Rice Lake load cells before someone finally called us. Total cost of unnecessary parts and labor: about $4,200. The actual fix—cleaning the junction box and replacing a damaged cable—took 40 minutes and cost $80 in parts. And the plant lost 14 hours of production for something that was visible on a simple insulation resistance test.

Here’s the part that hurts more than the downtime: if the load cell isn’t actually broken, replacing it doesn’t fix the underlying issue. The new cell will fail the same way if moisture is still in the junction box, or the ground loop is still there, or the cable is still damaged. You’ve spent money and gained nothing.

If the scale is used for trade—selling by weight—then you’re also on the hook for compliance. According to NIST Handbook 44, commercial weighing devices in legal trade must be accurate, repeatable, and subject to inspection. A repair that doesn’t fix the error is not just expensive; it’s a legal risk.

Do the Right Rice Lake Load Cell Testing

Before you order a replacement, run a proper test sequence. Not because I’m telling you to; because it’s cheaper and faster.

1. Inspect mechanically. Is the load cell side-loaded? Is there material built up under the bin? Are the mounting bolts torqued correctly? A bind of 1% of capacity looks like a zero shift.
2. Check the cable and junction box. Use an insulation resistance tester (megger) between conductors and shield. The spec sheet usually says 1,000 MΩ or higher. If you’re below that, find out why before you test the cell again.
3. Measure bridge resistance. Rice Lake load cells ship with a factory test certificate showing measured input and output resistance. Compare your readings to that certificate. If a 350 Ω bridge reads 355 Ω, don’t assume it’s still fine.
4. Watch the indicator. If the display is unstable, put your meter on the indicator’s test points and see if the signal is stable there. Sometimes the A/D converter or the power supply is the real problem.
5. Then, and only then, decide if the load cell needs replacing.

If you’re not comfortable doing that, a certified technician can. But ask them to prove the diagnosis before they quote a replacement. A good tech will show you the failed insulation test on the cable.

A standardized sequence doesn’t just save money; it saves time—which, in an emergency, is the same thing. Switching from “replace the sensor and see” to “test the loop first” cut our average troubleshooting time from 3.2 hours to 41 minutes across the last 30 emergency calls.

The Quiet Satisfaction of a Wrong Guess

After the spider’s nest story—the hydrostatic level sensor, not the load cell—I had a moment of doubt. The plant manager had already ordered a replacement sensor, and I’d told him not to open the box yet. If I’d been wrong, he’d have every right to be angry. I wasn’t, but the 40 minutes before the tank level matched the hand tape were not comfortable. When the level settled at 8.42 feet against a hand-measured 8.40, I felt that quiet satisfaction you get when the obvious answer turns out to be the wrong one, and the right one is hiding in plain sight.

Next time a Rice Lake load cell—or any sensor—starts acting up, don’t ask “what’s wrong with the sensor?” Ask “what is this sensor trying to tell me that I haven’t checked yet?” The answer is usually in the system, not the part.