How to Test a Load Cell: A Quality Inspector's Guide to Not Getting Burned

Everything You Wanted to Know About Testing Load Cells (But Were Afraid to Ask)

I review incoming equipment for a mid-sized manufacturing plant—roughly 200+ unique items a year, from benchtop centrifuges to 117 multimeters. A lot of what lands on my bench comes from Rice Lake. I've rejected about 12% of first deliveries this year due to specs being off, documentation missing, or—most commonly—load cells that don't perform as advertised.

If you're searching "how to test a load cell" or "Rice Lake load cell troubleshooting," you're probably in one of two camps: you've got a sensor that's reading weird, or you're trying to qualify a new supplier. Either way, here's what I've learned the hard way.

1. What's the simplest test to see if a load cell is broken?

Grab a multimeter—ideally a decent one (I use a Fluke 117, but even a $50 model works). Check the resistance between the excitation wires (usually red and black on a Rice Lake load cell). A healthy cell reads between 350 and 450 ohms for most standard units. Then check the signal wires (green and white)—you're looking for the same range. If you get an open circuit or a short, the cell is dead.

One thing I've learned (or rather, had burned into my memory after rejecting a batch of 50 units): you can also check resistance between the excitation and the cell body. Anything less than 5000 megohms means moisture has gotten in. That cell is on borrowed time.

2. How do I zero a Rice Lake load cell before testing?

You don't. At least, not with a screwdriver. Digital load cells like Rice Lake's iQUBE series zero themselves when you power them up. For analog cells, the zero offset is set at the factory, and you shouldn't touch it unless you have a calibration lab. What you should do is check the zero balance: power the indicator, make sure no weight is applied, and see what it reads. If it's not zero or within the spec (usually 1% of rated output), you've got a problem.

Everything I'd read before my first big project said "just re-zero it." In practice, messing with the zero pot on an analog Rice Lake cell caused a drift that took three calibration cycles to undo. Don't be me.

3. My load cell reads fine empty, but the numbers drift. What's going on?

Thermal drift is the most common culprit. If your plant runs at 40°C during the day and 20°C at night, the zero can shift. A quality load cell should stay within 0.001% of full scale per °C. Anything more, and either the cell is cheap or it's damaged. Also check your wiring—I caught a batch of bad cables in Q1 2024 (from a different vendor, not Rice Lake) where the insulation cracked under heat, causing intermittent shorts.

That issue cost us a $22,000 redo and delayed a launch by two weeks. Now every cable gets a heat test before acceptance.

4. What's the fail rate for new load cells, really?

I've been tracking this for four years. For brand-name cells like Rice Lake, the initial failure rate is about 0.5–1%. That means for every 100 you install, maybe one is DOA. But the curve changes after six months: if a cell survives its first thermal cycle (first winter in a cold plant, or first summer in a hot one), it's likely to last its rated lifespan. If you see a failure between month 3 and month 6, it's usually an installation issue—overtorquing, moisture, or electrical noise.

The conventional wisdom is that load cells are bulletproof. My experience with 200+ inspections says otherwise.

5. How do I test a load cell without a service manual?

You can still do a basic functional test. Hook the cell to a known-good indicator—the Rice Lake 920i if you have one. Apply a known weight (something traceable, like a 50 lb calibration weight—don't guess with a sack of cement). Check the display: the reading should match within the cell's rated accuracy. For a 500 lb cell rated at 0.05%, that means 50 lbs should read 49.975 to 50.025. If it's farther off, either the cell is drifting or the indicator needs calibration (note to self: always verify the instrument was on the right setting first—I wasted a morning once because the indicator was in grams instead of pounds).

6. When should I just replace the load cell instead of troubleshooting?

If a Rice Lake load cell fails the resistance test (open or short) or the insulation resistance is below 5000 megohms, it's done. Don't try to fix it—you can't reliably repair a sealed strain gauge. Also, if the cell has been physically struck (say, a forklift hit the scale), replace it. Internal damage isn't always visible, but it will show up as nonlinearity.

Here's where the time-certainty thing kicks in: I once spent three days testing a cell that had been water-damaged, trying to "save" it. The replacement cost $400. The production delay cost about $4,000. I do the math differently now.

7. What's the trick to getting accurate load cell readings on a Rice Lake system?

Three things, in order of importance: (1) Good wiring—use shielded twisted-pair cable, and keep it away from VFD lines. (2) Proper grounding—the cell body should be grounded, but the signal ground should float. Mix those up, and you get noise. (3) Regular verification—not full calibration, just a known-weight check every week. An ounce of prevention, as they say.

The upside of Rice Lake's ecosystem is that their indicators (like the 920i) do a lot of the filtering for you. The risk is relying on that as a crutch for bad installation. I've seen a plant spend $18,000 on a system that performed worse than a $5,000 one because they skipped the grounding step.

8. How do I check the accuracy of a benchtop centrifuge if it uses a load cell for balancing?

That's a niche question, but one I've had to answer. A benchtop centrifuge typically uses a load cell to detect imbalance before spinning up. The test is simple: place a known weight (like a 50g certified mass) in one tube slot, and see if the centrifuge gives an imbalance warning. If it does, the cell is working. If it doesn't trigger until the weight is 100g or more, the sensitivity is off—that's a safety risk.

I referenced this against the manufacturer's spec once, and the tolerance was ±5g for the warning threshold. If you're sourcing a used centrifuge (or just inherited one), this is the single most important test.