Don't Replace That Checkweigher Before You Test the Ground: A $3,200 Lesson

It passed the morning shift. Then, around 2:15 p.m., the automated checkweigher on line 7 started reading eighteen grams heavy. Not by much. But for that package, the tolerance was ±12 grams, so the reject gate kicked every third box into the bin. By 4 p.m., we had a pallet of "rejects" that were probably fine, and a line supervisor asking me why the Rice Lake unit was suddenly garbage.

If you're searching for automated Rice Lake checkweigher prices because yours just started doing something similar, I get it. I've been there. But I'm going to ask you to hold that thought for a few minutes. The checkweigher probably isn't the problem. I learned that the hard way—twice—and the second time cost me about $3,200 and a week of awkward conversations.

The Surface Problem: "The Load Cell Is Bad"

When a checkweigher drifts, the first instinct is to blame the load cell. It's the most expensive single component in the system, and it's the one that "wears out" in people's minds. I've written more root-cause reports than I can count where the initial assumption was "load cell failure" and the final finding was something else entirely.

Here's the thought process I used to hear from myself: load cell output is supposed to be stable. The indicator takes that tiny millivolt signal, converts it to weight, and displays it. If the displayed weight drifts, the load cell must be drifting. Therefore, replace the load cell.

That logic sounds airtight until you realize what the load cell actually is: a strain gauge bridge that produces a differential voltage measured in millivolts, powered by a reference excitation voltage. If the excitation voltage is stable but the signal path is compromised, the reading will shift no matter how new the load cell is. The problem isn't the sensor. It's the circuit around it. If this scale is used for commercial transactions, NIST Handbook 44 (2024) defines the tolerance classes. Verify the current edition at nist.gov before you rely on it.

The Deep Cause: Ground Loops, Moisture, and the Ugly Junction Box

The first time this happened to me—a food packaging line in 2019—we replaced a perfectly good load cell. The original unit went back to the manufacturer for testing, and it came back with a certificate saying "within tolerance." I felt like an idiot. Actually, I was an idiot, because I'd skipped the troubleshooting steps that every manual lists in the first ten pages.

Let me walk through what I should have done.

1. The Ground Is the First Suspect

I now use a digital earth resistance tester before I even open the junction box. The first time I did, the display read 23.8 ohms on what was supposed to be a dedicated ground rod. That's high. Most plant electrical specs I've worked with call for less than 10 ohms, and some of the better ones call for less than 5. High earth resistance creates a poor reference for the shield and can let noise into the load cell signal.

What I mean is that the shield's entire job is to give induced noise a path to ground. If that path is high impedance, the noise stays in the signal. That noise doesn't always show up as wild swings; sometimes it just looks like a slow drift that follows the temperature or the motor load. When I compared a good ground and a bad ground side by side—same scale model, same age, different wiring—I finally understood why grounding is a specification, not a suggestion. The ground was the problem, not the load cell.

2. Insulation Breakdown Hides in Plain Sight

The phrase "insulation tester vs megger" comes up a lot in our shop, because people wonder if they're the same thing. The short answer: a megger is a specific brand name that became the generic word for an insulation resistance tester. The longer answer, and the one that matters, is that not all insulation testers output the same test voltage, and using the wrong voltage can damage sensitive electronics.

For load cell cables and junction boxes, I test insulation at 50V or 100V, not 500V or 1000V. A standard high-voltage megger is fine for motor windings, but it's too aggressive for a low-voltage signal cable connected to a strain gauge bridge. I've seen brand-new junction boxes fail an insulation test because moisture had gotten into a cable gland. An insulation tester catches that. A multimeter alone won't.

And speaking of multimeters—the old 73 Series II multimeter we've had on the bench since before I started here is still good enough for a quick load cell output test. Disconnect the load cell from the indicator, connect the multimeter to the output leads, and measure the millivolt signal under no load. It won't tell you everything, but it tells you if the bridge is obviously dead or live. I use ours all the time.

3. The Manual Is Not a Trap

I know "read the manual" sounds like the most generic advice in the world. But the Rice Lake manuals I've downloaded are genuinely useful. The installation section for most systems specifies grounding, cable routing, and wiring in a way that would have saved me the 2019 fiasco. I pulled the Rice Lake manuals for all our scales from rice-lake.com after the second incident, printed them, and put them in a binder. That binder has prevented more unplanned downtime than any calibration service I've used.

I don't have hard data on how many companies replace load cells that were never bad, but based on my own repair records, my sense is that it's at least half of the "failed" load cells I was called in to fix. That number is anecdotal—our facility and a few customer sites—so take it for what it's worth.

The Real Cost of Replacing the Wrong Part

The second mistake happened in March 2022. Same symptom, different line. I was sure it was the load cell this time because the first one had been "fine," but this unit was old and had been through a washdown incident. I ordered a replacement, paid for expedited shipping, and scheduled a calibration tech to install and verify it.

Total so far: $2,340 for the load cell. $380 for overnight freight. $480 for the calibration visit. That's $3,200 before we counted the 45 minutes of production time during the swap.

The new load cell fixed the drift for about two hours. Then it came back. The calibration tech looked at the junction box—ugh, the junction box—muttered something about condensation, and pulled the plug. There was water inside the cable gland. The actual fix took 20 minutes of cleaning and new silicone sealant. The old load cell was fine. The new one is now a spare.

Here's the part that still bothers me: if I'd spent 20 minutes with the digital earth resistance tester and the insulation tester first, I would have found the water. Instead, I trusted my "experience" and ignored the basics. The most frustrating part is that the checkweigher would pass a calibration check at startup and drift an hour later. You'd think a freshly calibrated load cell would hold its zero, but it can't if the shield is picking up noise from a bad ground. The load cell didn't fail. The environment around it did.

The 30-Minute Test I Wish I'd Done

So, before you compare automated Rice Lake checkweigher prices, do this. It takes about 30 minutes. The short version: three things. The ground. The insulation. The connections. In that order.

1. Test the earth ground. Use a digital earth resistance tester to verify the ground rod resistance. If it's above the value in the manual or above 10 ohms, fix the ground first.
2. Inspect the junction box and cable entries. Look for moisture, corrosion, and loose terminals. If you've ever had direct washdown near the scale, assume there's water until proven otherwise.
3. Run an insulation test. Use an insulation tester that can be set to a low test voltage. In the "insulation tester vs megger" debate, remember: the right voltage matters more than the brand name.
4. Check the load cell output with a multimeter. A 73 Series II multimeter or equivalent is perfect for this. Compare the no-load and known-load readings to the values in the Rice Lake manual for your specific model.
5. Do a temperature soak test. Let the system sit at operating temperature for 30 minutes and watch the indicator. If the readout drifts with temperature, you have a signal integrity problem, not a load cell problem.

This checklist is the reason I now maintain our team's troubleshooting documents instead of just fixing things. It won't catch every possible failure mode, but it will catch the ones I keep seeing in our industry.

One caveat: this worked for us because we're an indoor facility with relatively stable conditions and controlled washdown procedures. If you're in a high-vibration outdoor setting or a chemical plant, your next step after insulation testing might be a protective cover or a different cable routing. Your mileage may vary.

When Buying New Is Actually Right

I don't want to sound like I think load cells never fail. They do. Bearing failure, gross overloading, lightning strikes—that's a real replacement trigger. But when the failure is drift on an automated checkweigher that has otherwise been stable, the odds favor an electrical or environmental cause.

The other time to buy is when you're facing a genuine deadline and can't afford to let a 30-minute test eat into your shift. I understand that. In Q4 2024, we paid a $460 rush fee and a 2-day calibration slot because the production window was worth $15,000. That's the right kind of expensive: buying certainty when you don't have time to troubleshoot. The wrong kind is replacing parts before you've spent 30 minutes testing.

What I don't stand by is replacing parts to save time, only to find out the real issue was something a digital earth resistance tester would have found in ten minutes. The expensive route should be the one you choose after diagnosis, not before it.

(Note to self: I really should write that ground test sequence into our PM schedule before someone else learns this the hard way.)

So, if the search term that brought you here was "automated rice lake checkweigher prices," I hope you now have a different question in mind: "what does the manual say about grounding?" That's the question that saved me $3,200. It might save you more.