The $3,200 Mistake: Rice Lake Bench Scales, a Blade Micrometer, and a Fluke 117 True RMS Multimeter

The Day the Scale Lied to Me

I remember the exact moment: Friday, 3:40 p.m., September 2022. The customer's quality engineer was standing next to me, holding a part that should have weighed 1.482 kilograms. The Rice Lake bench scale showed 1.478. We re-weighed the same part four times and got four different answers. That's when I knew the 400-piece order was in trouble.

I'm the calibration and test equipment guy at a small contract machine shop. I've handled this kind of work for eight years. In that time I have personally made and documented 14 significant mistakes, totaling roughly $6,200 in wasted budget. A big chunk of that came from one week in September.

The Setup I Trusted

We had the right tools. Rice Lake bench scales on the QC bench. A complete micrometer set in the drawer, including a blade micrometer for the slot-width check. A Fluke 117 true RMS multimeter for electrical troubleshooting. The scale had a calibration sticker that was 60 days old. The micrometer set was in its foam case, in the right order. Everything looked the way a QC bench should look.

The job was 400 precision-machined brackets. The spec said 100% weight verification on a bench scale and a slot-width check with a blade micrometer. I checked the first batch, recorded acceptable weights, and moved the parts to the pallet. Maybe 50 pieces in, the customer's engineer did his own sample and got lower readings. “Your scale or your procedure,” he said. “Either way, this lot is on hold.”

I doubled down. I re-tared the scale. I re-weighed a part. It still didn't line up. I told him we would re-check in the morning. In the back of my head, I remembered that a forklift had bumped the bench scale platform the week before. I ignored it. The sticker said the scale was calibrated. I did not want to lose the day.

How to Use a Fluke 117 True RMS Multimeter (and How Not To)

The next morning, I stopped re-weighing parts and started testing the scale. I grabbed the Fluke 117. At first I used it wrong. I set it to AC volts because I wanted to see if there was electrical noise. The reading jumped around and told me nothing. True RMS does not help if you are measuring the wrong type of signal.

The Fluke 117 true RMS multimeter is good at measuring non-sinusoidal AC waveforms. That matters for VFDs, motors, and dirty power. But the load cell circuit on a Rice Lake bench scale is DC. The 117 does DC too. You just have to turn the dial to the DC volts position. I know that sounds like a beginner move. It was a beginner move. I let the phrase true RMS make me feel like I was measuring something sophisticated. I wasn't.

Once I switched to DC volts, I checked the excitation voltage at the indicator: 9.93 VDC. Then I walked to the load cell junction box and checked again: 7.8 VDC. That voltage drop was the smoking gun. The cable had a broken conductor, probably from the forklift bump. The load cell still produced a reading, but it drifted depending on how the cable moved.

I called our rep and then Rice Lake support. They asked for the part number on the cable assembly. It was the 182591. I ordered the Rice Lake Weighing Systems 182591 service kit on Tuesday. The part that hurts is knowing this should have been a 20-minute diagnosis before the production run.

The Blade Micrometer Was the Second Failure

While the scale was in question, I tried to keep moving on the slot-width checks. The slot was deeper than the anvil of a standard micrometer, so I needed the blade micrometer from our micrometer set. I picked it up, found the slot, turned the thimble until it clicked, and wrote down the reading.

Here is the lesson: I did not check the blade micrometer against its standard first. With any micrometer set, you verify zero with the included standard before use. With a blade micrometer, this matters even more because the blade faces can collect chips and wear unevenly. I skipped the check because the previous guy had used it earlier that morning. The previous guy probably checked it. I did not verify. There was a tiny piece of swarf under the anvil face. It changed the reading by 0.0015 inches. Our tolerance was ±0.002. That is enormous.

The customer's engineer watched me record 0.3130 inch. He picked up the same blade micrometer, cleaned the faces, checked zero against the standard, and measured the same slot. He got 0.3115. The part measured 0.3110 on a properly zeroed tool. My reading was wrong. His first reading was wrong. Only the standard check made it right. We had to remeasure the slot on every part. That was my fault and my cost.

What It Cost

The final number was about $3,200. Actually, $3,180 in extra labor and calibration time, plus expedited shipping for the 182591 kit, plus a week of delay. The customer didn't cancel the order, but they added a six-month audit requirement. That was the part that stayed with me.

I learned a real thing about the old belief that a current calibration sticker means an instrument is correct. That belief comes from an era when instruments drifted slowly and predictably. A modern digital bench scale can be spot-on in the morning and off by half a gram in the afternoon if a cable gets pinched. The sticker doesn't catch that. The verification does.

If you use a bench scale for something as simple as postage, the same logic applies. According to USPS (usps.com), as of January 2025, a First-Class Mail letter (1 oz) is $0.73, and a large envelope (1 oz) is $1.50. If the scale drifts by a few grams, you either overpay or get hit with postage due. The stakes are lower than a 400-piece order, but the principle is the same.

The Checklist I Use Now

After that week, I wrote out a pre-run checklist. Not a fancy one. It fits on one page.

• Check the bench scale with a known test weight before the first part, not just zero.
• If the scale recently took a bump, inspect the cable and connector before trusting any reading.
• Check the blade micrometer zero against its standard. Clean the anvil faces first.
• When using a Fluke 117 true RMS multimeter, ask what I am actually measuring. If it's DC, use DC volts. True RMS only matters for AC.
• If the first measurement and the sample measurement disagree, stop. Re-measuring the same part with the same assumption is not the answer.

It's not a workflow masterpiece. It's a scar. But in the 18 months since, we have caught 47 potential issues before they became lots on hold. Some of them were small. A few would have been another $3,200 mistake.

This worked for us because we are a mid-size job shop with predictable production runs. If you're in food processing, pharmaceuticals, or legal-for-trade weighing, your process needs to be stricter than mine. I can only speak to domestic operations and a limited sample of about 200 production lots. Your mileage may vary. But I do know this: the scale wasn't the main problem. The main problem was me assuming the equipment was right just because it looked right.

Now I check. That's it.