Calibrated Does Not Mean Accurate. Here's What to Check.

I'm going to say something that makes my calibration engineer wince: a calibration sheet doesn't tell you whether a meter is accurate. It tells you whether a meter was compared to a standard on a certain date. Those are not the same thing.

Here at Extech, I'm the quality manager. I review every product manual and calibration sheet that ships out—roughly 200 unique items a year. In the last four years I've rejected more first-pass documents than I'd like to admit. Last year it was around 11%. I might be overshooting that number; I'd have to pull the spreadsheet. The point is, 'calibrated' appears on paper far more often than it is earned.

A meter can display 12.6 amps. So can a meter that is wrong.

You buy an Extech MA445 clamp meter. You open the box, switch it to AC amps, and read 12.6 A. That number is only useful if you know the maximum error. But what's in the manual? For many meters, the basic accuracy spec is something like ±1.8% reading + 10 digits. At 12.6 A, that means the real current could be anywhere from about 12.3 A to 12.9 A. Period. And that doesn't yet include calibration uncertainty.

The MA445 is a true RMS clamp meter, which means it handles distorted waveforms better than an average-rectifier meter. But true RMS doesn't mean true value. It means the internal math is designed to parse non-sinusoidal signals. That math is only as good as the calibration that verifies each range. A meter with an unverified calibration returns a false number with perfect confidence.

Same logic applies to an Extech pinless moisture meter. It estimates moisture by measuring capacitance around the sensor pad. That is an indirect measurement. It depends on material density, temperature, and whether there is a metal stud behind the drywall. If the instrument was not checked against a reference block at a known density, a 'calibrated' reading can drift by several percentage points and you won't see it coming.

Even a mechanical micrometer has the same DNA problem. You can read to 0.001 mm, but if the zero is off by 0.01 mm, the whole measurement chain is contaminated. I've seen a $30 micrometer with 0.01 mm error and a $400 one with exactly the same problem. The difference was documentation and maintenance, not the sticker.

Why 'calibrated' means worse than you think

The root cause isn't the electronics. It's the information gap. The buyer asks, 'Is it calibrated?' The seller says 'yes' because the unit went through a process. But the certificate doesn't say how close the unit is to the true value after calibration, or how far away it was before the calibration. That's the hidden layer.

Everything I'd read early in my career said accuracy is a property of the instrument. It's not. Accuracy is the result of instrument, calibration, and operator together. The instrument is only the middle part.

What I mean is that 'calibrated' is a process claim, not a performance claim. A certificate can show dates, initials, and a pass stamp. It may not show the size of the error unless you know where to look. Under ISO/IEC 17025, a calibration certificate is supposed to include measurement uncertainty. But not every certificate is issued under that standard, and even those that are may bury the uncertainty in a footnote.

The cost usually shows up later

In 2023 we approved a batch of 8,000 thermocouple probes based on a summary calibration sheet from an outside lab. The first page said 'Passed.' I didn't request the detailed as-found data until two months later, when a customer reported scattered temperature readings. A fifth of the probes were on the high edge of tolerance. The lab had calibrated them, but the summary sheet didn't give enough data to show how close they were to the edge. The redo cost us $22,000 and delayed a product launch by three weeks. I still kick myself for not asking about the as-found values.

My senior technician always told me to check the uncertainty block before I check the pass/fail block. I ignored that once. I don't do it anymore.

Saving $80 on a meter without calibration data is not a saving. It's an upfront fee for a later surprise. I've learned to ask 'what is NOT on this sheet?' before I ask 'is this calibrated?'

Hidden calibration fees are the measurement world's equivalent of a resort fee. The advertised price looks clean. The final bill doesn't.

How to read a Rice Lake weighing systems calibration sheet like a quality inspector

When someone hands me a Rice Lake weighing systems calibration sheet, I don't read it from top to bottom. The order matters, and the order is: as-found data, uncertainty, standards traceability.

The order matters

• As Found before As Left. The as-found column shows what the scale did before any adjustment. If it was sitting at the edge of tolerance, the load cells are aging. The as-left column tells you how it left the lab; the as-found column tells you whether it will survive until the next calibration.
• Uncertainty before Pass/Fail. A 'PASS' with no uncertainty is not a measurement. It's an opinion. In a good report, the uncertainty should be significantly smaller than the tolerance. In a vague one, it's either missing or too large to mean anything.
• Standards traceability before everything else. The sheet should list an ID for the reference weights and a note about traceability to NIST. If that block is empty, the readings have no foundation.

For a scale used in commercial transactions, the tolerance should also align with NIST Handbook 44. If a sheet says 'within manufacturer spec' but doesn't cite the accuracy class or scale division, that's a yellow flag. The same idea applies to a flow meter manual: find the K-factor section. If it only says 'factory calibrated,' call the manufacturer and ask for the K-factor and the date it was measured. A flow meter is a calibrated pipe. Without that data, it's just a pipe.

People think a higher price tag is what makes an instrument accurate. Actually, the price pays for the design, the components, and the guarantee that somebody checked it. The accuracy comes from the traceable calibration data that sits next to it. The most expensive meter in the world is a vague paperweight if you can't verify its last calibration.

The nice part about this approach is that it works for any instrument. The Extech MA445 clamp meter, the pinless moisture meter, the micrometer, the flow meter, the floor scale—they all talk in the same language. Some instruments are better documented than others. The documentation is what creates trust. That's why transparent data is more important than a low sticker price.

The short version

The meter isn't lying. The paperwork is incomplete. Before you trust any reading, check the calibration sheet for the as-found data, the uncertainty, and the traceability of the standards used. If those three blocks are clear, you have a real measurement. If they aren't, you have a guess with a nice font.

Choose vendors who give you the full documents without making you ask. In the long run, that's the cheapest calibration there is.