Don't Calibrate Your CMM Like It's 1999: A Technician's Quality Hygiene Checklist

If you're the person who signs off on inspection reports using a Hexagon CMM or portable CMM, you've probably had that sinking feeling. The part passed first thing in the morning. By lunch, it's out of spec. You re-measure. Same result. You check the calibration sticker. Everything's current. So what gives?

This checklist is for the quality inspector who's been burned by a 'passing' instrument that wasn't actually ready to measure. I'll walk you through 5 steps that go beyond the annual calibration certificate. Use this before any critical measurement session, especially if you're mixing Hexagon laser trackers with shop-floor micrometers. Let's cut the crap and get to it.

Step 1: Check Your Environment, Not Just the Device

You'd think temperature control is obvious. But the most frustrating part of CMM work? The part is fine. The lab is fine. You walk the portable CMM onto the factory floor, and suddenly your 0.005mm tolerance is a fantasy.

What to do:

• Temperature gradient: Put a thermometer on the part, the CMM bed, and the lab air. If any two differ by more than 1^oC, wait 30 minutes. Seriously. I've rejected a batch of 200 machined parts that measured perfect at 20^oC and failed at 24^oC. The vendor claimed 'industry standard.' Their standard wasn't ours.
• Vibration sources: If you're near a stamping press or a forklift path, your Hexagon support team will tell you the same thing I'm telling you: stop measuring. Vibration kills CMM accuracy. I've seen it cost a shop $22,000 in rework.
• Lighting (for visual tools): If you're using a comparator or microscope, shadows lie. Use a ring light. Check at two angles. Trust me on this one.

Step 2: The 'Quick' Check That Everyone Skips

Most operators run a standard calibration sphere check. That's fine. But here's the step that caught me out for a year: check the machine's compensation map.

Every CMM has a volumetric compensation. When was the last time you verified it wasn't corrupted? Here's a practical test:

1. Measure a known artifact (like a ring gauge or a set of ball micrometer standards) in three different positions on the table: center, front-left, back-right.
2. If the measurements vary by more than your tolerance, you need to re-run the compensation routine. I've had machines pass the daily touch-probe check but fail this test because someone accidentally loaded the wrong compensation file after a software update.

It's a 10-minute test. It saved me from shipping 8,000 parts that would've been out of spec at the customer's receiving dock. Good thing I found it before they did.

Step 3: Stop Trusting the Calibration Certificate Blindly

I have mixed feelings about OEM calibration certificates. On one hand, they're officially correct. On the other, they certify the instrument under ideal conditions—not your conditions. A certificate for an outside micrometer set tells you it was accurate at the lab. It doesn't tell you if the anvil faces are worn from 5 years of shop use.

Your checklist item:

• Gauge R&R on your actual parts: Grab 3 parts that cover the range of your typical measurement. Measure each part 3 times, with 3 different operators (or the same operator, if you're solo). Calculate the %GRR. If it's over 30%, the instrument isn't the problem—the process is.
• Visual inspection of anvils and probes: Use a magnifier. Look for nicks, wear, or debris. I once caught a ball micrometer that had a tiny burr on the spindle face. It was reading 0.003mm off on everything. The certificate was from 3 months ago. The burr was from last week.

Step 4: The 'How to Calibrate pH Meter Mettler Toledo' Trap

Alright, this one seems off-topic for a metrology article, but stick with me. If your lab uses a pH meter for any chemical testing (cleaning validation, coolant pH, etc.), you might face the same calibration maze. A Mettler Toledo pH meter needs proper buffer solutions. But here's the kicker: the buffer solution itself expires.

Your step-by-step:

1. Check buffer expiry dates. If your buffer is 6 months past its date, your calibration is garbage. I've rejected calibration results from a vendor who claimed a 'fresh' calibration with 14-month-old buffers. They didn't learn a thing.
2. Single-point vs. multi-point. A two-point calibration (pH 7 and pH 4 or 10) is standard. But if you're measuring extreme values, add a third point. The instruction manual says so, but most people skip it.
3. Electrode condition. If the electrode is dry or cracked, no calibration will fix it. I've had a technician spend 30 minutes 'calibrating' a pH meter that needed a $50 replacement electrode. He could've saved that time by looking at the electrode tip first.

The principle is the same for CMMs: the calibration is only as good as the reference standards you use. If your reference sphere is damaged, your CMM is lying to you. Check your references before you check anything else.

Step 5: Document the 'Why' Behind the Measurements

Here's the one thing nobody does: document why you chose a particular method or instrument. I started doing this after a quality audit in Q1 2024 cost us two days of back-and-forth. The auditor asked, 'Why did you use a portable CMM for this feature instead of a fixed CMM?' I had an answer, but no evidence that I considered the alternative.

Create a quick log entry for each critical measurement:

• Instrument selected (e.g., Hexagon portable CMM vs. fixed CMM)
• Reason (e.g., 'Part too large for fixed CMM, tolerances within PCT spec')
• Calibration status (certificate valid, last daily check passed)
• Environmental condition at time of measurement (temp, humidity, vibration source)

It's overkill 90% of the time. But that 10% of the time when someone questions your data? It saves your career.

Common Mistakes to Avoid (From Someone Who Made Them)

After 5 years of reviewing inspection reports, here are the biggest recurring screw-ups:

• Only checking at zero point. You zero your micrometer and it's perfect. But at 25mm, it's 0.01mm off. Check at least 3 points across the range.
• Skipping the reference artifact. 'My CMM is new, it doesn't need a calibration check today.' That's how I found a loose Hexagon support bracket on a brand-new machine. It was shipping parts for two weeks with a 15-micron error.
• Trusting the firmware update. Not all updates improve accuracy. Some change how the probe compensates. Always verify with a known standard after an update.

Small-clients note: If a calibration vendor treats your one-off ball micrometer or single pH meter like it's beneath them, go elsewhere. I've had a small $200 order from a startup turn into a $20,000 annual contract because they remembered who took them seriously. Today's small job is tomorrow's big account—treat it that way.

This isn't a perfect system. You'll still have bad days. But if you follow these steps, you'll have fewer of them. And when you do have a failure, you'll have the data to prove it wasn't your fault. That's worth a lot.