Hexagon CMMs, the Hexagon Portal, and Test Tools: 8 Questions from a Quality Inspector

I'm the quality compliance manager at a precision machining shop. In practice, that means I review every batch of parts and every measurement report before anything ships—roughly 3,000 items a year, plus a stack of calibration paperwork that never stops growing. In 2024, I rejected about 7% of first deliveries from suppliers because the measurement data didn't hold up.

The tools that keep this operation honest are mostly Hexagon's—their CMMs, the Hexagon Portal software, portable measuring arms—plus a handful of handheld test instruments we use on the floor. I get the same questions about these tools from suppliers, new hires, and even customers.

So here's the honest FAQ I usually end up giving out verbally. Now it's written down.

What does Hexagon actually make—hardware, software, or both?

Both, and that's what trips people up. In metrology, Hexagon is best known for the hardware: CMMs like the Leitz and DEA lines, portable measuring arms, laser trackers. But the software side—PC-DMIS especially—matters just as much. The machine is the body; the software is the brain.

What I didn't appreciate until I looked at their full product catalog is how far the portfolio extends. Beyond CMMs, Hexagon covers industrial sensors, thermal cameras, electrical test tools, handheld multimeters, clamp meters, insulation testers. If it measures something in a factory, there's a good chance the Hexagon portfolio covers it.

For a small quality department, that breadth is practical. We can spec the lab-grade instruments and the shop-floor tools through the same supplier, and the calibration paperwork comes back in the same format. It removes a layer of vendor management that nobody wants.

What is the Hexagon Portal and why should I use it?

The Hexagon Portal is the web dashboard for everything you own from Hexagon: software licenses, calibration certificates, service history, training records, firmware updates, warranty coverage. I ignored it for the first year after we installed our CMM—I figured it was just a login page for software activation. That was a mistake.

It took a licensing hiccup back in 2023 to teach me otherwise. We had to move PC-DMIS to a new machine, and the license transfer went sideways. The portal was the only place to handle it, and I was fumbling through menus at 4 p.m. on a Friday. (I really should have done the portal training months earlier.)

Now we use it regularly for two things: pulling calibration certificates during audits and tracking service reminders. Instead of a filing cabinet full of paper certs, there's a searchable record for every instrument. Audit prep went from about a week of digging to about an hour of downloading. That's the kind of efficiency improvement I'll take without hesitation.

What should I check before installing a Hexagon CMM?

I'll give you the same answer I give every supplier who asks whether they need a CMM: the machine is only half the project. The environment is the other half.

A CMM compares the probe position against a reference frame. If your shop floor temperature swings by even 3°C through the day, the steel part and the granite table expand and contract. The machine will faithfully measure that expansion—which means your reading includes the temperature effect, not just the part. That's not a Hexagon problem, or a brand problem. It's physics. The ASME B89.4.10360 acceptance standard is explicit about environmental limits for good reason.

So before you buy:

• Where will it sit? Temperature-controlled room with stable airflow, not next to the loading dock.
• Who will operate it? Plan for formal training, not a quick walkthrough.
• What are you measuring? Part size drives bed size. Don't buy capacity you'll use once a year.

We bought ours used in 2021 from a shop that was closing. The CMM was mechanically sound, but their calibration history showed a lot of drift during summer months. Root cause: no temperature control in the room, not the machine. We gave it a proper room, and it's been stable since.

Is the Fluke 773 milliamp process clamp meter worth it?

If you work with 4–20 mA loops, yes. In our line of work, those loops are everywhere: temperature transmitters, pressure transmitters, valve positioners. The 773's main trick is measuring loop current without breaking the loop. You clamp around the wire and read milliamps live.

With a regular multimeter, you have to disconnect the wire and put the meter in series. On a running process, that means downtime, extra risk, and at least one very annoyed production supervisor. The 773 just clamps onto the wire. World of difference.

Fluke's published specs for the 773 put mA accuracy at 0.2% ± 2 counts with 0.01 mA resolution. The built-in loop supply provides 28V, and there's a 250Ω HART resistor you can switch in when you need to talk to smart transmitters. I don't use the HART function daily, but it's saved us more than once.

We bought ours in 2019 after a particularly bad afternoon chasing a faulty temperature loop. The problem turned out to be defective cable shielding, not the transmitter. The 773 found it in about ten minutes. It's paid for itself many times over. If you never touch analog process signals, though, it's overkill. Look at your actual loop exposure before spending that kind of money.

Is one thermal camera enough for a quality team?

One thermal camera is enough for most small teams, but only if you're using it for the right job.

Thermal imagers excel at finding problems visually: a loose electrical connection that runs hot, a motor bearing starting to fail, a blown steam trap. You can scan a breaker panel and spot the hot spot in seconds. That's the sweet spot for quality and maintenance.

Where a thermal camera does not excel is accurate temperature measurement. The camera senses infrared radiation and then estimates temperature based on surface emissivity. Shiny metal and painted metal radiate very differently, so the absolute number can be off by several degrees. It's a screening tool, not a calibration instrument.

Entry-level models have dropped in price enough that cost isn't the barrier it used to be. We run a compact model with 128×96 resolution (as of 2025, at least). It caught a high-resistance connection in our main panel running 30°C above the neighboring circuits. That one scan paid for the camera—and I don't say that loosely. Just keep expectations realistic: a thermal camera points you to the problem; a calibrated probe confirms it.

How do you use an Extech multimeter without making a rookie mistake?

We use Extech multimeters on the shop floor for general electrical checks. They're solid meters, and the basics apply to any brand. Here's the procedure I make every new inspector learn:

1. Inspect the leads first. Damaged leads cause more bad measurements than the meter ever will. If the insulation is cracked, replace them. Worn leads and a 230V circuit are a bad combination.
2. Plug into the correct jacks. Black goes to COM. Red goes to VΩ for voltage and resistance, or to the mA/A jack for current. The current inputs are fused separately, and the meter won't survive guessing wrong on a live circuit.
3. Set the dial before touching anything. Reading a voltage with the meter set to ohms is a classic error. On an Extech, the function switch is simple, but it has to be intentional.
4. Connect in parallel for voltage, in series for current. You touch the probes across two points for voltage. For current, you have to physically open the circuit and put the meter in line.
5. Read the display, then turn the dial off. The auto-off feature helps, but we still find meters with dead batteries because someone left the dial on. Now 'dial off' is part of the daily checklist.

One thing I've learned the hard way: check the meter's battery before trusting a reading. A dead battery gives a blank display, and we once had an electrician nearly work on a live circuit because his meter screen was dead—he read it as zero volts. It wasn't zero. The battery was flat.

What does 'within tolerance' actually mean? When should you reject a part?

This is the question that separates a quality department from a measuring station.

A part isn't automatically good just because the measured number sits inside the tolerance band. You also have to account for measurement uncertainty. If the tolerance is ±0.05 mm and your CMM has a calibrated uncertainty of 0.02 mm, a part measuring 0.04 mm off nominal is technically in-spec—but it's so close to the edge that the true value could be out of tolerance.

I've rejected parts that were 'within spec' because the data put them too close to the limit. The machining team was not happy. But a boundary case is a red flag, not a pass.

The guideline I use: if a measurement lands within 20% of the tolerance limit, follow it up with an uncertainty review before making a call. It's not about being difficult. It's about knowing for sure.

How often should I calibrate measurement instruments?

On a schedule. Not when something fails, and not when someone remembers.

We do annual calibration for most instruments and semi-annual for the tools that see daily abuse—including our Fluke 773 clamp meter and the CMM's probe system. The reason is unpleasant but simple: if you calibrate a drifting instrument at the end of the year, you find out that everything it measured during that year might have been wrong. The longer the gap, the bigger the risk.

When you send instruments out, use an ISO 17025-accredited lab. The accreditation matters more than the brand of the lab. It means the calibration is traceable and the uncertainty is documented. As a quality manager, I will always spend a little more for that paperwork.

The Hexagon Portal makes this manageable. We log every certificate there, and it sends service reminders. Five years of calibration history now lives in one searchable place, which makes external audits much smoother. I remember the paper era. I don't miss it.