I Thought I Was Saving Money
In early 2019, I was managing the measurement lab for a mid-size manufacturing company. We needed a Hexagon CMM for quality control, an analytical balance for our R&D team, and a FLIR thermal camera to check for hot spots in electrical panels. I figured—smartly, I thought—I’d buy used equipment to save budget. After all, a used 5810R centrifuge from Eppendorf was also on my list, and I’d heard they last forever.
My reasoning: “The specs are the same, the hardware is solid, and we can just calibrate it ourselves.” That was my first mistake. Actually, my first expensive mistake.
What Actually Happened: Three Failures, One Root Cause
The FLIR Thermal Camera Disaster
I bought a used FLIR T640 (circa 2015) from a reseller for $2,800 (versus $9,000 new). It looked pristine. The first job was a routine switchgear inspection. I took thermal images, saw what I thought was a normal 45°C busbar, and signed off. Three weeks later, the same panel failed—arcing, shutdown, $12,000 in lost production. The post-mortem showed the busbar was actually at 85°C. My camera was reading 40°C low because the internal calibration had drifted. (I only found out when I sent it to a certified FLIR service center—which charged $680 for recalibration, plus shipping.)
At that point, I realized: I had no idea how a FLIR thermal camera works under the hood. I assumed once it was calibrated at the factory, it stayed accurate. That’s not true—uncooled microbolometers drift over time, especially if the camera has been through temperature cycles or physical shocks.
The Hexagon CMM That Couldn’t Keep Up
Around the same time, I sourced a used Hexagon CMM (a Global Classic 05.07.05) for $18,000—half the list price. The seller claimed it had been “recently calibrated.” I believed him (mistake #2). We installed it, ran a few test parts, and everything looked fine. But when we certified our first production batch for an aerospace customer, the Hexagon app (PC-DMIS) flagged measurement deviations that didn’t match the CMM’s own repeatability report. Turned out the seller’s “calibration” was a self-performed check with a gauge block, not a full ISO 17025 traceable calibration. The machine’s scales had thermal drift, and the probe was out of spec by 12 microns. That error cost us a re-inspection of 47 parts—$3,200 in labor, plus a 1-week delivery delay.
After that, I started using the Hexagon app to log all calibration events (a feature I’d ignored before). But the damage was done.
The Analytical Balance & Centrifuge Combo
Our lab had a 5810R centrifuge (used, of course) and two analytical balances (a Mettler Toledo XS205 and a Sartorius CPA225D). Both looked accurate—the balances showed stable readings, and the centrifuge spun at the set RPM. But after a series of protein precipitation experiments gave inconsistent results, we finally checked calibration. The analytical balances were off by 0.8 mg at 200 g—well within the tolerance for most work, but not for our critical assay. The 5810R centrifuge had a speed error of 150 RPM at 5,000 RPM because the tachometer sensor wasn’t adjusted. The cumulative effect: three months of research data that had to be discarded. (Roughly $15,000 in wasted reagents and technician time.)
I had assumed that because these instruments “looked fine” and were from reputable brands (Eppendorf, Mettler), they’d hold their calibration. That assumption was my initial misjudgment—and I only believed the truth after ignoring it.
The Deep Cause: We Treat Instruments Like Appliances
The surface problem seemed obvious: I bought cheap used equipment. But the real issue was deeper—our mental model of “measuring instruments” was wrong. We treated them like kitchen appliances: buy once, use forever, only fix when broken. But precision instruments are more like athletes: they need constant conditioning (calibration), monitoring (verification), and sometimes retirement after a certain number of cycles or temperature exposures.
Specifically:
- Thermal cameras (FLIR) require annual calibration with a blackbody source. Without it, drift is inevitable.
- CMMs (Hexagon, Zeiss, etc.) need environmental compensation—temperature, humidity, vibration—and regular probe qualification using certified artifacts. The Hexagon app can automate this, but you have to use it.
- Centrifuges (including the 5810R) need speed calibration and rotor balance checks. Even a slight imbalance can skew results in sensitive assays.
- Analytical balances need daily internal calibration and periodic external weights verification. A “stable” reading doesn’t mean accurate.
And the second layer: I assumed that any vendor could calibrate any instrument. That’s where the “professional boundary” concept comes in. A general calibration lab might have a go at a FLIR camera, but they don’t have the proprietary firmware access that FLIR-certified centers do. Similarly, a non-Hexagon service shop can adjust scales, but they can’t update the CMM’s error compensation map the way a Hexagon-certified technician can. I learned the hard way: the vendor who said “this isn’t our specialty—here’s who does it better” earned my trust for everything else.
The Price of Ignorance: A Full-Year Impact
Let me put numbers on it. Over 12 months (mid-2019 to mid-2020), these three mistakes—the FLIR camera, the Hexagon CMM, and the lab instruments—cost us approximately:
- $12,000 in production downtime from the switchgear failure
- $3,200 in re-inspection labor for the aerospace parts
- $15,000 in discarded research materials and extra technician hours
- $2,100 in rushed recalibration and repairs for all three devices
Total: $32,300—all because I thought I could skip the calibration step. And that doesn’t count the loss of credibility with customers and internal teams.
After the third rejection in Q1 2020, I created a pre-purchase checklist for any precision instrument. The first question: “Who will calibrate this, and what does the calibration include?”
What I’d Do Differently (Short Version)
Since you’ve read this far, you probably want actionable advice—but I’ll keep it short, because the problem was already clear:
- Never trust a seller’s “recent calibration” unless you see a certificate with an ISO 17025 logo and a lab that’s accredited for the exact instrument type. For a FLIR thermal camera, that means a specific accreditation for thermal radiometry.
- Use the manufacturer’s app—the Hexagon app (PC-DMIS, ROMER, etc.) has built-in calibration logs. Use them. For analytical balances, schedule internal daily tests with a certified weight.
- Budget for annual calibration upfront. For a used 5810R centrifuge, expect $300–500 per year for speed and temperature verification. For a Hexagon CMM, budget $1,000–2,500 per year for full calibration (based on quotes from Hexagon Metrology Services in January 2025; verify current pricing). For a FLIR camera, about $500–800 per year.
- Know your limit. If a lab says “we calibrate anything,” run. The best vendors tell you when to go elsewhere. For instance, I now send all my FLIR cameras to a FLIR-authorized center, and my Hexagon CMM to a Hexagon-certified technician—even if it costs more. That boundary is what makes their recommendation trustworthy.
In the end, the lesson wasn’t about buying new vs. used. It was about understanding how precision instruments work at a deeper level—and respecting the expertise required to keep them accurate. A $4,500 calibration mistake changed my entire approach. I hope you can learn from it without paying the tuition.
