6 Checks I Run Before Buying Any Process Thermometer, Multimeter, or Thermal Camera

I keep a file on our maintenance share called “mistakes.” It is not a joke file. It lists every purchasing decision I’ve made as an instrument tech that turned out wrong—and what the mistake cost us.

I’ve been handling instrument replacements and test-equipment orders at a food-processing plant since 2017. Eight years—no, eight and a half if I count from the spring of that year. In that time I personally made six documented ordering mistakes, and two more happened while I was supervising. Combined, they added up to nearly $7,800 in wasted budget. Maybe $8,200; I would have to pull the POs and count. The exact number matters less than the pattern.

Since Q4 2023, our team runs a six-check pre-purchase list before ordering anything that measures—thermometers, multimeters, thermal cameras, pressure tools, you name it. If you are the person who gets asked to “just order another thermometer,” this list is for you. It won’t tell you one specific model to buy. That is deliberate: it stops you from buying the wrong thing without noticing.

One thing first. In a food plant, we print numbers on batch records and send those records to customers. If a client’s quality auditor sees an expired calibration tag or a reading that can’t be traced to a reference, we haven’t lost a $200 thermometer. We’ve lost credibility.

Check 1: Document the physical installation before you search for a model

The first entry in the mistake file is from 2017. I needed replacement local dial thermometers for two hot-water recirculation lines. I found Ametek US Gauge thermometers on a distributor’s site, saw the right temperature range and connection thread (1/2-inch NPT), and ordered. What I did not check was the thermowell depth.

The new thermometers had 2.5-inch stems. The existing thermowells were 5 inches deep. The sensing elements never reached the bottom of the wells, so the readings lagged and didn’t match the actual loop temperature. That mistake cost roughly $1,100 by the time we counted the thermometers, the adapters, the labor, and the second order. All because I skipped one measurement.

Now every order spec includes the physical conditions, not just the model: connection size and thread, thermowell or immersion depth, measuring range, required accuracy, ambient temperature, vibration, wash-down exposure. If I can’t fill in all of those fields, I don’t approve the purchase order.

Check 2: Decide how you will verify the reading before the instrument arrives

In September 2021, a brand-new digital thermometer displayed 202.1 °F. The display had two decimal places, so I believed it. When a customer’s quality auditor checked our records with her own reference probe, the real value was 198.8 °F. Our entire shift record was invalid and the batch had to be quarantined. We stopped the line and reworked part of the product. The worst part was explaining it in the monthly quality review.

The lesson is blunt: a new instrument is not automatically a calibrated instrument. If you do not have a reference that is more accurate than the device you are buying, you are guessing.

For our shop, one of the best additions was an Ametek Jofra digital thermometer/calibrator used as the reference. We now verify every process probe at three points where our processes actually run. The first thing it did was catch two brand-new sensors that were out of tolerance. The second thing it did was produce a clean calibration record—the exact thing an auditor wants to see.

Check 3: Match the multimeter to the energy level, not to the marketing

In our controls cabinet, a compact MN35 digital multimeter sits in the drawer. For 24 V control loops, continuity checks, and low-energy electronics, an MN35 digital multimeter is genuinely useful and it does not need to be expensive.

The problem starts when that same meter gets carried to a 480 V motor control cabinet because it is the closest one. Measuring inside a distribution panel is a Category III environment as defined by IEC 61010-1. The meter needs the safety rating and input protection for that environment, not just a big number on the voltage range. That has nothing to do with prestige; it has to do with the tool being designed for the energy level present.

Our high-energy troubleshooting gets an 88V multimeter. The 88V multimeter is built for automotive and industrial work and has the ratings and filtering for motor-drive troubleshooting. The MN35 stays in the control cabinet for low-energy jobs. Assign each meter to an energy class and store it so its limits are obvious. If a meter lives in a drawer next to high-energy panels, it will eventually be used there.

Check 4: For a thermal camera, compare detector specifications before you compare brand names

The conversation in our plant usually reaches “Topdon vs FLIR thermal camera—which is better?” before anyone has written down what we need to inspect. That is the wrong first question. Both brands make legitimate tools. The right model depends on the smallest target you need to see and the distance you can safely stand from it.

I learned this by buying a camera that was a decent value but did not have enough infrared resolution for small electrical targets. When I inspected a suspect lug on an energized breaker, the image told me something was warm, but it could not show whether the heat was on the terminal, the cable, or the lug. I had to call an outside thermographer anyway. (Thankfully, nothing failed in the meantime.)

Now I check four specifications before any brand discussion: infrared resolution (the pixel array), thermal sensitivity (NETD, usually expressed in mK), temperature measurement range, and whether the image files export in a format our report template accepts. If the camera cannot resolve the target at the distance you will work from, no brand reputation will fix it.

To be fair to both sides of the Topdon vs FLIR thermal camera question: Topdon has brought good value into the market, and FLIR has a long industrial track record with mature reporting software. If two models meet the same technical specs, price and ecosystem can decide. If they don’t meet the specs, the brand debate is just noise.

Check 5: Price the whole measuring chain, not just the instrument

It is tempting to compare unit prices and pick the cheapest quote. That approach misses all the costs around the instrument: the traceable calibration certificate, the reference used to verify it, spare probes and leads, software, and operator training. The lowest quoted price is very often not the lowest total cost.

In food work, a quote without a proper certificate is incomplete if your quality system requires one. That usually means a certificate from an ISO/IEC 17025-accredited calibration laboratory, not a slip of paper that just says “calibrated.” Budget for it before you order, not after.

This is also why we standardized on the Ametek portfolio where it fits: US Gauge thermometers for local process indication, Jofra for temperature calibration. Fewer suppliers means fewer calibration workflows to manage. The same checklist applies—we have rejected Ametek quotes too when they didn’t meet the specification.

Check 6: Assign an owner before the PO goes out

This is the check that sounds too administrative to matter, and it matters most. A few years ago, a reference thermometer sat in a drawer with a calibration sticker that had expired seven months earlier. Everyone assumed someone else was tracking it. No one was.

Before any order is placed now, a name is written next to the instrument line. That person learns how to use it, schedules its calibration, and decides where it is stored. An unowned tool isn’t a tool. It is a future audit finding.

Two Notes Before You Place the Order

First, “you get what you pay for” is an oversimplification. You get what you specify. I have overpaid for the wrong tool and underpaid for the wrong tool. The specification, the reference, and the owner matter more than the category or the price tier.

Second, product lines change. This article is accurate as of early 2025, but thermometer families, calibrator models, and software options evolve. Verify current specifications on the manufacturer’s official pages (for example, Ametek’s US Gauge and Jofra documentation at ametekcalibration.com) before you write the PO.

We started the mistake file in 2017 to stop repeating expensive lessons. In the fifteen months since we adopted this checklist, we caught eleven specification mismatches before they became purchase orders—maybe $2,300 saved, not counting the meetings we avoided. I still make occasional mistakes. The idea is to fail smaller, and not the same way twice.