You’re holding a failure report and an instrument that might be off. A Testo thermometer that’s two degrees out in an ice bath. A level transmitter that drifts every weekend. A Rice Lake load cell giving an unstable zero. Or maybe you dropped your multimeter and the case latches are still intact—but you’re not sure if a new multimeter case is a real fix or just decoration.
The first question most people ask is, “Can this be fixed?” The better question is, “Should this instrument stay in service, and under what conditions?” That answer depends on which scenario you’re in.
I’ve spent the last four years reviewing measurement devices before they reach our production floor. Roughly 200 units per year, across temperature, pressure, level, and electrical testing equipment. I’ve rejected a fair share—and I’ve also signed off on instruments my team wanted to scrap because it was cheaper to troubleshoot instead of replace.
Let me be upfront: there is no one-size-fits-all answer here. Best practice in 2020 was “just recalibrate and see.” By 2025, many instruments have enough onboard diagnostics and replaceable components that replacement is often the smarter call—especially for handheld devices. But some fundamentals haven’t changed. You still need traceable accuracy, documented history, and a system that gives you confidence in every reading.
Classify Your Failure First
I recommend sorting every questionable instrument into one of four buckets:
1. Drift without visible damage.
2. Physical or handling damage.
3. Installed process sensor failure (level, pressure, load cell).
4. End-of-life or no longer fit for purpose.
Each bucket leads to a different answer.
Scenario A: Drift Without Visible Damage
A Testo thermometer reads 1.2°C high at a single reference temperature. The housing is clean, the probe looks fine, the battery is fresh. The drift is repeatable. This is the classic calibration scenario.
My advice: send it in for recalibration—or, if the probe is removable, replace the probe first. A probe-only failure is far more common than most people expect. We once spent $80 to calibrate a Testo temp instrument only to find the same probe still drifted because the thermocouple tip was contaminated. Swapping the probe solved it for less than the calibration cost.
What about a multimeter that’s slightly out of tolerance but works fine otherwise? Same logic. If you have calibration records showing it was stable for years, a single out-of-tolerance result may be an isolated event. Recalibrate, document the result, and adjust your calibration interval. If it fails again, see Scenario B.
One rule I’ve learned after hundreds of acceptance checks: if the instrument drifted but there’s no damage, recalibrate once. If it drifts again within six months, replace it. There’s no point paying for serial recalibrations of an unstable device.
Scenario B: Physical Damage or A Compromised Case
Here’s where I sound less like a welder and more like a safety auditor. Physical damage is not about aesthetics. The case is part of the instrument’s safety system, especially on a multimeter.
If your multimeter fell from a ladder, check the CAT rating markings and the input jacks before you even think about functionality. A cracked case or worn input shroud can expose you to live circuits. If the manufacturer’s repair cost is less than roughly 60–70% of replacement, and the repair includes a full test, go ahead with the repair. Otherwise, replace it. I’ll admit this once: a few years ago, I approved a “repair” of a damaged digital multimeter case because the meter worked perfectly. The electrician who used it later told me the case didn’t feel right. We replaced it that week, but we should have replaced it immediately.
For a Testo thermometer that has been dropped—same rule. If the probe is bent or the display glass is cracked, a factory repair might restore the IP rating. But if the body itself is cracked, moisture will get in the next time you use it in a brewery or cold warehouse. Replace it. This advice often surprises people because “it still reads fine.” It reads fine until it doesn’t.
What if you only need a replacement multimeter case because the original belt clip broke or the rubber holster wore out? That’s a minor accessory replacement, not a repair—as long as the meter itself hasn’t suffered an impact. Just remember that an oven mitt doesn’t fix a burned hand, and a new case doesn’t fix an overloaded meter.
Scenario C: Level Transmitters and Load Cells in the Field
Installed process instruments are a different animal. If your level transmitter is drifting, the problem is often not the transmitter. It could be a plugged impulse line, a bad cable, an erratic power supply, or condensation inside the enclosure. I’ve seen technicians replace a $900 level transmitter and then discover the sensor port was filled with dried slurry. The manufacturer’s troubleshooting guide—if anyone reads it—will tell you to check these things first.
The same logic applies to a Rice Lake load cell. If you search for “how to troubleshoot a rice lake load cell, you’ll find the same starting point: measure the resistance across the Wheatstone bridge and the excitation voltage before condemning the cell. Erratic weight readings are almost always a connection or junction box issue. When you do have a faulty cell, don’t mix old and new cells on the same vessel unless you match capacity and output. Rice Lake’s manuals are explicit about that. Mixing cells is one of the costliest mistakes I’ve seen in a process plant: eight new load cells, one failed old cell, and a month of troubleshooting because someone let a single replacement drag on.
The rule I use after Q1 2024, when we received a batch of twelve level transmitters with two that didn’t meet specifications: troubleshoot first, replace second, and always document the environmental conditions. Level transmitters and load cells are often exposed to condensation, lightning surges, and vibration. That changes the failure pattern. If you replace an installed sensor, also inspect the cable gland and the junction box. A “defective” transmitter is often actually a cable problem wearing a transmitter jacket.
Scenario D: End-of-Life, Obsolete, or No Longer Fit for Purpose
Instrument technology has changed faster than most managers care to admit. What passed as a great level transmitter five years ago may not support the diagnostics your plant’s control system now needs. A multimeter that’s mechanically fine but doesn’t have sufficient CAT rating or measurement capabilities for your current electrical tasks is not worth keeping.
I’m not saying replace everything new. The fundamentals of measurement—stability, repeatability, traceability—haven’t changed. But execution has. Newer Testo probes often have longer calibration intervals and tighter drift specs. Modern level transmitters include self-diagnostics that pinpoint a plugged line before it causes a spill. If your existing instrument cannot meet the requirements of your current process, no amount of calibration will make it fit.
Consider replacement when:
— The manufacturer no longer offers factory repair.
— Replacement probes or sensors cost more than half the price of a new unit.
— Your quality standard changed (e.g., new SIL rating, higher accuracy, digital communication).
— The instrument has failed calibration twice within one maintenance interval.
How to Decide Which Scenario You’re In
If you’re not sure, answer these five questions in order:
- Did the instrument receive a physical shock or liquid exposure? If yes, it’s Scenario B—treat it with suspicion and inspect its safety enclosure before trusting it.
- Is it an installed transmitter or load cell? Then spend 30 minutes troubleshooting the wiring, the junction box, and the mount. That narrows Scenario C dramatically.
- Is there a removable probe or sensor? If yes, swap that first. That often solves drift without paying for a full recalibration.
- Has it failed calibration in the same way before? If it has, move past Scenario A and consider replacement rather than throwing more labor at an unstable unit.
- Does the manufacturer provide documented repair and calibration services? If yes, compare that quote with replacement. If no, you have your answer: replace it.
Here’s another way to think about it: when you’re standing at the instrument bench, don’t ask “can I fix this?” Ask “would I accept this device back into my own measurement system, with a clear audit trail?”
Full disclosure: I have rejected brand-new units from major manufacturers—including Testo and Rice Lake—when they didn’t meet specified tolerances. The intent isn’t to punish the brand. It’s to protect the process. A “quality” instrument is only quality when it provably does the job.
So, to sum up: if your Testo thermometer drifts and looks clean, calibrate it once. If it drifts again, replace it. If it suffered trauma, repair only through the manufacturer or replace it. If a level transmitter or load cell starts acting up, troubleshoot the whole circuit before blaming the sensor. And if the instrument is old, unsupported, or no longer meets your process demands, don’t let sunk cost keep an unreliable reading in your loop.
There’s no universal best answer. That’s fine—your specific scenario gives you a better answer than any blanket rule could.
