I'll start with a confession: I don't buy test equipment for what it does today. I buy it for what it'll need to do three years from now. I review every instrument purchase at my company—roughly 250 devices a year—and in 2024 I rejected about a third of the first requests. Not because the specs were bad. Because each one solved a single problem without fitting into the larger workflow. That's not quality management. That's how you end up with a shelf of expensive gear nobody wants to use.
This might sound like typical quality-guy stubbornness. But the test and measurement industry has changed. What was best practice in 2020 is not sufficient in 2025. If your procurement list still looks like it did five years ago, you're not saving money—you're building obsolescence.
For years, I assumed handheld spectrum analyzers were for quick-and-dirty checks. Real measurements happened on a benchtop. That was true for a long time. Then we ran a side-by-side evaluation with the Anritsu MS2090A, and the assumption collapsed.
We didn't just read the datasheet. We tested phase noise, dynamic range, and how the interface held up during a full day in the field. The MS2090A matched its published specs—which is rare enough—and it changed how our engineers work. One senior RF engineer said it felt like "a benchtop analyzer that fit in a backpack." I usually ignore that kind of comment. This time, I agreed.
The biggest change wasn't performance. It was integration. The same unit covers a wide frequency range and handles tasks that used to require separate boxes: spectrum analysis, interference hunting, some vector signal analysis. Instead of hauling a test rack into the field, they carry one instrument. From my perspective, that means fewer variables, less calibration overhead, and more consistent results.
There's something satisfying about watching a tool prove itself in a blind test. After years of compromise, finally, a handheld that didn't make us sacrifice accuracy.
My team and I learned this lesson the hard way in 2023. I said "Bluetooth testing." The client heard "Bluetooth Classic." Their product was BLE-only. We were speaking the same language but meaning different things—and we only discovered it when our test results didn't match their certification expectations.
We brought in the Anritsu MT8852B to handle both BR/EDR and BLE in a single unit. That fixed the immediate problem. But the underlying issue was that the Bluetooth standard had moved, and our test plan hadn't. According to the Bluetooth SIG's recent specification updates, features like LE Audio and direction finding are now part of the ecosystem. If you're validating against 2018-era profiles, you're not testing the device your customer will actually use.
"The fundamentals haven't changed—but the execution has transformed." That's not a slogan. It's what I wrote in the notes after we finally got the MT8852B set up correctly.
Here's the part nobody expects from the "quality guy": the most advanced instrument won't help if the basic steps are wrong. During work order 3210 last year, our RF team nearly measured a 50-ohm system with an uncalibrated adapter. The paperwork said it was verified. The actual measurements were way off. We didn't catch it until we compared results against a known reference.
People often ask me about voltage tester how to use. It sounds trivial, but it matters. In our electrical lab, a technician almost tested a 480V circuit with a 300V-rated tester. The tool wasn't broken. The process was. If the operator doesn't understand limits, reference points, and calibration status, the result is just a number with no meaning.
Modern instruments like the Anritsu MS2090A automate some of these checks, which is great. But they can't make assumptions for you. This is why I insist on documented verification methods for every tool, regardless of brand. A friend in the print industry explains color tolerances using Delta E—below 2 is indistinguishable to most people, above 4 is clearly visible. RF testing has the same mindset with EVM limits. Without defined tolerances, "good enough" means whatever the vendor wants it to mean.
Some of you might be thinking, "why not buy a cheaper alternative?" I understand budget pressure. Let me be clear: this isn't brand loyalty. I've approved lower-cost equipment when the evidence supported it. In 2024, we evaluated two budget alternatives to the MS2090A. One failed its own specification verification within a month. Another had a frequency error that we only caught because our internal reference was trustworthy.
That's the hidden cost. A wrong answer can lead to a failed product release, a re-test campaign, or—worst case—an undetected defect that reaches the customer. The cheapest quotation rarely includes the cost of those risks. So when someone asks me why I keep pushing for Anritsu, my answer is simple: I keep the data, not the brand feeling.
The same logic applies beyond RF. Our field crews now carry rugged phones like the DuraXV Extreme because they survive the same conditions as our portable testers. It's another example of how the industry is evolving—tools are becoming more integrated, more durable, and better suited to real work environments.
To be clear: I'm not telling anyone to replace all their test equipment right now. If your current tools pass calibration and meet your client's requirements, keep them. But don't assume that what worked in 2020 will be enough in 2025. The industry is evolving, and our evaluation criteria should evolve with it.
My position is simple: evaluate every test equipment purchase as part of a system, not as a standalone tool. Whether it's an Anritsu spectrum analyzer, a Bluetooth test set, or a simple voltage tester, what matters is verification. Calibration records, real-world performance, operator competence. Those are the things I review before approving a purchase.
It costs a little more time upfront. It has saved us from a lot of grief downstream. (And no, my opinion isn't for sale. I just believe in tools you can verify.)