I coordinate emergency repairs for industrial automation customers—eight years, 400+ rush jobs, and a record number of 2 A.M. phone calls. When a filling line stops 36 hours before a scheduled restart, or a QC foreman needs 100% inspection before an overnight courier arrives, I'm the one who answers. In that world, measurement tools decide whether a 1 A.M. call ends with a fix or a four-hour argument.
The opinion I'll defend today: “you're a small shop, make do” is the most expensive advice in manufacturing. Small teams don't need more tools. They need the right tools, and the know-how to use them under pressure. I didn't always think this. Early in my career, I genuinely believed a $200 multimeter was enough for any plant with fewer than 50 employees. Three jobs rewired that belief—one about a pressure sensor, one about a digital microscope, and one about a megger insulation tester gathering dust under a workbench.
A Pressure Sensor That Wasn't Broken
In March 2024, a food-processing client called at 9:30 on a Wednesday night. A packaging-line glue station kept misfiring, and the overnight restart was scheduled for 6 A.M. The maintenance lead had already replaced three components, including a brand-new pressure sensor on the pneumatic feed line. The HMI displayed a perfectly healthy reading. The machine still misfired.
Here's something people don't realize until they've watched a line go down: a pressure sensor is four specifications, not one—range, accuracy, response time, and thermal drift. The replacement sensor had the right range and excellent accuracy. Its response time, though, was roughly 20 milliseconds slower than the original. Against a rapidly cycling air valve, 20 milliseconds shifted the glue timing just outside the tolerance window. The part was new. It was also the wrong part for that dynamic signal.
We logged the signal at the sensor's output, caught the lag, and called the sensor vendor's support line at 2 A.M. (they answered—which tells you a lot about a supplier). The fix took 11 minutes. If that plant had owned one instrument capable of capturing a 20-millisecond transient, they'd have found the problem in an hour instead of losing a shift. The “small budget” approach—trust the HMI, swap parts until something works—cost them more than any measurement tool on the market.
The Keyence Microscope Manual, and the 11 P.M. Judgment Call
The second lesson came from a client that machines precision parts for medical devices. It's a small operation—about 15 people—with quality requirements that would make a Fortune 500 lab nervous. One Tuesday in May, a rush order required 100% visual inspection of 800 components before an overnight courier pickup. The QC tech had the right machine for the job: a Keyence VHX-series digital microscope, capable of resolving features invisible to the naked eye.
One problem: nobody on that shift had used the microscope in six months. The tech stared at the instrument, then at the stack of parts, then opened a binder searching for the Keyence microscope manual to relearn the setup routine. Not because the manual is bad—it's thorough to a fault. But no one had documented the procedure for this part, on this machine, with these lighting conditions. At 11 P.M., with a courier deadline, “thorough” is not what you need.
What most people don't realize is that a digital microscope's workflow is easy to forget when you don't use it weekly—illumination, calibration lock, focus stacking, measurement toolbar. Skip one step and you can get a beautiful image that measures wrong. (Which, honestly, is worse than no image at all, because it looks trustworthy.) The missing piece wasn't the instrument; it was the access plan for its knowledge.
We recovered the job by downloading a current copy of the Keyence manual from their website (a genuinely pleasant surprise—as of 2024, Keyence makes manuals available for direct download, no support ticket maze), and a local rep walked the tech through the setup in a 20-minute call. The parts shipped on time.
But that near-miss changed how I advise small clients: a 15-person shop can absolutely justify serious inspection equipment—if the training and documented procedure come with it. The tool wasn't the failure. The missing quick-reference guide was.
How to Use a Megger Insulation Tester: The 45-Minute Fix
The third case is my favorite because it ended so fast. An automotive supplier's plant had been losing production to random ground-fault trips on one packaging line for three days. Three full days. The plant electrician was convinced a variable-frequency drive was at fault, and a replacement drive had already been ordered at four-figure cost.
I asked one question: “Has anyone done an insulation resistance test on the motor cable?” The silence told me everything.
Here's what I've learned about how to use a megger insulation tester in practice:
- Disconnect the motor leads from the drive.
- Select a test voltage based on the equipment rating—per IEEE 43, that's generally 500 V DC for motors under 1 kV.
- Apply the test for 60 seconds and record the reading.
- Compare each phase to the recommended minimum (roughly 1.5 MΩ for a 480 V-class motor at 40 °C) and to each other, because the pattern between phases tells the real story.
(Check the current revision of IEEE 43 and your motor's nameplate before testing—this is a field simplification, not the full standard.)
We grabbed the plant's megger—it had been sitting under a workbench for six months because the crew wasn't confident in it—and ran the test on each phase lead. Phase A read 40 MΩ. Phase B read 1.1 MΩ, well under the 1.5 MΩ floor. Phase C sat at 6.8 MΩ. The pattern pointed directly at moisture in one feeder cable section, not the VFD. We isolated the wet connection, dried and re-terminated it, re-tested, and had the line running before lunch. The replacement drive was cancelled.
We had about 30 minutes to make the call before the plant manager needed a go/no-go answer. Normally I'd repeat the test after a longer de-energized rest, but there was no time—we went with the three-phase comparison and a margin check. The surprise wasn't the wet cable. It was that the tool that caught it had been bought years earlier and never properly learned. That instrument was not exotic. It was just unknown.
But the Good Tools Cost Money
That's the honest objection, and I'm not going to wave it away. I've been the person arguing with a purchasing manager about a line item. Budgets are real. My old Fluke 73 Series II multimeter is still my daily carry—for continuity, basic voltage, and a quick sanity check, it's genuinely all I need.
At least, that's been my experience for 80% of routine checks. (Should mention: the damage doesn't happen in that 80%. It happens when a basic tool gets used for a job that needs something faster or more precise, because the right tool “wasn't in the budget.”)
The framing is the problem. We compare “good tool” against “no tool,” so the good tool looks expensive. In practice, the real choice is between a properly spec'd instrument and a chain of failed part swaps, last-minute courier surcharges, rental equipment, and lost production. In the pressure-sensor case above, the customer spent more on three replacement components plus a rush-service visit than a correct sensor would have cost. And the megger case: the plant had burned an estimated $18,000 in downtime and was about to spend four figures on a drive it didn't need. The tester that caught the real fault had been under the workbench the entire time.
This is exactly where the small-shop angle matters. Large plants have metrology departments that catch procedure gaps before they become emergencies. Small teams don't. When a 15-person shop buys a capable microscope or a real insulation tester, that isn't luxury spending—it's their entire quality and reliability department in one case. When the QC manager at that medical-parts shop signed off on the microscope, the calculation was simple in hindsight but agonizing at the time: several thousand dollars for a tool that might sit idle, versus the next overnight inspection with a courier waiting and no backup plan. It did sit idle for six months. It still paid for itself the first night it counted.
To the vendors reading this: treat the small inquiry seriously. I've watched the intern who placed a small order at a 12-person shop become the maintenance manager at a 200-person plant eight years later (happened twice). The suppliers that helped early—with manuals, with a straight answer, with a 20-minute refresher call—are the ones that got the call first. Today's small shop is tomorrow's spec sheet.
So here's my line, and I won't soften it: if your team is small, you can't afford to guess. Buy instruments that give you the truth, learn them from the manual, write down the 3 A.M. version of the procedure, and make sure more than one person knows it. “Make do” should describe lunch, not measurement.
The Bottom Line
Eight years of rush jobs taught me one thing about emergencies: they don't discriminate by headcount. A 15-person fab shop gets the same 2 A.M. failures as a Fortune 500 plant—it just has fewer people standing around the machine looking confused.
I'll leave you with this. In eight years, I have never been called because a customer's measurement equipment was too accurate for the application. I've been called dozens of times because it wasn't accurate enough—or because it was sitting under a workbench, waiting for someone brave enough to open the manual.
Stop making do. Start measuring.
