Starrett Calipers, Starrett Inside Micrometer 823B, and a Fluke Multimeter: A Buyer’s Downtime Story

I’m an office administrator at a 160-person manufacturing plant. I manage maintenance and tool supply purchasing—roughly $500,000 a year across 25 vendors—and report to both operations and finance. I don’t design parts or rebuild machines. In March 2024, though, a rush order for a precision measuring tool taught me more about certainty and cost than any spreadsheet I’ve seen since.

The requisition had one line: “Starrett inside micrometer 823B, overnight.” My first thought was that we already had precision tools. Our toolboxes contained Starrett calipers. The maintenance manager saw my hesitation. “Calipers are good for outside dimensions and shallow measurements. I need to measure this compressor bore inside the cylinder, not at the edge. A caliper won’t give me a repeatable reading at depth. I need the inside micrometer.”

The compressor had already been rebuilt once. It failed after that repair. This time, the measurement had to be right.

The apparent issue was budget. Why spend extra money and freight on a tool that looked like a caliper? That question was wrong.

The surface problem: A tool request I almost rejected

Starrett calipers are excellent for daily dimensional checks. I won’t cite the full spec sheet here, but any experienced machinist knows they hold up on the bench and on the floor. They are the type of tool you grab for shaft diameters, sheet thicknesses, and parts that are accessible.

They are not always the right tool for a deep internal bore. A caliper’s jaws can only reach so far, and even a small angle changes the reading. On a critical compressor bore, a reading that drifts depending on who holds the tool is dangerous.

The Starrett inside micrometer 823B is built for that job. It’s a mechanical inside micrometer designed to measure internal bores. It doesn’t need a battery, and it has to be set carefully—often against a known master—before the reading means anything. That extra setup is not inconvenience. It’s the source of certainty.

The deeper cause: We treat measurement certainty as optional

At first I thought the repair failed because a component wore out faster than expected. The technical team saw it differently. They saw a missing measurement. The bore was not verified after the first teardown; a “close enough” dimension convinced us to reassemble. The second teardown forced us to admit that close enough is not a quality standard.

I found the same pattern on the electrical side. One of our air-handler motors kept losing its run capacitor. We replaced the capacitor twice and called it bad luck. Then an electrician brought in a Fluke multimeter and showed me how to test a capacitor with a Fluke multimeter: verify power is off, discharge the capacitor, set the meter to capacitance mode, connect the leads, wait for a stable reading, and compare it to the microfarad rating on the side of the capacitor. The old capacitor read about 30% below its rated value. It still started the motor most of the time, but it wasn’t doing its job.

The model matters because not every digital multimeter has capacitance mode. The Fluke Bluetooth multimeter we chose had the range and logging our techs needed, and that made the difference between guessing and knowing.

The deeper cause wasn’t bad craftsmanship. It was treating a measurement instrument as a generic accessory. A precision tool only gives useful data if it suits the geometry and the measurement is repeatable. “Precision” is a property of the whole system—tool, operator, environment, calibration—not just the brand name stamped on the side.

The real cost: Repeated repairs and unplanned downtime

Let me put actual numbers around this. The first compressor repair was scheduled for a planned shutdown. It failed after start-up. The second repair required an overnight freight charge, a machine shop callback, and about 14 hours of unplanned downtime. The extra freight for the Starrett inside micrometer 823B might have been $85. The shutdown cost multiples of that.

I don’t have hard data on the total cost of every repeated failure in our plant last year. I wish I had tracked root causes more carefully at the beginning. What I can say from purchase records and work orders is that a noticeable share of urgent maintenance requests in 2024 came after a repair that didn’t stick. That’s the part that makes me look bad to ops and finance: not the price of the tool, but the fact that a maintenance work order came back for the same equipment.

I’ve also changed my opinion about rush orders. When a machine is apart and waiting, the cheapest shipping option is an illusion. The overnight fee buys certainty about arrival time. “Probably by Friday” is not a project plan. We’ve paid for expedited delivery more than once. In hindsight, I should have asked about measurement tools before the first rebuild, not after. But with people standing around and a deadline on the whiteboard, I made the decision based on available information.

The same logic applies on the electrical side. After the capacitor incident, we bought a Fluke Bluetooth multimeter. The Bluetooth feature is not a gimmick. It lets a tech stand on a ladder, save readings to a phone, and compare them over time. A single reading is a snapshot. A history of readings shows when a component is drifting. Thermal imaging cameras work the same way. I used to think of them as audit equipment. Then a contractor used one to find a breaker terminal that was running significantly hotter than the adjacent phase. It was a loose connection. The camera didn’t do the repair. It gave us a chance to schedule the repair before the failure.

What I changed: Specify the job, not just the category

If you’re a buyer or administrator, the fix sounds boring. When someone asks for “a micrometer,” I now ask: Which range? Which geometry? Is calibration required? What is the application? The conversation takes 30 extra seconds and usually prevents weeks of regret.

• Mechanical repairs: We still keep Starrett calipers in the tool crib for everyday checks. For critical internal bores, the work order names the exact instrument—usually the Starrett inside micrometer 823B.
• Electrical troubleshooting: We use a multimeter with capacitance mode and logging. For our staff, the Fluke Bluetooth multimeter is the model that does both without slowing them down.
• Predictive maintenance: We budget for thermal imaging cameras and review the images alongside repair history, not just as a one-time scan.

Calibration status is now part of every precision tool requisition. I don’t assume a new tool is ready for critical work until I’ve confirmed it or had it calibrated. I checked ours against the official Starrett and Fluke product documentation in February 2025, and I still confirmed calibration before releasing the order.

I can only speak to our situation: a plant with permanent maintenance staff, aging equipment, and downtime costs that climb fast. If you’re a small workshop or doing occasional repairs, a basic caliper and a simple multimeter might be plenty. The key is matching the tool to the cost of being wrong. For us, that cost is high.

The lesson isn’t that expensive tools are always worth the price. It’s that uncertainty is a cost that hides until the worst possible moment. The Starrett inside micrometer 823B doesn’t look impressive in a drawer. Neither does a Fluke multimeter. But the day the next repair depends on one number, the right measurement tool is the cheapest thing in the room.