If you're measuring parts for someone else, a Keyence encoder and a Keyence VHX digital microscope aren't "nice-to-have" tools—they're the difference between shipping good parts and paying for them twice. I learned that after I scrapped a 300-piece order in 2019 because I trusted a dial caliper over a Keyence encoder. The job cost me about $3,800. No—$4,200. I'm mixing it up with a different disaster. Either way, it hurt. Actually, it hurt twice: once in money, once in credibility.
I've been handling measurement and inspection for eight years. I've made enough mistakes to fill a checklist, and now I maintain that checklist so the other engineers in our shop don't repeat them. I'm not a metrologist. I'm just someone who buys machine time, measures things, and has paid the tuition for bad measurements.
The dial caliper vs. Keyence encoder lesson
In 2019, we were cutting small precision plates. The machine had a Keyence encoder on a linear axis. For reasons I still don't fully understand, I decided to verify the encoder position with a dial caliper. The caliper said 12.00 mm. The encoder said 12.11 mm. I called the encoder the problem because I wanted to trust something simple and mechanical.
The machine ran for a full shift. We made 250 plates, and every one had a feature shifted about 0.11 mm. The customer's inspection caught it. Our dial caliper didn't, because the caliper was mounted to a bracket that flexed under my grip. The encoder was right.
That's when I learned that resolution is not accuracy. A dial caliper looks precise, but its readout is only as good as the reference edge, the calibration, and the person holding it. A Keyence encoder gives you a digital signal tied to actual position. When they disagree, don't assume the cheap tool is the honest one.
Honestly, I'm not sure why the bracket flexed exactly that direction every time. My best guess is the clamp wasn't square, so the caliper body twisted slightly. It was a tiny error that became a huge problem because I ignored the better tool.
Adding to that, a caliper is a snapshot. You measure one spot, read one number, and move on. An encoder keeps generating data while the machine moves. For production, that continuous data is worth more than any number I can capture by hand.
Why I use a Keyence VHX digital microscope for surface checks
I used to inspect surface finish with a loupe and feel. That's fine for sorting gravel, not for finding cracks that only appear under the right light.
The Keyence VHX digital microscope is the tool I wish I'd bought three years earlier. It doesn't replace a CMM. It does something better for certain jobs: it lets you see the part in 3D, measure a crack, and capture the image so a customer can't argue with it. I caught a hairline crack once that a loupe would have never shown me. That one crack saved a $2,400 customer rejection.
Here's the part most buyers miss: it's not about the pixel count. It's about the lighting. The VHX's lighting modes and depth composition are what make a hidden crack obvious. If you're comparing digital microscopes, don't just look at the pretty sample photos. Ask to see it on your part, with your defect.
I didn't buy the VHX when I first started. I was a two-person shop, and my orders were small. The Keyence application engineer who showed up with a demo unit didn't make me feel small. He let me test it on my worst part—an ugly bracket with a machining mark—and said "that's exactly what this is for." That experience is why I still take small orders seriously. Today's $200 order can be next year's $20,000 order. Some suppliers never learn that.
Megger vs insulation tester—and the 83 multimeter I still trust
Measurement isn't just dimensions. Machine builders need electrical diagnostics too. My bench top still has a Fluke 83 multimeter I bought used years ago. It's a good meter for voltage, current, and continuity. But it is not an insulation tester.
I once used it to check a motor cable. The meter read OL—overload—so I thought the cable was fine. Actually, an insulation resistance test needs a high-voltage source to stress the insulation. A multimeter doesn't do that. When the customer asked for a megger test, the insulation tester showed 1.2 MΩ at 500 V, which is low enough to cause trouble. If I'd known then what I know now, I would have skipped the "good enough" check and used the right tool.
And yes, "megger" is a brand. It's so common that we say "megger the cable" even if the tester is a different brand. So, megger vs insulation tester: a Megger-brand insulation tester is one kind of insulation tester. Not every insulation tester is a Megger. If the drawing says "megger test," ask which standard and which voltage. Or use the exact tester the customer spec requires.
What I still use, and what I don't
So am I saying throw away your dial calipers and 83 multimeter? No. I still use them for quick checks, setup verification, and troubleshooting. A dial caliper is fine when the tolerance is 0.005" and you're measuring a clean surface. A multimeter is fine for finding out whether a wire is alive. But "fine" is not "good enough for machine control." When the data goes into an automated decision, use a tool with real resolution and repeatability, and check the spec sheet—not the price tag.
One more honest boundary: I don't know everything about these instruments. I've never fully understood why some insulation testers hold a reading and others don't. I also can't tell you if the VHX is the right choice for every surface inspection. What I can tell you is that my $4,000 mistake happened because I trusted a comfortable tool instead of a trustworthy one. That's the lesson I pass on.
Specifications mentioned in this article are from my own experience and from Keyence product information as of January 2025. Verify current specs at keyence.com before relying on them for production.
