-
What does SICK actually make?
-
Why choose a SICK IO-Link encoder instead of a standard one?
-
How do I choose a moisture meter that won't create rework?
-
Is the Fluke 189 true RMS multimeter still worth buying?
-
Which Fluke multimeter do I actually need without overspending?
-
Where do hidden costs hide in this kind of procurement?
-
What's the most common mistake that causes rework?
I'm the procurement manager at a 340-person automation integrator. For seven years I've managed a roughly $1.2M annual MRO budget, signed off on 1,200+ POs, and compared quotes from more than 40 vendors. This is not a vendor brochure. It's the questions my own engineers, our maintenance crew, and purchasing colleagues ask most often — with the answers I give when we're trying to keep total cost down and avoid rework.
Here's what we'll cover:
- What SICK actually makes (and doesn't make)
- When a SICK IO-Link encoder is worth the premium
- How to choose an industrial moisture meter
- Is the Fluke 189 true RMS multimeter still a good buy?
- Which Fluke multimeter you actually need
- Where hidden costs hide in sensor and test equipment orders
- The one verification step that prevents most order mistakes
What does SICK actually make?
Short answer: SICK is a German industrial sensor company. It makes photoelectric sensors, safety encoders, proximity sensors, flow sensors, IO-Link masters, and a lot more for factory automation. That's what my purchase orders say, anyway. It is not a medical device brand. If you searched for "SICK real thermometer high fever" expecting a body thermometer, you're looking in the wrong aisle. For temperature-related products, SICK does carry industrial pyrometers and infrared sensors for processes — think furnace monitoring, molten metal, or kiln surfaces. Those measure high heat, not fevers. Keep the brand name separate from the medical meaning in your purchasing system, because the search engine will not always do it for you.
Why choose a SICK IO-Link encoder instead of a standard one?
A plain incremental encoder sends pulses; the controller counts them. A SICK IO-Link encoder sends absolute position, diagnostic information, and configuration over a single cable. That may seem like a small change, but it changes how you troubleshoot. Last year, one of our robotic cells started throwing random position errors. The standard encoder passed the usual checks — the pulses were still there. The IO-Link version we replaced it with gave a diagnostic warning about shaft bearing temperature before complete failure. No temperature sensor attached to it; the encoder's internal model caught the trend. That one warning saved us roughly six hours of production downtime. I don't have hard data on industry-wide failure rates, but based on our experience, the diagnostic layer is worth the premium if your PLC supports IO-Link. Verify the connection type and protocol version before ordering. That's the step that matters.
How do I choose a moisture meter that won't create rework?
Define the material and the measurement point before you look at any catalog. I didn't do that once. I said "we need a moisture meter for incoming QC." The vendor heard "portable handheld for spot checks." We ordered a pin-type wood moisture meter. Our actual need was an inline sensor for measuring moisture in plastic pellets during production. Totally different devices. Now I start with four questions: what material, what moisture range, what process temperature, and does the sensor need to be in contact or non-contact? For production lines, I prefer sensors with temperature compensation, because uncompensated readings drift with ambient conditions. And speaking of drift, we had no formal calibration schedule at first. The third time a batch was rejected for moisture, I created a simple weekly reference check with a known sample. Five minutes of verification per week beats a forty-minute argument with quality later.
Is the Fluke 189 true RMS multimeter still worth buying?
If you're looking at a used unit with a valid calibration certificate, yes. Fluke's published spec for the 189 lists 50,000-count resolution and 0.025% basic DC accuracy. It was a legit workhorse: true RMS, a bar graph, and data logging. It's been discontinued, so any "new" 189 is old stock. Check the calibration sticker before you pay a premium. I've seen used 189s listed for more than a new 117, and that only makes sense if you need logging or higher accuracy. The direct replacement in Fluke's lineup is the 289, which adds more memory and a color display, but also costs more and uses more battery. Honestly, I'm not sure why the 189 retains such a loyal following. My best guess is trust — engineers know what it can do, and they don't want to learn a new menu tree while standing in an electrical panel. If you're on a budget, a proven 189 with fresh calibration is hard to beat.
Which Fluke multimeter do I actually need without overspending?
Pick based on the job, not the badge. For someone who checks residential or commercial wiring, a Fluke 117 is plenty — it has the non-contact voltage detector and enough accuracy for line voltage troubleshooting. For plant engineers working with variable frequency drives and motor circuits, the 87V is the one we carry; its built-in low-pass filter lets you measure true RMS voltage and current on a VFD output without the carrier frequency corrupting the reading. If you're doing long-term predictive maintenance logs, the 289 makes sense because it can capture data trends overnight. I once watched a team buy 289s for everyone because it was the flagship. Most of those guys never touched the logging button. That's a 40% cost difference for unused capability. The non-negotiable is input safety: choose a meter rated CAT III or CAT IV for the environment you work in. That's not an upgrade. It's the minimum.
Where do hidden costs hide in this kind of procurement?
In four places: freight, calibration, compatibility, and wrong-part reorders. The sticker price on a sensor is maybe half of the first-year cost once you include setup, shipping, and the labor to install it. I audited $180,000 in cumulative sensor and test equipment spending over six years and found about 12% went to expedited shipping and replacement parts caused by mistakes. One example: a $250 encoder with the wrong shaft diameter. We only noticed after a technician tried to slide it onto the motor. That $250 became $1,200 in field labor. We didn't have a formal approval chain for part-number verification then. Now our purchasing policy requires quotes from three vendors for anything over $500, and every order goes against a checking list. The policy doesn't slow us down much. But it has saved us from the worst kind of cost: the one you only see after installation.
What's the most common mistake that causes rework?
Assuming the part number and the application actually match. I've made this mistake more times than I like to admit. One recent example: I ordered a SICK IO-Link encoder listed as "M12 connector." I didn't verify whether the M12 was A-coded or D-coded. The vendor shipped what I specified; the connector didn't fit the pre-wired cable. We were using the same word, meaning different things. Result: a 10-minute loop of calls, a reorder, and a weekend install delay. After that, I built a simple rule — before any PO goes out, compare the part number against the connection diagram in the datasheet. That's a 5-minute check. Since we started, I estimate we've avoided about $8,000 in potential rework. What's on the checklist? Part number, connection coding, supply voltage, output protocol, shaft size or mounting configuration. That's it. If both the buyer and the supplier review those five items before shipping, most rework disappears. The thing is, it doesn't happen unless you make it a process.
