Last September, I was awake at 2:17 a.m. when my phone buzzed. The palletizer on Line 2 had stopped again, and the Z-axis was drifting about 10 mm high every few cycles. I got dressed, drove back to the plant, and made the mistake that changed my maintenance checklist forever: I assumed the first Balluff position sensor I saw was the problem.
I have been a maintenance planner for nine years. I have personally made—and documented—14 significant mistakes, totaling roughly $15,000 in wasted budget. This one was not the most expensive, but it was the most embarrassing. Now I maintain our team's pre-check list. This is why.
I asked for a Balluff LVDT sensor without checking the model
The axis used a rod-style linear sensor mounted along the guide rail. On the tag, it said “Balluff” and a model number. In our PM log, someone had written the line “replaces old LVDT.” In my head, every rod-style position sensor was an LVDT: linear variable differential transformer. That assumption was wrong.
From the outside, a position sensor is a position sensor. The reality is that linear position sensing has several different principles: magnetostrictive, inductive, LVDT, optical, even potentiometric. They are not interchangeable. The Balluff BTL series is magnetostrictive. An LVDT is a completely different device—it uses coils and produces a raw AC signal that needs external conditioning. The BTL has built-in electronics and can give you 4-20 mA, SSI, or IO-Link. They also mount differently.
I called the supplier and asked for a Balluff LVDT sensor based on that PM log. The sales engineer asked, “Do you need LVDT or magnetostrictive? Check the model number.” I didn't. I ordered the LVDT anyway. It arrived three days later and didn't even match the mechanical connector. That was $520 and another week down the drain. (I still wince when I see that invoice.) The delay was closer to two weeks—or rather, ten days—when you count the return shipping and the new part.
That was the first lesson: verify what is actually on the tag, not what someone wrote in the log.
I replaced the sensor before finding the real problem
After that, I ordered the correct Balluff position sensor. I was determined to make the installation perfect. I checked the mounting bracket with a Starrett angle finder because the alignment spec in the drawing called for a square mount. If you are not sure how to use a Starrett angle finder, the short version is: put it on a known flat reference, zero it, then hold it against the surface you need to measure. The digital display shows the angle in degrees. We found the bracket was 1.7 degrees off, shimmed it, and moved on.
When we powered the machine back up, the fault was gone for twenty minutes. Then the same drift returned, softer but still there. I wanted to blame the new sensor. My co-worker, thankfully, had more patience. He asked if I had checked the 24 V supply under load. I had not.
He grabbed the thermal imaging camera from the tool cabinet. On the 24 V distribution block, one terminal was visibly hotter than the wires around it—maybe 12 to 15 degrees Celsius hotter, I don't remember the exact number. Then he put a Fluke 36 clamp meter around the wire feeding that terminal. At idle, the current was normal. During a move, it spiked to over 1 amp, and the voltage sagged enough to make the sensor output flicker. The loose terminal was the real problem. The original sensor might have been fine. I had replaced it only because I didn't look past the part.
Here is the part that still bothers me: the new sensor had IO-Link, and if I had connected to it before running the axis, I could have seen the position signal drop during those voltage dips. Balluff's IO-Link tools would have told me the sensor was not the problem. Instead, I used the sensor as a scapegoat.
The checklist that stopped my repeat mistakes
In my opinion, the most expensive words in maintenance are “just replace it.” We have a pre-check list now. It is not a textbook six-sigma thing; it is a one-page sheet on the wall and in the CMMS:
- Confirm the exact model number and sensing principle: magnetostrictive, LVDT, inductive, optical, or something else. If the tag is damaged, call the OEM and ask.
- Measure supply voltage at idle and under load. A 24 V system that sags below around 22 V is a red flag.
- Scan the circuit with a thermal imaging camera before pulling anything. Hot terminals are a clue.
- Check actual current with a clamp meter on the sensor's supply branch. Spikes and drops matter more than the nameplate value.
- Verify the mechanical setup with a Starrett angle finder or an equivalent angle measurement tool.
- Record what you found, not what you expected. Future-me will thank past-me.
Since we started using this checklist, we have caught 18 potential problems in the past 18 months—things that would have led to unnecessary parts or extra downtime. One was a fused output on a PLC card that looked exactly like a sensor failure. Another was a bracket that had been knocked out of square, which we caught with the angle finder before ordering anything.
I am not saying the checklist is magical. At least, that has been my experience in a packaging plant with 20-year-old equipment. The change was not in the tools. It was in slowing down enough to ask: what else could cause this?
One more thing: this was in September 2024. Balluff's current catalog may have different model numbers. Check before you order.
If you are standing in a parts room at 2:00 a.m. looking at a Balluff position sensor, read the tag, check the supply under load, and give the sensor a fair trial. The sensor may be telling you the truth. The problem is if no one is listening.
