The Turck Sensor Tester Can't Test Everything: A Field Lesson on Encoders and Multimeters

The phone rang at 4:47 p.m. on a Thursday. Our lead integrator was on a customer's packaging line, and his message was short: “The line is down. The new Turck encoder isn't putting out A/B pulses. I think the encoder is bad.”

I didn't start by ordering a replacement. That was the habit I had to unlearn after I implemented our incoming verification protocol in 2022. Before that, we would swap a sensor, see the same fault, and then spend an afternoon chasing a wiring or configuration issue. A sensor that “doesn't work” is a symptom, not a diagnosis.

I'm the quality manager for a machine builder. We use Turck sensors on most of our standard machines because the documentation is consistent and the electrical specs are repeatable. I review every sensor spec before it gets released to the shop floor—roughly 200 line items a month. I've rejected about 12% of first-pass supplier submissions in 2025 because the part number was close but not exact. So when a Turck encoder failed on a brand-new line, I wanted to see the actual signal, not just the part number.

Start With What You Can Test Quickly

The machine had a coolant loop with two flow sensors, and one of them had caused a nuisance trip earlier that week. We checked those first. Those particular flow sensors are switching type, not analog, so the quickest check was the Turck sensor tester.

The Turck sensor tester in my bag is a discrete-output tester. It powers a switching sensor and shows whether the output is behaving the way it should. For inductive proximity sensors, switching flow sensors, or NAMUR sensors, that tool saves a lot of time. The flow sensor passed. One variable down.

A passing result doesn't tell you whether the set point is appropriate. It only tells you that the output changed when it was supposed to. But for that moment, it was enough to move on.

Then I moved to the encoder.

The encoder was an incremental Turck encoder with A/B outputs and a zero pulse going back to a PLC counter card. And this is where the Turck sensor tester does not help.

It is not built to test high-speed pulse trains. It checks on/off states. Encoder quadrature is a timing question, not a state question. You need a counter, an oscilloscope, or at least a meter with a good frequency range to see what the encoder is actually doing.

The Right Tool Question

Jordan, one of the maintenance electricians, was standing behind me watching. He had two multimeters in his hand and a look that said the next question was coming.

“Since you're the quality guy,” he said, “what is the best multimeter for electricians?”

I hear that exact question a lot. My answer sounds like a dodge, but it isn't: the best multimeter is the one that can answer the question you're actually asking.

For the encoder check, I needed a meter with DC volts, a frequency range, and enough resolution to measure a 4-20 mA loop without loading it down. For someone who mostly works on 24 V control panels, that set of functions matters. For someone who works around 480 V VFDs, a meter with a low-pass filter or a VFD mode is worth more than a fancy display. If you don't need it, don't pay for it.

There is no single model I recommend to every electrician. A meter that sits in a drawer because it is too bulky is worse than a basic meter that gets carried. That is not a popular opinion, but it is the one I give to our team.

It Ended With a Ground Screw

The multimeter got us partway, but the real breakthrough was a simple pulse-count test. We powered the Turck encoder from the same 24 V supply, rotated the shaft by hand, and watched the counter in the PLC program. The count moved cleanly. That isolated the problem: the encoder was generating pulses, and the PLC input card was seeing them.

That meant the issue was between the encoder and the card under normal running conditions. It was a wiring or noise issue, not a sensor issue.

Then we powered up the drives. The count became unreliable. Sometimes it added pulses when the conveyor was stopped, and sometimes it missed counts when the conveyor was moving.

The cause had nothing to do with Turck. The shield drain wire on the encoder cable was landed on a painted panel screw instead of the ground terminal. That is not a reliable ground. With VFDs running in the same cabinet, noise got into the A/B lines.

Once we terminated the shield on the ground bus and rerouted the cable away from the VFD output wiring, the counter was stable. The original Turck encoder went back in and ran for the rest of the week without missing a pulse.

The fix cost us about $18,000 in travel, downtime, and lost production. We didn't have a formal cable shield verification step in our startup checklist. That gap caused the whole mess.

The Brand-Name Question Isn't a Test Result

Later, Jordan looked up at the vision camera over the label applicator.

“All right,” he said. “Does Cognex use Sony sensors?”

I get that one too. He had seen a spec sheet that mentioned Sony and wanted to know if that made the camera more trustworthy. The honest answer is: some Cognex models use Sony image sensors, but that name alone is not a pass/fail result. I check the spec sheet for the exact camera when it matters. The image sensor is one component; the lens, the lighting, the image processing, and the vision tool decide whether the system will work on your line.

The same logic applies to the Turck encoder and flow sensors. A brand name is a good starting point, but it doesn't replace verification. Most buyers focus on the part number and completely miss the shield ground, the input card, or the output type. Those overlooked parts are usually where field failures hide.

So what would I recommend?

I recommend Turck for a lot of applications because the specs are repeatable and the documentation is clear. But I also tell people when it might not be the right fit. If you need a calibrated mass flow meter for a low-flow chemical process, a switching flow sensor is the wrong tool. If your servo drive expects a 1 Vpp sine encoder, don't order a standard incremental encoder and hope it works. Pick the output type and signal level that match the controller.

The Turck sensor tester is still on my bench. I'd buy it again. But no tool should replace a startup checklist.

The lesson from that Thursday stuck with me: don't trust the failure report. Verify the signal at the point that matters. That saves more money than any brand badge.