Excellent PLC
By Michael Reeves – Maintenance Supervisor, Petrochemical Facilities
I’ve been around long enough to know that when a non-contact current sensor starts acting weird, nine times out of ten it’s not actually broken. The Bently Nevada 163947-00-05-02-84-90 proved that rule again during a routine inspection at our unit last year.
What looked like a sensor issue turned out to be a classic grounding oversight.
What We Noticed First
Operators complained about:
-
Random current spikes showing up in trends
-
Nuisance alarms with no real load change
-
Different readings between identical motors
-
No clear pattern tied to process conditions
The motor itself was running smoothly. Vibration data was clean. Only the current signal looked suspicious.
Early Assumptions (Mostly Wrong)
At first, everyone pointed fingers at the sensor:
-
“Maybe the sensor electronics are drifting.”
-
“Could be a bad batch.”
-
“Let’s just swap it and see.”
I’ve learned not to replace parts until I understand why they failed—especially when the part costs real money.
Digging Deeper
We checked:
-
Sensor mounting and alignment
-
Cable routing
-
Power supply stability
-
Shield termination points
That’s when it became obvious: grounding was inconsistent across the cabinet.
Some shields were grounded at both ends, some at neither, and one grounding bar wasn’t bonded properly to plant ground.
The Real Cause
The issue wasn’t the 163947-00-05-02-84-90 at all.
It was:
-
Ground loops introducing low-frequency noise
-
Floating reference potential affecting signal stability
-
Electrical noise from nearby VFDs coupling into the sensor wiring
Once the grounding reference shifted, the sensor output followed it—exactly as physics says it should.
What Actually Fixed It
-
Standardized single-point grounding for sensor shields
-
Bonded cabinet ground bar to verified plant ground
-
Rerouted sensor cables away from power conductors
-
Verified grounding continuity during commissioning
After that, the noisy spikes disappeared instantly. No sensor replacement needed.
Lessons From the Field
-
Non-contact sensors still depend on clean electrical reference points
-
Bad grounding can mimic sensor failure perfectly
-
Replacing hardware without fixing grounding just wastes budget
-
If multiple sensors misbehave at once, look at infrastructure first
Final Thoughts
The 163947-00-05-02-84-90 is a solid piece of hardware. In our case, it did exactly what it was supposed to do—respond honestly to a bad electrical environment.
Fix the ground, and the “sensor problem” magically goes away.
That’s maintenance reality.
— Michael