By Jeffrey C. Keeton, PE, Finleyville LabsAS WITH MANY
troubleshooting adventures, once you arrive at the source of a problem the solution seems to have been screaming at you the whole time. I received a call about an erratic level transmitter. Big John, our instrument technician, said the transmitter (in a bubbler application) was intact, had air on it, and its terminal housing was not filled with water (it had been raining so hard even Noah’s project was postponed due to weather). He showed me the HMI computer screen where the level was displayed and the value was jumping around erratically.
“Did you check the current in the transmitter loop?” I asked in what I thought was a relatively benign tone. “No,” he replied, “I climbed to the top of the @#$#@ tank to check the instrument and then forgot how. Of course I checked it, it was jumping around just like on the computer. I've never seen a transmitter do this before. I don't think there’s anything wrong with my instrument, it’s your @#$#@ computer that has a problem!” His response was surprising. He’s usually in a foul mood. The rain must agree with him.
I decided to talk to the system operator. He hadn’t been out in the rain nor required to climb to the top of any tanks lately. He told me there were a number of instruments acting funny and proceeded to show me on his HMI display screens. The screens revealed several normally stable analog values now jumping around erratically (the signals were from pressure, level, temperature, and flow transmitters). In this part of the plant, we have twenty (20) Modicon series-800 and series 984-145 PLCs, networked together via Modbus-Plus with a PC based Intellution HMI. Almost exclusively our transmitters are the standard garden variety type with 4-20 mA loop powered inputs to the PLC.
With multiple analog transmitters going haywire, I started looking for the common threads and the first suspect was the power supply. We have one linear 24VDC power supply supplying all I/O for a given processor. “Besides the erratic displays, is your process still running normally?” "Well," he said, "my pumps are still running and my feed valves are staying open, but I'm getting all kinds of alarms, high level, low level, you name it."
So the problem was common to the analog I/O but apparently not to the discrete I/O since the valves and motor relays seemed to be functioning fine. I brought up a configuration screen displaying raw input counts from the analog transmitters. That's when the janitor walked by and said, “The numbers are going up and down like a wave.”
Aha! A sine wave! A diode in the power supply had shorted allowing partial AC through; enough to vary the analog signals; yet not enough to seriously disturb the DC solenoids and relays. Sure, the janitor had just proven his ability to troubleshoot control systems, but I have yet to prove myself competent to sweep a floor.
A friend once told me when he experience this type of situation, he envisioned himself standing there with a cartoon donkey head. I scribbled down the part number from the power supply and quickly left for the store room before big John decided to pin a tail on me.
Jeffrey C. Keeton, P.E. works at Finleyville Labs, Inc. He can be reached by phone at 724-348-0714; by fax at 724-348-8326, or by e-mail at Jeff@FinleyvilleLabs.com