Industrial instrumentation intervew quesrions

The Instrumentation Insider: Article #2 – The Deep Dive on Calibration, Logic, and Gremlins

Welcome back to the shop floor, folks. Last time we covered the basics—the 4-20mA lifeblood, the difference between a thermocouple and an RTD, and why water is the enemy of everything you hold dear.

Today, we are getting our hands dirty. We are talking about calibration logic—not the textbook "hook it up to a calibrator" stuff, but the why behind it. We are also tackling the gremlins: the ground loops, the digital handshakes, and that one valve that always sticks on a Tuesday morning.

Pull up a stool, wipe the grease off your glasses, and let's hit these 20 questions hard.

---

1. Why do we perform a "5-point calibration" instead of just a zero and span?

Answer: Because instruments aren't perfectly linear. A zero and span (2-point) check only tells you the endpoints are right. A 5-point check (0%, 25%, 50%, 75%, 100%) reveals the "linearity" error—that ugly S-curve in the middle. If the middle points are off, you need to adjust the trim tables in a smart transmitter, not just the zero and span screws.

2. What is the difference between "As-Found" and "As-Left" data?

Answer: As-Found is the reading before you touch the instrument. This is the most important data you will collect because it tells you if the instrument was drifting dangerously during operation. As-Left is the reading after you adjusted it. If you only record As-Left, you are hiding the evidence of a problem that existed for months.

3. I have a ground loop. How do I kill it?

Answer: A ground loop happens when you have two different ground potentials trying to use the same signal wire. The fix? Isolate. Use a signal isolator in the loop, or break the shield connection at one end. Pro-tip: Only ground the cable shield at the control room end. Leave the field end floating. It stops the shield from becoming a current-carrying conductor.

4. Can I calibrate a pressure transmitter with a hand pump while it's still in the process?

Answer: You can, but it's called a "field check," not a full calibration. You are introducing a known pressure at the manifold block while isolating the process. However, you are not checking the impulse lines. If those lines are plugged, your field check will pass, but the process reading will still be wrong. Always blow down the lines first.

5. What is "Damping" and why does my operator hate it?

Answer: Damping is a software filter that smooths out the signal to stop the display from bouncing around. Operators hate it because it creates lag. If you set damping to 5 seconds, and a pipe ruptures, the operator won't see the pressure spike for 5 seconds. Rule of thumb: Use as little damping as necessary to get a stable reading.

6. Why does my Coriolis flow meter read mass flow, but my DCS wants volume?

Answer: Because mass is the truth, and volume is a liar. Volume changes with temperature and pressure; mass does not. If your DCS wants volume, you have to feed the Coriolis meter's density output into the DCS and let it do the math (Mass / Density = Volume). Don't let the meter do it internally unless you have to.

7. What is "SIL" and why do I keep hearing about it?

Answer: SIL stands for Safety Integrity Level. It's a rating (SIL 1, 2, or 3) that tells you how reliable a device is at performing a safety function when you need it. A SIL 2 transmitter has a higher probability of working on demand than a SIL 1. If your plant is doing a HAZOP review, pay attention—they are deciding which loops need the expensive SIL-rated gear.

8. My Rosemount 3051 reads fine, but the HART communicator says "Communication Error." Why?

Answer: You have too much resistance in the loop. HART communication requires a minimum load resistance (usually 250 Ohms) between the power supply and the transmitter. If you are reading the 4-20mA fine on a multimeter, but can't talk to it, you don't have enough impedance. Throw a 250-ohm resistor in series with the positive lead and try again.

9. What is "Wet Leg" vs. "Dry Leg" in level measurement?

Answer: For a closed tank DP transmitter:

· Dry Leg: The high-pressure side connects to the bottom of the tank. The low-pressure side connects to the top (vapor space) and is left empty (dry). Used for clean, non-condensing vapors.

· Wet Leg: The low-pressure side is filled with a seal fluid (like glycol) to compensate for condensation. The transmitter is calibrated to subtract the constant hydrostatic pressure of that seal fluid. If the seal fluid evaporates, your reading goes to hell.

10. Why do we use "Square Root Extraction" for DP flow?

Answer: Because flow rate is proportional to the square root of the differential pressure. If you don't extract the square root, you get a parabolic reading (at 50% DP, you are only at 25% flow). In modern transmitters, you enable "Square Root Output" internally. In older systems, the DCS did it. If you enable it in both places, your reading will be wildly wrong.

11. What's the best way to test a thermocouple with a multimeter?

Answer: Check the resistance (continuity) and the millivolt output. A good thermocouple will show a very low resistance (less than 10 ohms). If it shows infinite resistance, the element is open. Then, heat the tip with a lighter and watch the millivolt reading climb. If it doesn't climb, the junction is dead.

12. My ultrasonic level transmitter lost its echo. Now what?

Answer: It's either "false echo" or "weak echo." False echo means it's hitting an obstacle (like a ladder or baffle)—do a "False Echo Learn" or "Sweep" function to tell the transmitter to ignore that return. Weak echo usually means the sensor face is covered in condensation or foam. Angle the sensor, or better yet, use a guided wave radar if you have heavy vapor.

13. What is the difference between "Fault Tolerance" and "Redundancy"?

Answer: Redundancy is having two of everything (e.g., two transmitters on one pipe). If one dies, the other takes over. Fault Tolerance is a single system that keeps working even when a component fails (e.g., a dual-chamber DP cell). Redundancy is expensive; fault tolerance is elegant. You usually want both for critical ESD valves.

14. Why does my control valve "stutter" when I'm at 10% open?

Answer: You are experiencing "Stiction" (Static Friction). The valve packing is too tight, or the seat is worn. The actuator pushes against the stem, but nothing moves until the pressure builds up enough to overcome the stick. Then it jumps past the setpoint, the controller backs off, and it sticks again. Time to check the packing nuts or lap the valve seat.

15. What is "Zero Elevation" and "Zero Suppression"?

Answer: This applies to DP level transmitters.

· Zero Elevation: The transmitter is mounted below the bottom tap. There is a constant head of liquid on it. You "elevate" the zero to ignore that static pressure.

· Zero Suppression: The transmitter is mounted above the bottom tap. You have to "suppress" the zero to tell it that 0% level exists when there is still some liquid head below it.

16. How do I deal with a "Noisy" 4-20mA signal?

Answer: Three culprits: (1) Variable Frequency Drives (VFDs) running too close to the signal cable—route the cable at least 12 inches away. (2) Bad shielding—replace the cable with a shielded twisted-pair. (3) A dying power supply—check the voltage at the transmitter; if it's below 10V, the signal will be unstable.

17. Is it okay to use Teflon tape on instrument fittings?

Answer: For NPT (tapered) threads, yes—but sparingly. Never use it on the first two threads (to prevent contamination from getting into the process). However, for compression fittings (like Swagelok or Parker), never use Teflon tape. The seal is made by the ferrule biting into the tubing, not the threads. Tape will just cause leaks and false torque readings.

18. What is a "Pulse" output and why does my flow meter use it?

Answer: Pulse outputs are digital signals (frequency) that represent a discrete volume per pulse (e.g., 1 pulse = 1 gallon). They are immune to electrical noise over long distances. Your DCS counts the pulses and multiplies them by the K-Factor to get total flow. If you lose a pulse, you lose accuracy—but you won't get a "4mA" drift like you would with analog signals.

19. Should I calibrate my smart transmitter in the field or the shop?

Answer: The shop is for the full calibration (temperature, pressure, linearity) in a controlled environment. The field is for the trim—a sensor-specific adjustment that accounts for the exact plumbing and orientation on the pipe. In reality? We do the heavy lifting in the shop during turnaround and only do a quick "re-range" and "zero trim" in the field.

20. What is the single biggest mistake new techs make?

Answer: They don't check the process conditions before they troubleshoot. They spend two hours recalibrating a transmitter, only to realize the pump was off, the block valve was closed, or the tank was actually empty. Always—always—verify the physical process first. Is there pressure? Is there flow? Is the valve actually moving? Your tools are useless if the process isn't running.

That's the real-world stuff, folks. Remember: 70% of instrument problems are mechanical (plugged lines, loose wires, moisture). Only 30% are electronic. Start with the simple, stupid stuff first.

Next up in Article #3: "The Digital Shift: Foundation Fieldbus, Profibus, and Wireless—Love Them or Hate Them."

Stay safe, keep your multimeter on the right setting, and I'll see you next time!