I85PREP
⌘ K
ISA CCST Level II study materials
ISA CCST Level II study resources

ISA CCST Level II: Calibration

ISA CCST Level II · Work through your exam, one topic at a time.

All guide sections

Calibration

The Calibration domain tests whether you can run a defensible calibration: choose a traceable standard, compute error against tolerance, record as-found and as-left data, and decide what to do when a device fails. Level II expects you to know the difference between calibrating an instrument and calibrating a loop, and to understand smart-transmitter adjustments.

Core rules

  • Trace every reference standard to a national standard. A common rule is a 4:1 test-uncertainty ratio between the reference and the device under test; 3:1 is sometimes accepted with justification.
  • Tolerance is the permissible deviation, usually expressed as a percentage of span. Error is calculated in engineering units, not just in milliamps.
  • Perform a five-point check, typically 0, 25, 50, 75 and 100 percent, both upscale and downscale. Record hysteresis and deadband.
  • Record as-found data before any adjustment. If as-found is out of tolerance, perform an impact assessment before returning the device to service.
  • On HART transmitters, distinguish sensor trim from output trim. Zero trim should not be used to correct a sensor error; use the correct trim procedure and a traceable input.
  • A loop calibration includes the field device, wiring, input card, controller scaling and HMI indication.

Worked example

A 0-100 psi transmitter has a tolerance of ±0.5 percent of span, or ±0.5 psi. At a 50 psi input, the output reads 11.90 mA. The expected output is 12.00 mA, so the error is -0.10 mA. Convert to engineering units: 0.10 mA ÷ 16 mA × 100 psi = 0.625 psi. That exceeds the ±0.5 psi tolerance, so the as-found result fails. Do not simply adjust the span; first determine whether the error is zero shift, span error or sensor drift, then trim and repeat the five-point check.

Traps

Adjusting span to compensate for a zero error; using a reference with only a 2:1 ratio; recording as-left values as as-found; testing only 0 and 100 percent; ignoring ambient temperature effects; and failing to document the serial number and calibration due date of the test equipment.