How to run an instrument loop check that holds up at turnover
The loop check is where an I&E job gets proven. Every transmitter, switch and valve has been installed, tubed, pulled, terminated and megged — and none of that matters until the tech at the device and the operator at the screen agree on what the number is. A good loop check record is also what the client's commissioning team, your QC manager and, if something goes wrong a year later, an investigator will look at. The free workbook above gives you the check sheet, a log to track every loop, and a calibration record, with the math done for you.
Before you start: what has to be done first
- Device installed and hooked up per the install detail, with tubing pressure-tested and the process connection isolated or in a safe state.
- Terminated both ends and the terminations checked — see the termination wire list template.
- Calibrated or bench-verified to the range on the instrument index.
- I/O card live and the tag configured in the PLC, DCS or SIS, with the HMI graphic available.
- Documents in hand: loop drawing, datasheet, alarm/trip schedule or cause-and-effect, and the permit for the work.
Section A — wiring, continuity and insulation
Ring out each conductor from the device to the field junction box and on to the marshalling terminals. Confirm the wire tags and ferrules match the drawing, polarity is right, and the shield is grounded at one end only (normally the panel end) and cut back and taped at the field end. Ground loops from shields landed at both ends are one of the most common causes of noisy 4-20 mA signals.
Insulation resistance: lift the conductors at the I/O card and disconnect the transmitter before you put a megger on the cable. An insulation tester will destroy loop-powered electronics and I/O channels. Use the test voltage your project spec calls for; our insulation resistance testing reference covers how readings are interpreted.
Section B — 5-point signal injection
Apply 0, 25, 50, 75 and 100% of the calibrated range at the process connection (pressure source, simulator) or simulate the output from the transmitter with a HART communicator. At each point record:
| Column | How it works |
|---|---|
| Applied value | LRV + % × (URV − LRV), in engineering units. Calculated. |
| Expected mA | 4 + 16 × %. Calculated. |
| Measured mA | What your meter reads at the I/O terminals. |
| mA error % span | (measured − expected) ÷ 16. Calculated. |
| HMI reading | What the operator screen shows. |
| HMI error % span | (HMI − applied) ÷ span. Calculated. |
| Result | PASS when both errors are within ± your tolerance. |
If the mA is right and the HMI is wrong, the problem is configuration — range, scaling or units in the controller. If the mA is wrong, it's the device or the wiring. Recording both is what lets you tell the difference on paper.
Section C — alarms, trips and interlocks
Drive the signal through each alarm and trip setpoint from the alarm schedule or cause-and-effect chart and record where it actually tripped. Then prove the action: the HMI alarm annunciates and logs, the pump stops, the ESD valve closes, the first-out indication is correct. Check what happens on a bad signal — under-range, over-range, open wire — and that the controller does what the philosophy says. Safety-instrumented trips are usually client-witnessed; get initials on the row.
Section D — final element and fail position
For control valves, command 0-25-50-75-100% and record the travel you actually see; the sheet calculates the deviation against your position tolerance. For every valve, pull the signal and the air and confirm it goes to its fail position (fail closed, fail open, fail last). Time the stroke both ways on on/off and ESD valves and confirm the open and closed limit switches match the HMI.
The loop check log
One line per loop, with planned date, status, tech and client witness. Status moves Not Started → Ready for Check → In Progress → Passed or Failed - Punch → Re-check Passed → Client Accepted. The progress block counts each status and shows % complete, and an overdue flag catches loops past their planned date. Keep failed loops visible with their punch number — the history is part of the record.
The calibration record
Record as-found before you touch anything, then adjust and record as-left, five points each. Error % of span = (measured mA − expected mA) ÷ 16. Enter your test standard and its calibration due date: if the standard was out of cal on the calibration date, the sheet flags the record "OUT OF CAL — results void", because it is. Next due date is calculated from your calibration interval. An as-found failure on an instrument that's been in service is information someone needs — don't adjust it away without writing it down.
Mistakes that cost re-checks
- Checking the mA and skipping the HMI. Half of loop check failures are scaling or configuration, not wiring.
- Verifying the alarm but not the action. The alarm banner came up; did the valve close?
- Megging with the transmitter connected. Expensive, and it happens every job.
- Signing a loop complete with open items. If a section failed, the overall result says FAIL until it's re-checked.
- Out-of-date test equipment. A calibrator past its due date undermines every record it touched.
Beyond paper check sheets
Printed check sheets get rained on, lost in a truck and signed a week late. Field PM QA/QC workflows step each loop through wiring, injection, alarms and client acceptance with hold points and e-signatures, and the form library ships with ready-made industrial QC forms. Signed records file into the job book automatically. Start a free 30-day trial.
Related templates
- Instrument index & I/O list
- Termination wire list, tubing pressure test & hazardous area inspection
- Inspection request log — schedule client witness points
- Punch list template — track failed-loop punch items
- Ohm's law calculator — loop resistance and voltage checks
