A loop check on an analog input proves that a milliamp signal leaving a field device arrives at the controller, gets converted correctly, and shows up on the right tag in the right engineering units. It is the moment where the wiring drawing, the input scaling, and the tag database all get tested against reality at once. The technique is deliberately simple: you inject known currents at five evenly spaced points and read back what the host displays. When five injected values map cleanly to five expected percentages on the expected tag, you have proven the whole chain end to end, and when they do not, the pattern of the error usually tells you exactly what is wrong.
Loop-Check a 4-20 mA Input in one line: To loop-check a 4-20 milliamp analog input, you disconnect or source at the field device and inject 4, 8, 12, 16, and 20 milliamps with a loop calibrator, then confirm the PLC or DCS displays 0, 25, 50, 75, and 100 percent of range in the correct engineering units on the correct tag. This verifies that the wiring is continuous, that the raw-count to engineering-unit scaling is right, and that the point is not swapped with another, before the loop is trusted for commissioning.
The five-point method exists because two points can hide a scaling fault that five will expose. You put a loop calibrator in current-source mode at the field end, either in place of the transmitter or in series where the transmitter would drive the loop, and you drive 4 milliamps first. At 4 milliamps the host should show the bottom of range, which is 0 percent, and whatever engineering value that represents, such as zero pressure or an empty tank. You then step up to 8, 12, 16, and 20 milliamps, which correspond to 25, 50, 75, and 100 percent, and at each step you read what the controller displays. Evenly spaced injected currents should produce evenly spaced displayed values, and any point that does not fall on the line is a clue.
The single most important discipline is confirming the tag, not just the value. It is easy to inject 12 milliamps, see 50 percent appear somewhere on the screen, and declare success, when in fact the 50 percent showed up on a neighboring point because two field wires were landed on the wrong terminals. A proper loop check names the tag before injecting and watches that specific tag respond, ideally with a second person at the console calling out the tag identity and the value. Proving that this current on this pair of wires moves this named point is the whole reason the check exists, and skipping the tag confirmation defeats it.
The pattern of a failed check is diagnostic. If all five points are shifted by the same amount, the scaling has an offset error, often a zero-versus-4-milliamp mismatch where the input expects a live zero but is configured for a dead zero or the reverse. If the points are correct at the bottom and grow more wrong toward the top, the span is scaled wrong, and the error is proportional to signal. If the readings jump around or drop out, you are chasing a wiring or continuity problem rather than a configuration one. Reading the shape of the error before touching anything saves a great deal of guessing.
Inside the controller the milliamp signal is not stored as milliamps. The analog input card converts the current to a raw count, a number spanning the card's resolution, and the program then scales that raw count into engineering units using a configured low and high value. A loop check is really a test of that two-stage conversion. Many technicians find it worthwhile to watch the raw count as well as the scaled value during the check, because the raw count tells you whether the card is receiving the current correctly and the scaled value tells you whether the engineering conversion is right. A correct raw count with a wrong scaled value points squarely at the scaling configuration rather than the wiring or the card.
The classic scaling mistakes are all visible in a five-point check. A point configured for a 0 to 20 milliamp input rather than 4 to 20 will read 20 percent when you inject 4 milliamps, because the card treats 4 milliamps as one fifth of full scale instead of as the live zero. A point with the engineering range entered backward will fall as you raise the current. A point with the wrong high value will track correctly at the bottom and diverge toward the top. Because each of these leaves a distinct fingerprint across the five points, injecting only two points, or worse only checking 20 milliamps, can pass a loop that is badly misconfigured in the middle of its range.
It is also worth confirming the units label and the direction of any inversion. A level transmitter reading a tank might be configured so that a full tank is the high current, or it might be an inverted range where the high current means empty, and the loop check is where that intent gets confirmed against the drawings. Reading the engineering value out loud at each point, not just the percentage, catches units that are technically scaled correctly but labeled or ranged for the wrong quantity. The goal of this part of the check is that 12 milliamps not only shows 50 percent but shows the specific engineering value that 50 percent of that range is supposed to mean.
Beyond scaling, a loop check is the cheapest proof of end-to-end wiring continuity you will ever run. When an injected current at the field device produces the right response at the controller, you have implicitly proven that the conductors are landed, the terminals are tight, any barriers or isolators pass the signal, and the card channel is alive. This matters most at commissioning, when hundreds of points were wired by different people over weeks and the drawings may not match the field. A systematic loop check catches the crossed pair, the point landed one terminal high, and the transmitter wired to the wrong channel, all of which look fine on paper and fail only when a real signal is injected.
Doing the checks against a live control system rather than a bench simulator is what makes swaps visible, because a swap only reveals itself when the wrong tag lights up. This is where a cloud SCADA platform such as Merobix helps a commissioning crew, since the person injecting current in the field and the person watching tags can share the same live view, and the injected values are visible in the same historian the operators will use. Trending the point as you step through the five currents gives an unambiguous staircase on the correct tag, and if the staircase appears on a different tag, the swap is caught on the spot rather than months later during an upset.
The habit of loop-checking before commissioning pays off long after the plant is running. A documented five-point check gives you a baseline that a future troubleshooting session can compare against, so when an operator later reports a point reading strangely, you know it was proven good at startup and the fault is newer than the wiring. Recording the injected values and the displayed values, ideally captured automatically as trend data, turns the loop check from a throwaway commissioning task into a permanent record of how each analog input behaved when it was known to be correct. That record is often the fastest way to settle whether a suspect reading is an instrument problem or a scaling change someone made later.
Two endpoints can pass a loop that is misconfigured in the middle, especially when the input is set for the wrong range or has an inverted scaling that happens to touch the correct values at the extremes. Injecting 4, 8, 12, 16, and 20 milliamps produces five evenly spaced points, and any scaling offset or span error shows up as a point that does not fall on the straight line. The shape of the deviation across the five points also tells you whether the fault is an offset, a span error, or a wiring problem.
A swap only reveals itself when the wrong tag responds. If you inject current on a field pair and name the tag you expect to move, but a neighboring tag moves instead, the wires are landed on the wrong terminals or the transmitter is on the wrong channel. This is why a proper loop check confirms the tag identity, not just that some point somewhere showed the expected percentage, ideally with a second person watching the named tag at the console.
Watching the raw count alongside the scaled engineering value separates a wiring or card problem from a scaling problem. If the raw count tracks the injected current correctly but the engineering value is wrong, the fault is in the configured low and high scaling rather than the field wiring. If the raw count itself is wrong or erratic, you are dealing with continuity, the input card, or the current source rather than the software conversion.
Safety & engineering notice. This article is general educational information, not site-specific engineering, safety, or legal advice, and it does not reflect any particular facility. Standards and regulations (for example OSHA, API, IEC, ISO, NFPA, NIST, and NERC CIP requirements) change and vary by edition, jurisdiction, and application. SCADA and remote monitoring cannot verify physical isolation, atmosphere, lockout/tagout, permit status, or a safe go/no-go decision. Qualified personnel must perform site-specific engineering, hazard analysis, and safety review, and confirm current requirements with the authority having jurisdiction, before acting.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.