KEYSTONEIndustrial Services
Interactive trainer

PLC digital I/O troubleshooting

A 24 V DC I/O circuit: a START button into an input card, a rung in the CPU, an output card driving an interposing relay. Six faults, a voltmeter, and the skill that separates PLC people from parts-swappers: reading the module LEDs against real terminal voltages.

Before you start — job brief

PLC I/O runs on 24 V DC and is safe to probe live — that is how you troubleshoot it. The discipline here is different: never force an output to "test" it with people at the machine, and never trust an LED by itself. An LED is the module's opinion; the meter reads the truth at the terminal, and the fault lives wherever they disagree.

  • The program is given as correct in this trainer — the rung monitor shows you exactly what the CPU sees
  • Input LED = the card's side of the story; terminal voltage = the field's side
  • Work the boundaries in order: field device → terminal → module → program → output terminal → load
  • Forcing I/O bypasses every interlock in the program. It is a test of last resort, done with the machine safe — never a fix

Every I/O call is a boundary problem

A PLC digital I/O circuit is a chain of hand-offs: a field device hands 24 V to a terminal, the terminal hands it to an input point, the point hands a state to the program, the program hands a command to an output point, and the output hands power to a load. The fault always lives at exactly one hand-off. Troubleshooting is not inspecting everything — it is finding the one boundary where the two sides disagree.

The two ways an LED can lie

Module LEDs are the most useful and most misread indicators on a panel, because each one only reports its own side of a boundary.

  • Input LED dark with a fed terminal. The field delivered 24 V and the card refuses to acknowledge it — a failed point, or a lifted common taking out the whole card at once.
  • Output LED lit with a dead terminal. The LED follows the logic state, not the copper. A lit output LED over a 0 V terminal is a failed output point every time.

In both cases the meter at the terminal is the tiebreaker. The LED is the module's opinion; the voltage is the fact.

The order that solves it

  1. Prove the field supply. Rail to 0 V after the fuse. A dead rail explains every dark LED at once — stop and fix that first.
  2. Prove the device. Operate it and read its output. A button that never delivers 24 V is a device or supply problem, not a PLC problem.
  3. Prove the terminal. The same 24 V must arrive at the input terminal. Device good, terminal dead — the wire between them is open, and the full supply sits across the break.
  4. Check the LED against the terminal. Fed terminal, dark LED — the module side failed. Before condemning the point, read terminal-to-COM.
  5. Read the rung. With inputs proved, the monitor shows whether logic solves. In this trainer the program is given correct; in real life this is where you separate hardware from configuration.
  6. Prove the output terminal. Rung true and LED on mean nothing until the terminal reads voltage. Dead terminal, lit LED: output point.
  7. Prove the load. Voltage across a relay coil that never clicks convicts the coil. From there you are into the starter circuit — a different drawing, and our contactor trainer takes over.

Terminal-to-common: the reading everyone skips

Probing a terminal against the 0 V bar tells you what the field delivered. Probing terminal-to-COM tells you what the point actually sees — and they differ in exactly one important failure. When the card's common lifts, every terminal still reads 24 V to the bar, yet terminal-to-COM reads 0 and COM itself floats up to 24. One dead point is a point or a wire; a whole card of dark LEDs over perfect field voltage is the common, almost every time. The normally-lit NC STOP input going dark with everything else is the giveaway you can spot from across the room.

Sinking, sourcing, and why the card cares

Current needs a loop. A sinking input card returns current from the field device to its COM; the field device sources the 24 V. Swap in a device that expects the opposite arrangement and you get the classic "good device, dead input" call — the voltage is present, but no loop exists for the point to measure. If this circuit looks familiar, that is because it is the same idea as our PNP/NPN sensor trainer wearing work boots.

Forcing is a test, not a fix

Every PLC will let you force an I/O point, and there is a place for it: proving a wiring path on a safe, empty machine. But a force bypasses every interlock the program owns, and it does not clear when you walk away — it waits. Force nothing while people are at the machine, log every force you set, and clear them before you leave. A forced output is not a repair; it is a loaded tool left on the bench.

Practice the rest of the panel

This trainer is one of 52 free tools we publish — the VFD, three-phase motor and 4–20 mA loop trainers pick up where this one ends. No sign-up; they work on a phone in front of the machine.