TL;DR: PLC faults fall into four families — wiring, configuration, logic, and intermittent. A systematic 7-step method (define symptom → confirm power → read the ladder → check inputs → trace field wiring → check outputs → verify fix) eliminates random probing and finds root cause in the fewest steps. Multimeter discipline and the half-split method do the heavy lifting. Practising on live fault scenarios before arriving at a real machine dramatically shortens your first-time diagnosis.
Every experienced controls technician has stood in front of a machine that should be running and isn't. The new technician starts poking randomly. The experienced one pauses, defines the symptom precisely, checks the power supply, reads the ladder, and finds the fault in eight minutes. The difference is not talent — it is method.
This guide covers the full systematic approach: the four fault families, the 7-step method, multimeter discipline, and the half-split technique that halves the fault window at every step.
The Four Fault Families
Before any diagnosis, you need a mental map of where PLC faults live. There are four families. Your job is to identify which family you are in, then apply the correct test procedure for that family.
| Fault family | Typical symptom | First check | Tool |
|---|---|---|---|
| Wiring fault | Input never changes state despite field device operating | Supply voltage at device, then continuity back to card | Multimeter |
| Configuration fault | Input changes but PLC ignores it, or wrong value | Verify tag address, I/O card slot, scaling in software | Software monitor |
| Logic fault | Rung stays false even when all inputs appear correct | Contact types (XIC/XIO), interlock conditions, coil address | Ladder monitor |
| Intermittent fault | Fault clears itself, reappears under load or heat | Connector seating, shielding, data logging over time | Multimeter + data log |
The majority of field faults are wiring faults. That is the family to eliminate first.
The 7-Step Method
The 7-step method is the structured path from symptom to root cause. Every step eliminates a portion of the fault window. Skip a step and you risk missing the fault or replacing parts unnecessarily.
Step 1: Define the symptom precisely
Write it down before touching anything. "Motor does not start" is a symptom. "I think the proximity sensor is bad" is a hypothesis — write that separately. You need the symptom isolated before you start forming hypotheses, or you will confirm the hypothesis you already believe instead of finding the actual fault.
Good symptom definitions:
- "Conveyor does not start when operator presses Start button"
- "Motor trips on overload within 30 seconds of start, no load"
- "Output Q0.3 energises but the contactor does not pull in"
Step 2: Confirm power — PLC running, I/O powered, no fault lights
Before reading any ladder logic, confirm the system is actually powered and healthy. Check:
- PLC RUN LED is green (or equivalent healthy state)
- I/O cards have their power LEDs lit
- No fault/error LEDs are active on the processor or cards
- Supply voltage at the 24 VDC or mains distribution point is within tolerance
A PLC in fault mode, or a card with a blown supply fuse, will show symptoms that look identical to logic or wiring problems. Eliminate the power layer before moving on.
Step 3: Read the ladder — find the output that should be ON
Open the online ladder monitor and find the output coil for the device that is not operating. Look at the entire rung. Which contact in the rung is false when it should be true? That false contact is your fault target.
In a well-structured program, the output coil name tells you exactly what is supposed to happen. Follow the rung from left rail to right until you find the break in power flow.
Step 4: Check inputs at the I/O card
For the false contact you identified in Step 3:
- Check the input LED on the physical I/O card for that channel — is it lit?
- Compare the LED state to the tag state in the ladder monitor.
Two scenarios:
- LED is off, tag is false: The problem is in the field — the field device or its wiring is not delivering the signal. Move to Step 6.
- LED is on, tag is false: The problem is in software — wrong tag address, card addressing mismatch, or a configuration fault. Check the I/O tree in the software and verify the correct input channel is mapped to the correct tag.
Step 5: Trace field wiring from card to device
Now you are in the wiring layer. Use the half-split method (described below) to find the break.
Start by measuring voltage at the I/O card terminal. If no voltage is present, the supply or return is broken. If voltage is present at the card but the LED is not lit, the card input itself may be suspect — but check the common return before condemning the card.
Work backwards from the card toward the field device. Each measurement halves the remaining fault window.
Step 6: Check the output card and field actuator
If you have reached this step, the input chain is healthy and the logic says the output should be ON. Now check the output side:
- Is the output card LED lit for that channel?
- LED lit but actuator not responding: the problem is in the wiring from the card to the actuator, or in the actuator itself.
- LED not lit but tag is true: the output card or its fuse has failed.
- Measure voltage at the actuator terminals while the output is commanded ON.
- Correct voltage present, actuator not moving: the actuator is the fault.
- No voltage at actuator terminals: trace back along the output wiring to find the break.
Step 7: Apply the fix and verify
Make the minimum change needed to address the root cause. Then:
- Clear any test forces you applied during diagnosis.
- Cycle the machine through a full operational sequence.
- Confirm the original symptom is gone.
- Confirm no new symptoms have appeared.
- Document the fault, root cause, and fix in the maintenance log.
Step 7 is frequently skipped. A machine that starts after a fix is not the same as a machine that has been verified. Verify before you close the job.
Multimeter Discipline
A multimeter is the primary diagnostic tool for wiring faults. Used incorrectly it gives misleading readings. Used correctly it finds any wiring fault in a few measurements.
The correct sequence:
- Set the range before probing — never probe first and then range. An auto-ranging meter on a live 415 V circuit takes a fraction of a second to settle. In that time you can misread a voltage that will mislead your diagnosis.
- Verify the meter first — test the meter on a known voltage source before moving to the suspect circuit. Dead batteries cause the meter to read low. A faulty lead causes floating readings.
- Confirm supply voltage at the source — before probing individual devices, confirm the supply rail is at the correct voltage. A 24 VDC supply sagging to 18 V explains everything.
- Measure at the device terminal — if the device terminal reads the correct voltage and the device is not responding, the device is the fault. If the terminal reads 0 V, the supply side has a break.
- Use continuity mode only on de-energised circuits — continuity mode sends a small test current. On a live circuit, other voltage sources confuse the reading and you risk a short. Always de-energise before using continuity.
- For a sensor: check supply pin, then signal pin — brown (+24 V), blue (0 V), black (signal). If brown and blue are correct but black never switches, the sensor is faulty or its sensing face is obstructed.
- Document every reading — write down where you probed, what you expected, and what you measured. If the fault is intermittent, the log becomes your evidence.
The Half-Split Method
The half-split method is binary search applied to a control circuit. Instead of testing each component sequentially from one end, you start at the midpoint of the circuit and eliminate half the remaining fault window with each measurement.
For a circuit from supply through fuse → OL contact → coil → return:
- Measure at the midpoint: the OL contact output.
- Correct voltage: the fault is between the OL contact and the coil (second half).
- Zero volts: the fault is between the supply and the OL contact (first half).
- Measure at the midpoint of the confirmed half.
- Repeat until isolated to a single component.
A 16-component circuit has at most 4 measurements to isolation with half-split, versus up to 16 measurements testing sequentially. For intermittent faults where you have limited time in the fault state, speed matters enormously.
The midpoint of a PLC control circuit is almost always the boundary between the field side and the I/O card. Starting there immediately identifies whether the fault is in the field or in the PLC system — the most useful split available.
Handling Intermittent Faults
Intermittent faults are the hardest to diagnose because the fault clears itself before you can isolate it. Four approaches work:
Data logging: Configure the PLC to log the timestamp and state of the suspect input or output to a data file or historian. When the fault reappears, the log shows you exactly what changed and in what sequence.
Force the fault condition: If you know the conditions that trigger the fault (high ambient temperature, specific machine position, high load), recreate those conditions deliberately and run through your half-split.
Vibration test: Many intermittent faults are loose connections or cracked solder joints that only open under vibration. With the system safely powered down, apply mechanical vibration to suspect cables and connectors while monitoring continuity.
Environmental correlation: Log fault timestamps against shift start times, ambient temperature readings, and machine cycle counts. Patterns like "always after 2 hours of runtime" or "always in the first 10 minutes of a shift in winter" point to thermal or condensation causes.
Applying the Method to the Four Fault Families
Wiring fault: Enter at Step 5. Confirm supply voltage first, then half-split from the I/O card boundary toward the field device. Multimeter in voltage mode throughout.
Configuration fault: Enter at Step 4. The input LED and the tag state disagree. Check the I/O configuration tree — verify channel assignment, card slot addressing, scaling, and tag cross-reference. No multimeter needed.
Logic fault: Enter at Step 3. The inputs are correct but the rung is false. Force each suspect input in the online monitor and watch power flow through the rung. A forced input that does not energise the rung confirms that contact is the problem. Check contact type (XIC vs XIO), the correct tag name, and whether an interlock condition upstream in a different rung has suppressed the permissive.
Intermittent fault: Start with data logging before any hands-on diagnosis. You need evidence of what happened and when before you can half-split anything.
What Needs a Real Meter
A browser simulator — including ours — cannot replace hands-on experience with a real multimeter and a live panel. Here is what you genuinely need real hardware to learn:
- Developing the muscle memory of safe probe placement on live terminals
- Recognising the physical signs of a loose connector or burned contact
- Understanding the feeling of a high-resistance connection versus a good one
- Correctly identifying polarity, phase, and common returns by sight in a real wiring arrangement
- Responding safely when something unexpected happens during live probing
The simulator builds the diagnostic reasoning — the mental model of the circuit, the habit of following a method, and the recognition of fault patterns. That preparation makes your time on real hardware much more productive.
Frequently Asked Questions
Q: What is the most common PLC fault in industry?
A: Field wiring faults account for the majority of PLC-related maintenance calls — typically 60–70% depending on the industry. Open circuits from vibration-loosened terminals, corroded connections in outdoor enclosures, and damaged cables from repeated flexing are the most frequent causes. Logic faults and I/O card failures are less common but take longer to diagnose when they occur.
Q: How do I know if a PLC fault is a hardware fault or a software fault?
A: Use the I/O card LED as the boundary. If the input LED state and the tag state in the ladder agree — both false, both true — the problem is either in the field wiring (LED off when it should be on) or in the logic (tag is right but the rung is still false). If the LED and the tag disagree, you have a software addressing or configuration fault. That split identifies which layer the fault lives in.
Q: When should I force an output to test it?
A: Force an output when the output coil is showing true in the ladder but the actuator is not responding. Forcing confirms whether the problem is in the ladder (the coil isn't really true) or in the output card and field wiring (the card isn't converting the tag state to a physical signal). Always clear forces before returning the machine to service, and always confirm it is safe to energise the actuator before forcing.
Q: How do you find an intermittent PLC fault?
A: Data logging is the most reliable tool — configure the PLC to timestamp transitions on the suspect input. If the fault is wiring-related, flex and load the cable while monitoring continuity with the circuit de-energised. If it is temperature-related, correlate fault timestamps with ambient temperature logs or heat gun the suspect area while monitoring the input. Intermittent faults often have environmental patterns that point you directly to the cause.
Q: What is the half-split troubleshooting method?
A: The half-split method applies binary search to a control circuit — instead of testing each component from one end, you measure at the midpoint of the circuit to eliminate half the fault window at once. If voltage is present at the midpoint, the fault is in the second half; if absent, the fault is in the first half. You then halve the remaining window with the next measurement. This finds any single fault in a linear circuit in log2(N) measurements instead of N measurements.
Practise the 7-step method on live injected faults in the browser: the PLC fault scenario track starts with a free first scenario that walks you through the full diagnosis cycle — symptom, ladder read, input check, fix, and verify. The wiring fault labs add the multimeter layer — supply voltage, signal checks, and continuity on a simulated control circuit.
For the complete troubleshooting training hub — all eight fault scenarios, the eight wiring fault labs, and the fault family catalogue — see the PLC troubleshooting hub.
