Good PLC troubleshooting narrows a fault through safe observations and controlled tests instead of random online changes. A reusable log records symptom, operating state, evidence, hypothesis, test, result, decision, change, rollback, and final verification so the next responder can reproduce the reasoning.
The short answer
Make the process safe, preserve the initial state, and describe the symptom precisely. Check power, hardware diagnostics, communications, I/O quality, interlocks, sequence state, modes and recent changes before editing logic. Form one testable hypothesis at a time, record expected and actual results, use approved temporary forces only when authorized, and verify normal operation and cleanup after the correction.
- Safety, permits, production authority and an approved fallback precede online diagnosis.
- Preserve timestamps, diagnostics, active alarms, versions and recent changes before they disappear.
- Change one controlled variable at a time and distinguish observation from inference.
- Remove forces and temporary bypasses, restore the approved baseline and document final verification.
Decision guide
| Log field | Purpose | Good entry |
|---|---|---|
| Symptom/context | Define what failed and under which mode, product, time and sequence state | Observable behavior with timestamps and affected scope |
| Evidence/hypothesis | Separate facts from the proposed cause | Diagnostic buffer, I/O quality, network state, logic path and one falsifiable hypothesis |
| Test/result | Record a safe check and its expected outcome | Authorization, method, expected result, actual result and attached evidence |
| Correction/verification | Control the approved change and closure | Revision, approval, rollback, retest, force/bypass cleanup and operating confirmation |
A practical workflow
- 01
Secure the equipment state and confirm safety, operational authority, communication channels and rollback.
- 02
Capture the exact symptom, sequence state, alarms, diagnostics, versions, online differences and recent work.
- 03
Trace the signal and control path from field condition through I/O, communications, logic, interlocks and output.
- 04
Test the highest-value hypothesis with the least intrusive authorized method and record the result.
- 05
Apply a reviewed correction, test normal and abnormal paths, remove temporary measures and release the verified baseline.
Evidence and acceptance
The final log should contain ticket or event reference, date and timezone, equipment state, people and authority, symptom, diagnostic exports, versions and checksums, hypotheses, tests and results, forces or bypasses, approved change, backup, rollback method, retest, residual risk, operating confirmation, and follow-up action.
Common mistakes
- Editing logic before capturing diagnostics and the original version.
- Forcing I/O or bypassing an interlock without authorization, visibility and removal control.
- Changing several variables at once and losing causal evidence.
- Assuming a cleared alarm proves the root cause is fixed.
- Closing the event without backup, revision, retest, cleanup and reusable learning.
Engineering and security boundary
Primary sources
Technical claims and product boundaries were checked against these primary sources. Fact-checked
Frequently asked questions
What should be captured before going online?
Symptom, time, operating mode, safety state, active alarms, hardware and network diagnostics, released version, current backup and recent authorized changes.
When is forcing an I/O acceptable?
Only under the site’s authorized safety and operational procedure, with explicit scope, visibility, time limit, monitoring and confirmed removal.
How do you separate symptom from cause?
Record observed facts first, state a falsifiable hypothesis, predict the test outcome and update the hypothesis from the actual result.
Can AutoNexit diagnose a PLC live?
No. It preserves the evidence and reasoning; qualified personnel use approved engineering tools for live diagnostics.