Troubleshooting Unfamiliar Industrial Equipment Step by Step
Troubleshooting unfamiliar industrial equipment without a manual: capture the symptom, map power and control, stay safe, then run checks in order.
Troubleshooting unfamiliar industrial equipment: start from the symptom, not the manual
When you are troubleshooting unfamiliar industrial equipment without a usable manual, start with the symptom written in plain words. Make the machine safe to touch, sketch a rough map of its power and control path from the nameplate and what you can see, then run checks from least to most invasive. Log every result, so each check narrows the list of likely causes.
Start with the symptom, not the documentation
A missing manual feels like the main obstacle, but the symptom is the one thing you already have. Write down exactly what the machine does and does not do. Does it hum but not turn? Does it trip ten seconds after reaching speed? Does it lose pressure only after warm-up? Each detail rules some causes in and others out. A vague label like 'broken' rules out almost nothing.
Capture four things before you touch anything:
- Timing: on start, under load, after warm-up, or at random.
- Indicators: which lights are on, which are off, and what any display says, copied word for word.
- Sound and smell: a hum, a click, a burning odour, a rattle that stops when the machine stops.
- Recent change: a new product batch, a changed setpoint, a replaced part, a power dip, a washdown.
Nobody on shift may remember the last change. Ask the operator who ran the machine last, and check the previous shift handover. If the answer is 'nothing changed', write that down too. It is still a finding.
Write the symptom as one sentence a colleague could act on: 'Conveyor drive trips on overcurrent about 20 seconds after start, only when the belt is loaded.' That sentence drives everything that follows. If you are unsure where to begin, start with the first minutes after the machine goes down.
Build a working map from what you can verify
A rough map gives you a place to start. Without one, every check looks equally likely.
Reading the nameplate and labels
The nameplate gives the voltage, current, power, speed and frame size the machine was built for. Drive and controller labels add model, input and output ratings, and sometimes a fault code table. Panel labels, terminal numbers and wire markers show what the builder considered important. Photograph every label before you open anything, because labels fade and cover panels get removed.
Tracing the power and control path
Follow the path from where power enters the cabinet. Note the disconnect, any fuses or breakers, the contactor or drive, and the motor cable. Then trace the control side: start and stop devices, emergency stops, interlocks, sensors and any PLC inputs you can see. Draw this as a block sketch with five boxes: source, protection, switching, load, control. Mark every guess with a question mark.
Treat the sketch as a hypothesis, not a diagram. Wires may have been re-routed, labels may be wrong, and a jumper may be doing a job nobody documented. A partial map is still enough. It tells you where power enters, which device is the likely switch, and which check is the least risky to run first.
Make the machine safe before any test
An unknown machine carries more risk than a known one. You may not know where stored energy sits: a capacitor in a drive, a pressurised accumulator, a spring-loaded brake, or a raised load. You may not know that a sensor can restart the machine automatically. Circuits may be unlabelled, so a wire you think is dead may be live from another source.
The basic order is short. Isolate the energy source at the disconnect. Lock and tag it so nobody can re-energise it while you work. Verify zero energy with a proper tester on the circuit you are about to touch, and confirm the tester works on a known source before and after. Do not rely on a panel light or a drive display.
This article does not repeat the full steps. Follow your site's lockout tagout procedure on every unfamiliar machine, even when the fault looks trivial.
Run checks from least to most invasive and log each result
Order your checks in three tiers, and do not skip ahead.
Sort checks by cost and risk
- Visual and physical: look for burn marks, loose terminals, leaks, blown fuses, a seized coupling, an open guard or a dirty sensor. With power isolated, turn the shaft by hand and note any stiffness or noise.
- Isolated measurements: winding resistance, insulation tests, continuity on a control circuit, and pressure readings on a hydraulic line. These need the machine isolated and a clear method.
- Powered tests: only after the safe method is confirmed and the people near the machine know what you will do. Change one thing at a time.
Cheap checks come first because they remove many causes at once. A loose terminal can explain a surprising range of symptoms. An expensive component swap explains very few.
Record pass and fail so the list narrows
After every check, write one line: what you checked, the result, and the time. For example: 'Motor winding resistance phase to phase, balanced, 14:20, passed; winding not the cause.' A passed check rules a cause down or out. A failed check points to the next branch. Without the record, the next shift starts from zero and repeats the same checks.
| Time | Check | Result | Effect on causes |
|---|---|---|---|
| 14:05 | Terminal tightness at drive | Loose on L2 | Supply connection ranked up |
| 14:20 | Motor winding resistance | Balanced | Winding ruled out |
| 14:35 | Brake release voltage | Not present | Brake circuit checked next |
The log is not paperwork for its own sake. It is the working memory the next person on the machine will rely on.
Where a diagnosis tool helps, and where your judgment still decides
Fault Line AI is a web app you open in a browser, so there is nothing to install. You describe the fault in plain words, for example 'motor hums but won't turn', and it returns a ranked list of likely causes and the next check to run. Each check comes with a reason, so you can see why it sits where it does.
Mark each check as passed or failed, and the list re-ranks and the next check changes to match. When a check passes, its cause is marked confirmed. Built-in step-by-step checks cover four fault families: motor hum or no start, hydraulic pressure loss, drive overcurrent or trips, and overheating or shutdown. For other faults you get a general list of checks.
Guided troubleshooting works without signing in, but its results are not saved, so they cannot be picked up later or on another device. To keep a record, sign in and save a symptom report with the machine name, asset tag and severity.
The limits matter. Fault Line AI does not connect to or control the machine, and it does not ask follow-up questions. It does not record the parts you used or the outcome of a repair, and it does not offer a library of repair procedures beyond its built-in checks. It suggests checks for you to perform. You read the labels, the sketch and the lockout status, and you decide what is safe. Use it to order your thinking, not to replace it.
Next time you face an unfamiliar fault, try the guided troubleshooting in Fault Line AI without signing in. Create an account only if you want to save a symptom report with the machine name, asset tag and severity.
Frequently asked questions
What do I do when there is no manual or wiring diagram for the machine?
Work from the nameplate, the labels and what you can trace on site. Build a rough sketch of power and control, mark what is a guess, and start with visual and isolated checks. Look for the OEM documentation, but do not stop work while you wait for it.
How do I tell whether a check is safe to run on a machine I don't know?
A check is safe only when the energy is isolated, locked and verified at zero for the circuit you will touch. If you cannot confirm that, do not run the check. Stop and ask someone who knows the machine, and follow the lockout steps before you continue.
Is a diagnosis tool worth using on a machine I have never worked on?
It can help you order checks and keep your reasoning in one place, especially for the four built-in fault families. It cannot see the machine or decide what is safe. Treat its ranked list as a starting point for your own checks.