Skip to content
Faultpath

Reference

Troubleshooting patterns

Eight moves chosen because they should generalise beyond the machine they are practised on. That is the design intent and not yet a measured result — no study has been run showing they predict behaviour on unseen equipment. They are what the diagnostic profile measures, and being told you missed one is not much use without knowing what it was.

Every example below is taken from a scenario you can play. None of it is written twice — if a move appears on three different machines here, that is the case for it generalising, made with worked examples rather than an assertion. It is a case, not a proof: the test is whether it predicts behaviour on a machine the technician has never seen, and that experiment has not been run.

Expected versus actual machine state

3 machines · 14 scenarios

Read what state the machine expected to reach against the state it is in, and treat the gap as the question.

Upstream / downstream localisation

3 machines · 14 scenarios

Establish where in a chain the quantity is still correct and where it is not, and narrow the fault to the span between.

Correlating symptoms with history

3 machines · 13 scenarios

Use when it started and what changed to separate wear from failure, and failure from something someone did.

Supply present versus load responding

3 machines · 3 scenarios

Confirm that supply reached a device before concluding the device is at fault — and know that supply present is not proof the device is drawing it.

  • Line 4 case conveyor

    control voltage reaches the coil, so the wiring to it is intact and the coil is what is not responding

    Performed by: Measure voltage at the contactor coil.

    Conveyor commanded to run but nothing happens
  • P-201 transfer pump

    the motor is drawing a reduced but real load, which excludes a disconnected drive train

    Performed by: Measure motor current on all three phases.

    P-201 makes some pressure but nowhere near duty
  • PK-3 case packer

    24 V reaches the coil terminals — which proves the circuit got there, not that any current is flowing through it

    Performed by: Measure voltage at the clamp valve pilot.

    PK-3 will not clamp at all

Command versus response

2 machines · 3 scenarios

Separate what the controller asked for from what the field device actually did. A command that was issued is not a movement that happened.

Healthy-reference comparison

2 machines · 2 scenarios

Compare the affected branch against an equivalent one that is working, and use the difference to localise the fault.

Common-mode versus local fault

1 machine · 2 scenarios

Decide whether one branch is affected or all of them, and read that as the difference between a local fault and a fault in what feeds them all.

Motion versus feedback indication

1 machine · 2 scenarios

Separate an actuator that did not move from a position confirm that did not arrive. The controller reports both identically.

  • PK-3 case packer

    the clamp physically closes while the machine reports it did not — the move happened and the confirm did not

    Performed by: Watch the clamp cylinder through a cycle.

    PK-3 stops mid-cycle saying the clamp never closed
  • PK-3 case packer

    the clamp does not move at all, so this is an actuation fault and not a confirm fault

    Performed by: Watch the clamp cylinder through a cycle.

    PK-3 will not clamp at all