Elevator control systems: generations, feedback and a troubleshooting order

Most time lost on an elevator fault is lost to random part swapping. A controller is a state machine with a defined set of inputs and outputs; if you test the inputs first, the controller usually tells you what it thinks is wrong.

Four generations you will still meet

Generation Tells What it means for diagnosis
Relay / contactor logic Rows of relays, terminal strips, printed ladder diagrams on the door Every function is observable. Read the print, watch the relay, find the open contact.
Early microprocessor (EPROM era) Seven-segment displays, DIP switches, personality chips Fault codes exist but are coarse; position is often counted, not measured.
Serial / networked controls CAN or RS-485 loops, node addresses, hall fixtures on a bus One bad node or a damaged loop takes out a group. Fault may be reported far from the cause.
Destination dispatch Kiosks, group controller, per-passenger assignment Mechanical faults look like scheduling faults. Verify each car runs standalone before blaming the dispatcher.

Position feedback: the difference between counting and measuring

Older systems count pulses and assume the car is where the count says. Modern systems use an absolute encoder on the machine or a linear tape in the hoistway, so position survives a power cycle. This distinction explains a common misdiagnosis:

  • Incremental encoder with a count: after a rope slip, brake slip or manual rescue, the count is wrong and the car levels badly or “loses” floors. It needs a relearn, not a new drive.
  • Absolute feedback: level is right but speed or direction faults appear – suspect the drive, the encoder coupling, or a parameter set that was never matched to the machine.

A troubleshooting order that avoids board swapping

  1. Power first. Measure all phases under load, not open circuit. A sagging phase produces faults that look like drive failures.
  2. Safety chain continuity. Prove the chain is closed at the controller terminals, then work outward. If it is open, find which device opened before touching configuration.
  3. Door circuit in isolation. Bridging the closed circuit as a test only tells you whether the fault is in the door circuit or further down. Do not leave it bridged and do not bypass an interlock in service.
  4. Feedback. Confirm the drive sees the car moving and in which direction. Wrong direction feedback after work on the machine is nearly always encoder wiring or a swapped pair.
  5. Parameters. Only then look at parameter sets – and compare them against a known-good unit of the same model rather than a generic list.

Documentation that pays for itself

Record, per unit: controller model and firmware, drive model, encoder type and coupling, parameter file backup date, and the last three faults with date, load and position. Units that carry this history are resolved in a fraction of the time, and the history is what makes a fault pattern visible rather than anecdotal.

Safety: fault codes and parameter lists are manufacturer-specific and change between firmware revisions. Always work from the documentation for the exact model and revision in front of you.

Related reading: system anatomy for the mechanical context, and the forum’s fault code threads for unit-specific experience.

Scroll to Top