A Mach3-controlled G0704 produced intermittent limit-switch errors while running a test program, even though no axis was near an overtravel switch. The spindle was stopped and no workpiece was installed, so cutting load and spindle vibration were not required for the fault to occur. Treat this as a false input transition until the switch circuit, wiring, input assignment, and breakout-board interface are isolated.
Separate Observations from Conclusions
| Observation | Engineering conclusion |
|---|---|
| No overtravel indicator illuminated when the error appeared | The diagnostic display did not identify a sustained switch input; a brief transition could still have occurred between screen updates. |
| Output #1 blinked during operation | No evidence connects this output activity to the limit error. Do not treat it as the cause without correlating it to the fault. |
| The fault occurred with the spindle stopped | Spindle operation is not necessary to reproduce the problem. |
| The error stopped when the Z-circuit wires were removed from the C-10 board | The fault was localized to the connected Z input path or its interaction with the board, rather than proving that the board itself had failed. |
Isolate the Intermittent Input
- Run the same test program under controlled conditions and observe the limit-input diagnostics.
- Disable or disconnect one limit-switch circuit at a time. Repeat the test after each change until the error stops.
- After identifying the Z circuit, inspect the complete path: switch actuation, terminals, conductor continuity, and movement-sensitive sections of cable.
- Replace the suspect solid-conductor bell wire with stranded wire and replace the switch if its contacts are suspect.
- Reconnect the circuit and repeat the original program long enough to determine whether the intermittent fault remains.
Solid wire can develop an intermittent break when machine movement repeatedly flexes it. A diagnostic screen may miss a short interruption, so failure to see a limit indicator does not clear the field circuit.
Interpret the Input-Port Test Correctly
The Z circuit originally used port 12 and produced the same error when moved to ports 10 and 11. Because the behavior followed the circuit across three ports, the test does not uniquely prove that one input channel is defective. A shared board problem remains possible, but so do a remaining field-wiring fault, a common/reference connection problem, or the way the Mach3 input is configured.
Disconnecting the field wires from the C-10 and observing that the error disappears is useful localization evidence. Before replacing the breakout board, test the disconnected switch and cable independently while flexing the cable and operating the switch. Then verify that the Mach3 input assignment and active-state interpretation match the actual switch circuit.
Verify the Repair
Restore the Z limit circuit, run the same proving program with the spindle stopped, and monitor both the limit diagnostics and the physical wiring. The repair is verified only if the original operating sequence completes repeatedly without a limit error. If the fault returns, record whether it follows the Z field circuit, remains on one board input with the field circuit removed, or changes with cable movement; that distinction determines whether to continue with wiring, switch, configuration, or C-10 board diagnosis.
FAQ
Why does Mach3 report overtravel when no axis is near a limit switch?
An intermittent switch contact or conductor can create a transition too brief to remain visible on the diagnostics screen. In this case, elimination testing localized the behavior to the Z limit circuit.
Does a blinking Mach3 Output #1 cause a random limit error?
The available evidence shows Output #1 blinking but does not establish a causal connection. Correlate the output transition with the error before changing output logic.
Does the same error on C-10 ports 12, 10, and 11 prove the board is bad?
No. Because the error followed the connected Z circuit across ports 12, 10, and 11 and disappeared when its wires were removed, test the switch, stranded cable, common connection, and Mach3 input configuration before condemning the board.
