When a Mach3 plasma-cutting machine loses both the keyboard and mouse as soon as the driver and PLC cabinet is energized, treat the event as an electrical-interference or shared-power-path problem until testing proves otherwise. In the reported case, replacing motherboards and ports did not help; USB and PS/2 devices behaved alike, the wireless joystick remained usable, and the fault persisted with the LPT connections removed and the computer moved 3 m from the cabinet.
Separate the Confirmed Symptoms from the Hypotheses
| Observation | Engineering implication |
|---|---|
| Keyboard and mouse work before the cabinet is energized | Cabinet energization is the repeatable trigger. |
| USB and PS/2 devices both fail | The evidence does not support blaming only the USB protocol or one connector. |
| Changing motherboards and ports does not correct the fault | A single failed computer port or motherboard is less likely. |
| The wireless joystick continues to operate | Determine whether only the input devices fail or whether Mach3 and the entire computer stop responding. |
| The condition remains with LPT connections disconnected | The LPT interface is not required for the observed failure. |
| Moving the computer 3 m away does not help | Distance alone does not remove the disturbance; investigate power and common-reference paths as well as radiated coupling. |
Identify Which Switching Event Causes the Dropout
Do not group cabinet power, motion, and plasma ignition into one event. Similar reports differed: one fault appeared when the driver cabinet was switched on, while another appeared only during arc ignition combined with stepper operation. Isolate these operating states before changing wiring.
- Start Mach3 with the driver and PLC cabinet off and confirm that the keyboard and mouse respond.
- Energize the cabinet without commanding motion or plasma ignition. Record whether either device disappears immediately.
- If the inputs remain available, command motion without plasma ignition and repeat the check.
- If motion remains stable, test the torch-control output and then the actual ignition event under the machine's approved operating procedure. Record the exact transition that causes the failure.
- After a failure, determine whether reconnecting a device causes the computer to enumerate it again, whether a reboot is required, or whether the entire computer has stopped responding.
If the keyboard still communicates but Mach3 will not jog, check the on-screen Jog On/Off state. That control-state issue does not explain an operating system losing both input devices, but it prevents a Mach3 configuration problem from being mistaken for electrical interference.
Distinguish Conducted Power Disturbance from Interface Coupling
- For a diagnostic test, power the computer from a suitable UPS, disconnect the UPS input from the wall supply, and reproduce the cabinet switching event. Follow the UPS ratings and safety instructions.
- Disconnect nonessential signal links between the computer and machine, then repeat the test. The original fault persisted without LPT connections, so also inventory every remaining electrical reference, including auxiliary power, shields, communication cables, and peripheral grounds.
- If operation becomes stable on isolated UPS power, investigate the computer supply path and cabinet-generated conducted disturbance. If it still fails with machine interfaces removed, investigate coupling through the remaining cables and the local electromagnetic environment.
- If the failure requires a connected control interface, reconnect one interface at a time until the symptom returns. The last restored connection identifies a path to inspect; it does not by itself prove which component is defective.
Do not remove protective earth as a troubleshooting shortcut. Protective bonding is a shock-protection measure, and changing it can create a hazardous machine even if the symptom changes.
Inspect Shared Auxiliary Power and Shield Paths
One reported installation powered an optically isolated interface from the computer power supply. During plasma ignition, a meter connected to the nominal 12 V THC supply moved beyond its 0-20 V measurement range. That result does not establish the waveform or its peak value, but it is sufficient to reject the reading as a verified 12 V supply condition and to investigate the shared supply path.
- Document whether the PLC interface, THC, relay interface, or optical-isolation circuitry receives power from the computer supply.
- If machine-side circuitry shares that supply, test it with a separate appropriately specified power source so the computer is not part of the machine-side return path. A 24 V supply was proposed in the evidence, but use 24 V only if every connected device is rated for it.
- Inspect shields, connector shells, cabinet bonds, and cable terminations for open or intermittent continuity. Perform any movement-assisted continuity check only with the equipment de-energized and under the site's electrical safety procedure.
- Do not assume that adding another ground conductor, changing supply phases, using Bluetooth peripherals, or substituting metal conduit will solve the fault. These changes were reported without consistent success and do not identify the coupling path.
Verify the Correction Under Each Operating State
A repair is not verified by a single successful startup. Repeat the sequence that originally produced the failure: cabinet energization, motion, torch-control operation, and actual arc ignition where applicable. Confirm that the keyboard, mouse, wireless controller, Mach3 interface, and computer remain responsive after each transition.
Record the final wiring arrangement, power-source separation, shield termination, and the trigger that previously caused the dropout. If the symptom remains, preserve the isolated test results and compare them as a decision tree: failure on battery-backed computer power points away from the wall supply alone; failure with all machine interfaces removed points away from those interfaces; failure only after one link is restored identifies that link as the next inspection boundary.
FAQ
Why do my Mach3 keyboard and mouse stop when the control cabinet turns on?
The cabinet transition is introducing a disturbance through an unidentified path. Because USB and PS/2 devices both failed and the symptom remained with LPT disconnected, test computer power, shared auxiliary supplies, return paths, shields, and remaining interfaces instead of replacing another port.
Will moving the Mach3 computer farther from the plasma cutter fix input dropouts?
Not necessarily. The reported fault remained after the computer was moved 3 m from the cabinet, so distance did not eliminate the relevant path; use isolated-power and interface-disconnection tests to distinguish conducted from coupled interference.
Can I power the THC or opto-isolated interface from the computer supply?
Treat that arrangement as suspect when ignition or cabinet switching affects the computer. One nominal 12 V THC circuit drove a meter beyond its 0-20 V range during ignition; test a separate correctly rated supply, but do not apply the proposed 24 V unless every connected device accepts it.
