Machine Controller Faults on Stone Equipment: Alarms and I/O Checks

A bridge saw that will not start, a CNC router that homes halfway and stops, an edge polisher that runs the motor but ignores the foot pedal — on modern stone equipment, the first suspect is rarely the mechanics. It is the control cabinet. Every one of these machines runs on a programmable logic controller (PLC) or a dedicated motion controller paired with a human-machine interface (HMI) screen, a bank of input/output (I/O) modules, and a chain of safety devices that all have to agree before the machine will move. When something will not run, the fastest path to a fix is almost always inside that cabinet, not under the gantry.

This guide covers reading an alarm or fault message correctly, checking the I/O status indicators on the controller rack, verifying control power and fuses, and tracing the emergency-stop (E-stop) chain and safety relays that cause more "machine will not start" calls than any mechanical failure. It also covers the shop-specific headaches of running control cabinets in a wet environment, and when to stop troubleshooting and call the machine builder. Servo drives, encoders and homing faults, motor starters and overload relays, limit and proximity sensors, and phase converters or VFDs each have their own detailed guides; here they are mentioned only where they intersect with the controller itself.

What a Controller Fault Actually Is

The PLC or motion controller reads inputs (sensors, switches, pedals), runs logic, and drives outputs (motor starters, valves, drive enables). The HMI screen is a window into that logic, not the logic itself — it shows whatever fault text the machine builder programmed, and that text is only as useful as the manual that explains it. The same word, "fault," means a different register on every builder's system, so treat the exact code number as the starting clue, not a plain-English guess.

Most builders separate alarms from faults by severity: an alarm is a warning the machine can keep running under (low coolant, a filter due for cleaning), while a fault stops motion and usually needs an acknowledgment once the condition clears. Knowing which tier you are looking at tells you how urgent the problem is before you open a panel.

A machine can also refuse to run with no fault displayed at all. That is a permissive condition, not a fault: the logic is waiting on something it was never told is satisfied — a door interlock, a "ready" signal from the water pump, or an operator selection never made on the HMI. These silent "won't start" conditions eat the most time precisely because nothing on the screen points at a culprit.

Remember the controller only knows what its sensors tell it. A limit switch or encoder that is physically fine but wired to the wrong input, or a VFD not actually receiving the run command the PLC thinks it sent, produces a controller-side symptom with a field-side cause. Those specific failures have their own guides; what belongs here is how to tell, from the cabinet, that one of them is in play before chasing the wrong subsystem.

The Practical Diagnostic Sequence

Work the problem in the same order every time. Read the exact alarm text and code first. Check the I/O status LEDs second. Confirm control power and fuses third. Trace the E-stop and safety-relay chain fourth. Only after those four checks come up clean does it make sense to start pulling covers off motors or drives.

Reading the Alarm or Fault Message First

Write down the exact code and text before touching anything, including any sub-code shown in smaller font below the headline message. Pull up the alarm history screen if the HMI has one and look at what fired immediately before the fault that stopped the machine — the real trigger is often a sensor dropout a few seconds earlier, not the fault that finally halted motion. Cross-reference the code against the machine's electrical manual rather than guessing from memory; code numbering is specific to the builder and sometimes the firmware revision.

If the manual is missing, most builders can email the fault-code table from the serial number. Resist hitting reset more than once or twice without reading the code first — repeated blind resets waste time and can mask an intermittent fault until it becomes a hard failure mid-cut instead of at startup.

Checking I/O Status LEDs on the Rack

Every input and output module on a PLC rack has its own status LED — a second readout independent of whatever the HMI displays. An input LED lit means the controller is physically receiving a signal at that terminal; if the LED is on but the HMI's soft indicator shows off, the fault is in the logic program or terminal mapping, not the field wiring. An input LED that stays dark when it should be lit (door closed, guard in place) points at the field side: the sensor, its wiring, or its power.

Compare the live LED pattern against the machine's wiring diagram, usually stored in a pocket on the cabinet door. A rack-level fault LED (separate from individual point LEDs) usually means a communication problem between modules or the CPU, and behaves very differently from a single bad input — it often means reseating a bus connector rather than troubleshooting one sensor.

Control Power, the 24 VDC Supply and Fuses

Nearly all modern industrial controllers, sensors and relay coils run on 24 volts DC rather than line voltage, because 24 VDC stays below the roughly 30-volt threshold generally treated as the point where DC shock becomes dangerous, while still driving standard sensor and relay loads reliably. That low-voltage rail comes from a dedicated power supply fed by line voltage, protected by its own control-circuit fuse or breaker separate from the main motor disconnect. A machine that is completely dead — no HMI backlight, no status LEDs anywhere — almost always traces back to that control power rather than any single component.

Check the supply's own output reading before assuming the fuse is the problem; a supply that has quietly degraded can show an intact fuse while delivering sagging voltage that causes intermittent faults rather than a clean dead stop. Keep spare fuses of the exact rating on hand and labeled — an undersized substitute nuisance-trips under normal load, and an oversized one defeats the protection it was there to provide.

The E-Stop Chain and Safety Relays

"Machine will not start" is, more often than any other single cause, an open link in the emergency-stop chain. E-stop buttons on commercial stone equipment are typically wired in series through every station — the operator pedestal, the far end of the gantry, any pendant — into a safety relay or safety PLC, and that relay has to see every contact closed and every guard interlock satisfied before it allows the main contactor to energize. One E-stop left latched, a guard door that did not fully close, or a pinched wire in a flexing cable track will hold the whole machine in safe-stop with no further explanation than "not ready."

Safety relays usually carry small indicator LEDs showing which channel is open; walk the chain by physically checking each E-stop and interlock in sequence rather than assuming the nearest one is the culprit. Dual-channel safety circuits are built so a single wire fault cannot be mistaken for a safe condition — a half-working chain will usually refuse to reset at all rather than let the machine run degraded. Treat that refusal as the system doing its job, not a nuisance to bypass.

Advanced and Trade-Specific Troubleshooting

A sensor can be verifiably correct at the terminal — LED lit, a meter confirms voltage present — yet the machine still behaves as if the input is not satisfied. This almost always means the logic program is not looking at that input the way you expect: a latch set by a previous fault and never cleared, an interlock requiring two or three inputs to agree at once rather than the one you are watching, or a timer that holds the permissive false for a fixed window after a reset. This is one of the clearest signals to call the machine builder rather than keep probing, since altering logic without the source program risks a worse, less predictable fault.

Wet stone shops are rough on control cabinets in ways a dry metalworking shop never sees. Slurry mist finds its way into junction boxes through cable glands that were not fully tightened at commissioning, and mineral-laden water plus vibration accelerates corrosion at crimp connections and terminal blocks far faster than in a clean environment. Corrosion buildup on a terminal raises its resistance, which can read as present on a meter under no load but drop out under the higher current draw of an actual switching event — exactly the kind of fault that comes and goes and defies a five-minute inspection. Periodically opening enclosures near wash-down areas and checking glands, drip loops and terminal strips for white or green deposits catches this long before it becomes a Monday-morning no-start.

Telling a mechanical fault from an electrical one quickly saves diagnostic time. If a fault code names a specific axis, try jogging that axis manually at low speed from the HMI's manual/jog screen, with guards in place: a mechanical bind shows up as excess current draw, unusual noise, or the axis refusing to move at all, while a sensor or wiring fault typically lets the axis move freely but never reports the position or completion signal the logic is waiting for. A controller that commands motion and gets silence back from its feedback device is the classic signature covered in the servo, encoder and homing fault guide; rule out mechanical binding first before assuming the electronics are at fault.

Water systems — recirculation pumps, flow switches, low-level float sensors in the settling tank — feed their own inputs into the same controller on most modern bridge saws, and a tripped flow or level fault stops the saw just as surely as an E-stop will, often with a less obvious message buried in a secondary status screen. Confirm the physical condition first (is the tank actually low, is the pump actually running) before assuming the sensor itself has failed.

Maintenance and Working Safely in the Cabinet

Keep a written alarm log for every machine, even a notebook clipped near the HMI, recording the date, the exact code, what fixed it, and how long the machine was down. One fault is a data point; the same fault three times in a month is a pattern pointing at a wearing component, a marginal supply, or an environmental problem long before it causes a production-stopping failure. This log is also the single most useful document to hand a service technician.

Back up the controller's parameters and program on a schedule, not just after a crash. Most PLCs and motion controllers store machine-specific settings — axis limits, calibration offsets, custom fault thresholds — that cannot be reconstructed from memory if a CPU or memory module fails. Store the backup off the machine, both on a labeled USB drive in the shop office and in a cloud folder; a backup sitting only on a laptop that visits the shop once a quarter is not a real backup plan.

Know the line between what a shop electrician should fix and what belongs with the machine builder. Fuses, contactors, relays, sensors, connectors and wiring are normal shop-level repairs. Anything that requires changing the PLC program itself, re-tuning a servo drive's internal parameters, or re-flashing controller firmware should generally go to the builder's support line, both because unauthorized logic changes carry real safety risk and because warranty terms on newer equipment often exclude damage traced to them.

Working inside an open control cabinet means working near energized circuits, governed in U.S. shops by OSHA's 29 CFR 1910.147, the control of hazardous energy. Under that standard, only an authorized employee — someone trained and permitted to lock out or tag out a specific machine for servicing — should open a cabinet to service it, and a qualified person is specifically someone trained and proven proficient in that machine's energy-control procedure. In practice, that means de-energizing and locking out before reaching into a cabinet for anything beyond reading an LED or screen, never relying on the E-stop button alone as a substitute for lockout, and never working alone inside a live cabinet on a call that might need a second set of hands.

Stock a small kit of known-good spares for every controlled machine: the exact fuse ratings used in that cabinet and a spare control relay or two of the same model. A five-minute fuse swap turns into a half-day delay the moment the right fuse has to be ordered and shipped, and on a shop's primary bridge saw or CNC router that delay shows up directly in missed delivery dates.

These same fault-chasing habits apply across the shop's other controlled equipment. The emergency stop and blade guard safety practices that govern the E-stop chain above, and the lockout/tagout and machine guarding procedures that apply before anyone opens a cabinet, are worth reading in full. For faults that trace to a servo axis rather than controller logic, the servo drive, encoder and homing fault guide picks up where this one leaves off, and persistent water-system faults are easier to chase after reviewing the shop's coolant water filtration and recycling setup.