Equipment suddenly stops, screen is clean, no red alarm, no fault code—this is probably the most troublesome "ghost fault" in equipment maintenance. You restart it, and it runs normally again, for an unknown duration—maybe a few hours, maybe a few days—then suddenly occurs again. Production halts, troubleshooting has no clue, not even an alarm, not even a direction to follow. The reason these faults are difficult is precisely because they don't follow the usual rules, and the system's self-diagnosis function simply cannot catch them. Combined with real maintenance cases, the culprits of such "no-alarm shutdowns" are, in nine out of ten cases, hidden in the following three obscure places.
Cause 1: "Contact Resistance" in the Control Loop
This cause is listed first because it is cunning enough and most easily overlooked.
You look at the PLC, the indicator light is green, no problem. But in reality, in the circuit controlling the pull-in of a certain key relay or contactor, there is a hidden "poor contact" issue. For example, oxidation on the relay contact surface, or contact resistance generated after prolonged operation.
A real case: a Siemens 840D system equipment would randomly shut down without any alarm. Maintenance personnel checked the wiring, the power supply, the servo modules, repeatedly struggling for a week. Finally, they discovered the problem was on terminals 13 and 14 of a relay—when measured with a multimeter, the contact resistance was actually over 100 ohms.
What does 100 ohms mean? In a 24V DC control circuit, an extra 100-ohm resistance divides the voltage, causing insufficient voltage in the pull-in coil of the subsequent servo power supply module enable relay, unable to energize properly, and the equipment naturally stops. When you go to measure with a multimeter, the contact may be connected again, the resistance disappears, everything seems "normal". This intermittent characteristic makes it extremely difficult to detect.
Cause 2: "Transient Disconnection" in the PLC Communication Network
Nowadays, most equipment is "networked": PLCs communicate with servo drives and remote IO modules via bus networks like Profinet, Profibus. Once a network transient disconnection occurs, the system immediately triggers a safety stop, but there may be no time to pop up an alarm on the screen, or the alarm flashes briefly, and the diagnostic buffer only leaves a vague message: "DP slave or PN device failure".
What causes "transient disconnection"?
Unreliable physical connections. For example, using a poor-quality RJ45 plug that is slightly undersized and cannot lock in place; when equipment vibrates, it makes and breaks contact intermittently. Replacing it with an industrial-grade plug may solve the problem.
Another is interference. A new high-power device, such as a laser welder, has been added near the equipment, and its start/stop generates electromagnetic interference sufficient to cause the network to flash-break. This interference comes and goes quickly, impossible to catch with a regular multimeter, only detectable with an oscilloscope, or by locating the interference source.
Cause 3: "Transient Voltage Drop" in the 24V DC Power Supply
If the previous two were "signal" issues, this one is an "energy" issue.
PLCs, sensors, relays all require a stable 24V DC power supply to operate. If this power supply has problems, even just a voltage dip or interruption of a few milliseconds, it is enough to stop the equipment.
A very typical case: there were three servo drives in the equipment; after one was replaced from Teco to Panasonic, irregular shutdowns and jerky movements began to occur. They checked the servo, parameters, enable signals—all normal. Finally, they found that the problem was with the 24V switching power supply supplying the enable relay. It had aged, and its power supply capacity declined; when all three servos worked simultaneously, the voltage momentarily dipped, causing a momentary interruption in the relay.
Even more troublesome, the original Teco servo was not sensitive to such short transient interruptions and could ride through; but the newly replaced Panasonic servo had more sensitive detection, stopping immediately upon detecting an interruption, without generating an alarm. This kind of hidden power supply degradation often shows normal static voltage when measured with a multimeter; the problem only reveals itself under load.
Troubleshooting Approach and Suggestions
When encountering this kind of "ghost fault," don't panic, and don't rush to replace parts. Follow this line of thinking to avoid many detours:
First, implement "heartbeat" monitoring. This is the most effective method. In the PLC program, add counting or latching functions to all critical signals that could cause a shutdown, such as emergency stop, safety circuit, enable feedback, communication heartbeat. Whichever signal changes first, that's where the problem lies, clear at a glance.
Second, focus on checking the safety circuit. Many no-alarm shutdowns are actually caused by a momentary break in the safety circuit. Carefully inspect the wiring and contacts of the emergency stop buttons, door switches, safety relay—look for looseness or oxidation.
Third, don't trust static measurements. Using a multimeter to measure normal voltage does not mean it's normal under load; testing continuity normally does not mean there's no contact resistance. For a suspect relay, replace it directly. For a suspect power supply, try replacing with a new switching power supply.
Fourth, pay attention to new changes. Did the fault start after recently replacing a new component? Was a high-power device newly added to the workshop? These new changes are very likely the clues.
These faults are hard to repair because they do not follow preset alarm logic. But once you understand the physical nature behind the fault—contact resistance, signal transient interruption, power voltage drop—even the most "ghostly" problem will eventually reveal itself.