Every programmable controller is designed to survive a blink of the mains. IEC 61131-2 sets the equipment requirements and tests for programmable controllers, including a short internal hold-up that lets the CPU ride through a sub-cycle interruption without resetting. That built-in margin is measured in milliseconds; it is not enough to finish a controlled stop, flush a data log or warn a SCADA master that the line has gone down. A small nickel-metal hydride backup module on the 24 V control rail is what converts those milliseconds into the seconds needed for an orderly shutdown.
IEC 61131-2 defines the operating and immunity requirements for programmable controllers, including the DC power port. A 24 V DC controller is specified to work across a wide rail - commonly about 19.2 to 28.8 V - and to tolerate brief supply interruptions. Reference designs cite a hold-up time on the order of ten milliseconds at full load, and the standard's interruption classes distinguish very short dips (for example a PS2 class that rides an interruption under 10 ms with more than a second between events) from longer losses.
Real controllers match this: a Mitsubishi MELSEC iQ-F unit, for example, documents continued operation through an instantaneous power failure of 5 ms or less, and industrial power supplies and motor controllers describe a 'faultless ride-through' around 10 ms. Beyond that window the internal bulk capacitor is exhausted and the rail collapses.
The first animated figure expands the critical first second after mains loss. At t0 the AC input disappears; the power supply's bulk capacitor holds the 24 V rail for its internal hold-up - the few milliseconds the PLC handles alone. Without added storage the rail then sags through the controller's undervoltage threshold and the CPU resets ungracefully, losing scan state and any data not yet committed.
With a NiMH backup module, a rail-monitor circuit detects the sag while the internal capacitor still holds, connects the nickel pack through a blocking path and DC-DC stage, and keeps the 24 V rail alive for seconds. In that window the PLC finishes its current scan, writes retentive registers and the data log to non-volatile memory, executes a defined stop sequence, and transmits a last-gasp frame over the fieldbus.
Three distinct needs are often conflated. Hold-up is the sub-second survival of a brief dip that needs no software action - the capacitor's job. Orderly shutdown needs seconds of regulated power to save state and park actuators safely - the NiMH module's job. Bridging needs minutes of power to keep running through a transfer to a generator or a UPS - a larger battery's job. Confusing them leads either to a capacitor-only design that resets on a 100 ms loss, or to an oversized, expensive battery where a small module would do.
The second figure compares the time domains: the internal electrolytic covers milliseconds, a supercapacitor bank comfortably covers seconds of high-power but little energy, and a NiMH module covers tens of seconds to minutes at the modest current of a CPU, its I/O and a communication module - the sweet spot for graceful shutdown and last-gasp reporting.
Data-acquisition and remote-terminal applications add a communication requirement. When mains is lost, an RTU or DAQ gateway should timestamp the event, store the last valid sample and transmit a power-failure message so the SCADA master distinguishes a real outage from a broken radio link. AMI meter networks use the same idea under the name 'last gasp', and the energy needed is the transmit pulse current multiplied by the time to associate and send - typically a few hundred milliseconds at a few hundred milliamperes, repeated for retries.
A NiMH module supplies that pulse with energy to spare for retries, and unlike a supercapacitor it also carries the controller through a long enough window to log context, close files and place outputs in a safe state.
The 24 V control rail is the natural attachment point: a string of twenty NiMH cells sits near the rail voltage, charges from it during normal operation, and is buffered by a small DC-DC converter. Sealed NiMH offers a flat plateau, low internal resistance for the transmit pulse, an aqueous chemistry with no thermal-runaway risk in a sealed cabinet, and far better energy per volume than a supercapacitor when the requirement is seconds-to-minutes rather than milliseconds.
Kept at a controlled float or top-up charge and discharged only on real outages, the module sees an easy duty - long standby with rare, shallow events - where low self-discharge and calendar stability dominate over raw cycle count.
The design brief therefore starts from four measurements: the 24 V rail current of the CPU, I/O and communication modules; the time the application needs to save state and send its last-gasp; the fieldbus transmit pulse current and retry count; and the cabinet temperature range. Paper B turns those into a chemistry and Ah calculation; Paper C maps the module onto the IEC 61131-2 and IEC 61000-4-11 type tests and the cell-safety evidence.
Weijiang Power manufactures sealed nickel-metal hydride cells and compact 24 V backup modules that extend PLC and data-acquisition ride-through from milliseconds to an orderly shutdown. Send us your controller rail current, the hold-up or last-gasp time you need, the fieldbus protocol and cabinet temperature, and our engineers will design a welded NiMH module with charge management, blocking and protection matched to the 24 V rail. See module formats on the products page.
