A semiconductor fab loses a batch of 300 wafers — roughly $150,000 — to a voltage sag that lasts 200 milliseconds. The sag drops two phases to 62% of nominal, just long enough for the RF generators on the etch tools to trip on undervoltage. The facility has a UPS on every critical tool. But the UPS did not catch this sag — its transfer threshold was set at 55% to avoid nuisance switching onto battery for sags the tool power supplies could ride through. The UPS did exactly what it was configured to do: ignore sags above the transfer threshold. The problem is that 92% of utility voltage disturbances are sags — not outages — and most are shallow enough that a battery-based UPS lets them pass. A UPS protects against outages. A DySC protects against the 92% of events a UPS ignores.
What a voltage sag actually looks like on a factory floor
A voltage sag (IEC term: voltage dip) is a short-duration reduction in RMS voltage to between 10% and 90% of nominal, lasting from half a cycle (8.3 ms at 60 Hz) to one minute. The most common cause is not a lightning strike — it is a tree branch contacting an overhead distribution line on a windy day, a utility recloser operating to clear a temporary fault on a neighboring feeder, or a large motor starting across the street. The utility clears the fault in 100–300 ms and the voltage recovers. In that 100–300 ms, a VFD with a 15 ms ride-through capacitor bank trips on undervoltage, a robot controller browns out and loses position, and a CNC spindle drive faults. The lights never flickered — the sag was too short for the human eye to register — but the production line stopped.
How a DySC works: stored-energy-free sag correction
A DySC (Dynamic Sag Corrector) is fundamentally different from a UPS. A UPS stores energy — in a battery, flywheel, or supercapacitor — and converts it to AC through an inverter when the input fails. A DySC stores no energy. Instead, it uses a high-speed solid-state switch (IGBT-based) that opens within half a cycle when the input voltage sags, simultaneously injecting the missing voltage through a series transformer fed from the remaining healthy phases or from the line itself via a parallel converter. The load sees a corrected sine wave within 2 ms of the sag onset. When the sag ends, the switch closes and the DySC returns to bypass mode. There is no battery to replace every 3–5 years, no thermal management for a battery room, no inverter running continuously, and — critically — no transfer threshold to configure. The DySC corrects every sag, regardless of depth (down to 50% of nominal on a single-phase DySC), up to its rated correction duration (typically 1–5 seconds). After 5 seconds, if the sag persists, it is an outage — and a DySC passes the outage to the load, tripping the process. For outage protection, you still need a UPS.
DySC vs UPS: the comparison that matters for production equipment
| Factor | DySC | Battery UPS |
|---|---|---|
| Protection target | Voltage sags (92% of utility events) | Outages (8% of utility events) |
| Energy storage | None — corrects from line | Battery bank, flywheel, or supercapacitor |
| Response time | < 2 ms (half-cycle detection + injection) | 2–10 ms (transfer switch + inverter ramp) |
| Runtime at full load | 1–5 seconds (sag correction duration) | 5–60 minutes (battery capacity) |
| Maintenance item | None (no consumables) | Battery replacement every 3–5 years |
| Footprint (100 kW unit) | ~0.3 m² (wall-mount or floor cabinet) | ~1.5–2.5 m² including battery cabinets |
| Efficiency in normal mode | >99% (bypass: switch closed, no conversion) | 92–96% (double-conversion: rectifier + inverter always active) |
When a DySC is the better answer than a UPS
A DySC makes more sense than a UPS when the dominant power quality problem is sags, not outages. This describes most urban and suburban industrial facilities served by underground distribution — outages are rare (one every 2–5 years), but sags happen 10–50 times per year depending on the feeder's exposure to trees, weather, and neighboring loads. A DySC also wins when the protected load does not need outage ride-through — the machine can coast down safely during an outage, but must not trip during a sag. Semiconductor etch and deposition tools, automotive welding controls, injection molding machines, and CNC machining centers typically fall into this category. A third scenario: when battery maintenance is the bottleneck. A large UPS battery bank in a hot factory environment — especially one without air conditioning in the electrical room — sees its battery life cut from 5 years to 2–3 years. The replacement cost and downtime for battery swaps can exceed the UPS purchase price over a 10-year period. A DySC has no battery — that operational cost line item is zero.
When a UPS remains the only choice
The UPS is mandatory when: the process cannot tolerate an outage — a data center, a hospital operating room, a continuous chemical process where coast-down creates a hazardous condition; the sag depth exceeds what a DySC can correct (typically below 50% of nominal for single-phase DySC, or below 55% for three-phase); or the sag duration exceeds 5 seconds (DySC correction limit) — rare on utility distribution but common on weak generator-backed microgrids where a large motor start pulls down the voltage for 10–15 seconds. A UPS also handles the transition from sag to outage — a DySC corrects the sag, but if the sag deepens into an outage, the load drops. A double-conversion UPS rides through both.
For facilities that need both sag correction and outage ride-through, the combination architecture — a DySC on the upstream feeder correcting sags for the entire bus, and a smaller UPS downstream on the critical loads that need outage protection — is more cost-effective than upsizing the UPS to cover every load for outage duration. The DySC eliminates 92% of the events; the downstream UPS handles the remaining 8%.
Integration with industrial power supplies and machine-level protection
The DySC discussion fits into a broader power quality strategy. At the machine level, a DIN-rail power supply with a 20 ms hold-up time covers the shortest sags — the ones below 1 cycle. A DySC at the sub-panel level covers sags from 1 cycle to 5 seconds. A UPS at the critical load covers outages beyond 5 seconds. Three layers, each protecting against a different class of event, with no overlap and no gap. For a machine builder, specifying the right combination means the machine survives the power quality at the customer's site — wherever that site happens to be.
If your facility experiences more than 5 voltage sags per year that stop production, a DySC pays for itself faster than a UPS — it corrects the exact events that are stopping your machines, without the battery replacement cycle that inflates UPS total cost of ownership.
What is the difference between a DySC and a constant-voltage transformer (CVT)?
A CVT (ferroresonant transformer) regulates voltage using a resonant LC circuit and magnetic saturation of the transformer core. It corrects steady-state voltage variations — a feeder that runs 5–10% high or low continuously — but responds too slowly (5–30 cycles) to catch a sub-cycle sag. A CVT also runs at 75–85% efficiency and generates significant heat. A DySC responds in under 2 ms and runs at >99% efficiency in bypass. They address different problems: CVT for chronic voltage regulation, DySC for transient sag events.
Can a DySC protect an entire production line or just individual machines?
A three-phase DySC can be sized from roughly 25 kVA to 2 MVA, making it suitable for individual large machines (a 500-ton injection molding press), a sub-panel feeding a row of CNC machines, or an entire production line's control power bus. The sizing criterion is the total kVA of the loads that must ride through a sag — not the total connected load. Motors that can coast through a sag without tripping (pumps, fans, conveyors with mechanical inertia) do not need to be on the DySC. Control power, servo drives, PLCs, robot controllers, and process-critical heaters do. Separating these loads at the sub-panel level is typically more cost-effective than sizing one large DySC for the entire facility.
