A standard automatic transfer switch (ATS) does one thing reliably: it opens one power source and closes another,in that order,with a deliberate dead time between the two operations. That dead time — typically 50–300 milliseconds depending on the switch mechanism — is a feature,not a bug. It prevents the two sources from ever being connected simultaneously,which would back-feed the utility grid and create a severe safety hazard for line workers. But those 50–300 milliseconds also reboot every microprocessor-controlled device downstream. For a wastewater lift station running VFD-driven pumps,a 200 ms interruption means the VFD DC bus collapses,the pump coasts down,and the restart sequence takes 30–90 seconds after power returns — during which the wet well continues to fill.
A no-break changeover — also called a closed-transition transfer or synchronized transfer — eliminates the dead time by briefly paralleling the generator with the mains. The ATS controller monitors the voltage,frequency,and phase angle of both sources in real time. When the generator is running and stable,the controller waits for the two waveforms to align within a tight synchronization window (typically ±5° phase angle,±0.2 Hz frequency,±5% voltage). When they align,it closes the generator breaker while the mains breaker is still closed,creating a momentary parallel condition. Once the generator is carrying the load,the mains breaker opens. The entire sequence completes without a single cycle of interruption to the downstream bus.
The Synchronization Window: What the Controller Must Verify Before Paralleling
Closing two unsynchronized AC sources onto the same bus is not a nuisance trip — it is an arc-flash event. If a 480 V generator is 180° out of phase with the 480 V mains when the breaker closes,the instantaneous voltage across the breaker contacts is 1,356 V peak,and the resulting current surge can exceed 50 kA for the first half-cycle. To prevent this,the ATS controller — or more accurately,the synchronizing controller — verifies four conditions before issuing a close command:
Phase angle difference must be within ±5° — at 60 Hz,one cycle is 360° in 16.7 ms,so 5° represents 230 µs. The controller predicts the zero-crossing time for both waveforms and issues the close command such that the breaker contacts mechanically meet when the two waveforms are within 5° of each other. Frequency difference must be within ±0.2 Hz — a larger frequency difference means the phase angle is constantly drifting,and the breaker close time prediction becomes unreliable. Voltage difference must be within ±5% — a larger voltage difference drives reactive power flow between the two sources during the parallel period. And the parallel time must be limited — typically to 100 ms or less — to comply with utility interconnection requirements,which prohibit sustained parallel operation without a dedicated protection relay and utility approval.
The LOVATO Electric ATL Series automatic transfer switch controller handles 2–3 power sources (two mains plus a generator,or two generators plus a mains) with an expandable I/O architecture. Its built-in synchronoscope function monitors voltage,frequency,and phase angle on all sources simultaneously,enabling closed-transition transfer with programmable sync window parameters. For applications that do not need synchronization on day one but may in the future,the ATL Series supports both open-transition (break-before-make) and closed-transition (make-before-break) modes in the same hardware — the transfer mode is configurable in software.
Where No-Break Changeover is Mandatory
Data centers with Tier III or Tier IV uptime certification require no-break transfer. The Uptime Institute's Tier III specification mandates concurrently maintainable power paths,which means the transfer between utility and generator cannot interrupt the IT load. Hospitals with life-safety critical circuits — operating room isolation transformers,ICU ventilators,and anesthesia machines — require no-break transfer per NFPA 110 Class X,which defines a maximum interruption of 0.004 seconds (one-quarter cycle at 60 Hz). Semiconductor fabs with continuous Czochralski crystal-growing furnaces cannot tolerate even a half-second power interruption — a single crystal growth run takes 18–36 hours,and a power dip mid-run scraps the ingot at a cost of $50,000–500,000 depending on wafer diameter.
Where Break-Before-Make is Sufficient
Most commercial and light industrial applications tolerate a 50–300 ms interruption without consequence. Office buildings,retail facilities,and warehouses with LED lighting and desktop computers ride through the dead time on the stored energy in power supply capacitors — an ATX power supply holds up for 16–20 ms,which is shorter than the ATS dead time,but a UPS bridge fills the gap. Wastewater pump stations,irrigation pumps,and HVAC chiller plants with across-the-line starters tolerate a brief interruption because the motors restart automatically when power returns. In these applications,the added cost of a synchronizing ATS — typically 2–3× the cost of a standard open-transition ATS for the same current rating — is not justified by the consequence of the interruption.
How long does the generator need to be stable before a synchronized transfer?
The generator must reach rated voltage and frequency — typically 10–15 seconds after start — and then stabilize for an additional 5–10 seconds before the synchronizer attempts to match phase. The total time from generator start signal to closed-transition transfer completion is typically 15–30 seconds. An uninterruptible power supply (UPS) must bridge the entire 15–30 second window,which is why facilities requiring no-break power always pair a synchronized ATS with a UPS system. The UPS handles the generator start-up window; the synchronized ATS handles the transfer back to mains without a second UPS-draining interruption. For power quality solutions upstream of the ATS,see our power meter and electrical power product ranges.
What prevents back-feeding the utility during the parallel window?
Three independent layers of protection. First,the synchronizing controller verifies the phase,voltage,and frequency alignment before issuing the close command — if any parameter is outside the programmed window,the close command is blocked. Second,a protective relay (typically ANSI device 27/59 for under/over-voltage and 81O/U for over/under-frequency) provides a hardware-level interlock independent of the controller software. Third,the utility interconnection agreement specifies a maximum parallel duration — usually 100 ms — after which the mains breaker must open regardless of whether the generator is carrying the load. The ATS controller enforces this timeout in hardware firmware. These three layers together satisfy both the safety requirement (no back-feed) and the reliability requirement (no single point of failure can cause a sustained parallel condition).
For more on transfer switch control and power distribution,browse our PLC catalog for controllers that integrate ATS logic with broader facility automation.
