MarineNovember 14, 20268 min read

Yacht Shore-to-Generator Changeover: Why AV Drops

Why a yacht's Crestron rack and network drop during shore-to-generator changeover, and how UPS sizing and alerts prevent it.

Shore-to-generator (or generator-to-generator) changeover on a yacht is a brief break-before-make interruption: power drops for a short window while the switchboard resynchronizes to the new source, and any AV or network equipment without battery backup loses power and reboots. A properly sized UPS bridges that gap so Crestron processors, switches, and NVX endpoints never see the interruption at all. The fix is not a single big battery — it's matching UPS runtime to the actual length of the changeover, deciding which loads deserve protection, and wiring in remote alerts so a fault is caught before a guest notices.

This matters more on a vessel than in a house because the changeover happens routinely — every shore-power connection, every generator swap, every time the vessel picks up or drops shore power at a marina — and because a rebooting Crestron processor mid-changeover means scenes, lighting states, and touch panels all have to re-initialize before the crew or owner can use them again.

Why does the rack drop power during a changeover at all?

A changeover is a break-before-make transfer: the switchboard has to disconnect from the outgoing source before it connects to the incoming one, so there's a genuine gap with no power on the bus, even on a well-maintained vessel. Equipment that depends on continuous AC — a Crestron processor, a network switch, an NVX encoder — will drop the instant that gap exceeds its internal power-hold time, and most control and network gear has none worth mentioning. A UPS sitting between shore/generator power and the rack is what actually survives the gap, because its battery picks up the load the instant input power disappears. Crestron's own UPS literature describes this directly: the device "constantly monitors the incoming power line and instantly transfers power to battery backup in the event of such a power loss, ensuring the connected components continue to operate unhampered," and separately "will also ensure a smooth changeover from primary power to a backup generator" — language written for exactly this scenario (CEN-UPS1250 product page).

That particular Crestron unit is listed as discontinued, so it isn't something to specify for a new build or refit — but the behavior it describes (transfer time measured in single-digit milliseconds, battery picking up the load automatically) is the standard a replacement UPS on any new installation should be held to. Ask your integrator which current UPS platform they spec and what its measured transfer time is before you accept a design.

How much runtime does a changeover actually need versus a longer outage?

This is the sizing question that trips up a lot of designs: a shore-to-generator changeover is seconds long, while a genuine outage — a tripped generator, a fuel problem, a dead shore cable — can run for minutes or longer. Crestron's own spec sheet for one of its UPS models shows this trade-off concretely: at half load the unit holds roughly 12 minutes of runtime, but at full load that drops to under three minutes, which is the gap between "bridges the changeover with margin to spare" and "barely covers the changeover itself" (CEN-UPS1250 specifications). A smaller, factory-integrated Crestron Green Light UPS shows the same shape at a different scale — about 10 minutes at half load, about 3 minutes at full load (GLA-UPS product page).

The practical takeaway: don't size a UPS to its nameplate capacity and call it done. Size it to actual connected load, leave meaningful headroom below full load, and treat the changeover event and a real outage as two different design problems — the first needs seconds to a couple of minutes of bridging, the second needs either a much bigger battery bank or a managed, graceful shutdown.

What should get load-shedding and a managed shutdown instead of a bigger battery?

Not every piece of gear in the rack deserves the same priority during an extended outage. Crestron's UPS documentation describes "configurable load-shedding" that allows only certain components to be shut down, preserving battery runtime for the devices that matter most, plus sequenced power-up and power-down so equipment doesn't all restart at once and trip breakers or damage components (CEN-UPS1250 product page). On a yacht, that typically means deciding in advance: does the network switch and control processor stay up through a longer outage while displays, amplifiers, and non-essential AV shed first? That's a design conversation to have with your Cave Group integrator — the priority order belongs in the commissioning documentation, not improvised during an actual outage.

A managed shutdown matters just as much as a managed restart. A control processor that loses power mid-write to its configuration, or a network switch that reboots into an inconsistent state, can take longer to recover than the outage itself lasted. Sequenced, delayed-start outlet banks — a feature built into rack UPS hardware for exactly this reason — give the equipment time to come up in the right order rather than all at once.

Why does a yacht's power frequency complicate this more than a house would?

A house sits on one grid frequency for its entire life. A yacht does not: it may run 60Hz on its own generators but tie into 50Hz shore power in parts of the Mediterranean, or the reverse depending on build origin and cruising grounds. Fixed-frequency equipment that assumes 60Hz can behave unpredictably on a 50Hz shore tie unless it's been specified or configured for dual-frequency operation. This is a detail worth raising with your marine integrator during design, not discovered at the dock — ask specifically which rack components are frequency-tolerant and which aren't, and whether the UPS itself regulates output frequency independent of what's coming in from shore or generator.

How do you catch a changeover fault before anyone on board notices?

Remote monitoring is what turns "the system rebooted and nobody knew until the guest complained" into "the integrator saw the alert and pushed a fix before anyone noticed." Crestron's XiO Cloud platform is built to send alerts by email or text when it detects an issue with a device or a room, with configurable severity levels so the right person gets notified for the right kind of fault (Alerts | XiO Cloud Manual). On the networking side, Peplink's InControl platform works the same way for WAN and router health — it aggregates device data into reports, lets a technician remotely access any device's web admin without a site visit, and supports email notifications for important network events (Peplink InControl 2). Together, those two alerting layers cover the control/AV side and the network side of a changeover-related fault.

For a new build or refit, confirm during commissioning that both of these are actually wired up — not just installed, but configured with real alert recipients and sensible severity thresholds. A UPS that silently switched to battery and recovered without anyone noticing is a near-miss worth knowing about, and that information only surfaces if alerting is turned on.

What should get tested after a refit or new build, before the vessel leaves the yard?

A changeover-survival test belongs on every commissioning checklist, not just a generic power-on check. At minimum: force a shore-to-generator changeover (or simulate one) with the full AV/network load connected, and confirm the rack stays up through the gap without any processor or switch rebooting. Separately, pull shore power entirely and time how long the UPS actually bridges at the real connected load — not the nameplate figure — and confirm load-shedding and sequenced restart behave as programmed if that runtime is exceeded. Finally, verify XiO Cloud and Peplink InControl alerts actually fire and reach the right person during that test, since a system that behaves correctly but never tells anyone is only half-solved. Cave Group designs and commissions marine Crestron and network systems around exactly these tests as part of a refit or new-build scope; see the marine page for how that fits into a broader vessel technology package, and the related yacht refit technology playbook for how this kind of acceptance testing fits into a full yard-period scope.

Frequently asked questions

Why does my yacht's AV rack reboot every time shore power or generators change over?

Because changeover is a break-before-make transfer with a genuine gap in available power, and equipment without battery backup loses power the instant that gap exceeds its internal hold time — which for most control processors and network switches is close to zero. A correctly sized UPS placed ahead of the rack bridges that gap so the equipment never sees the interruption.

How long does a UPS need to run to survive a changeover versus a real outage?

A changeover is typically seconds long, so a UPS sized with reasonable headroom below its full-load capacity easily bridges it; a genuine outage can run minutes or longer, and published runtime figures for rack UPS hardware show runtime dropping sharply between half load and full load, so sizing to actual connected load rather than nameplate capacity matters.

Should every piece of AV and network gear in the rack get UPS protection?

Not necessarily at the same priority. Load-shedding features built into rack UPS hardware let you keep essential gear — typically the network switch and control processor — running longer while shedding displays, amplifiers, or other non-essential loads first; the priority order should be agreed with your integrator during design, not improvised during an outage.

Does 50Hz shore power versus 60Hz generator power cause problems for AV equipment?

It can, for equipment that assumes a fixed frequency rather than tolerating both; this is a detail to confirm with your integrator for any vessel that cruises between regions with different shore-power standards, since the answer depends on the specific hardware specified for your rack.

How do I know if a changeover fault happened if no one was on board to see it?

Remote alerting is the answer: platforms like Crestron XiO Cloud for the control system and Peplink InControl for the network send email or text alerts when they detect an issue, so an integrator can be notified and often fix the problem remotely before anyone on board even registers it. Confirm these alerts are configured with real recipients during commissioning, not just installed and left at defaults.

Sources

Work With Cave Group

Planning a yacht project?

Cave Group designs, installs, and supports these systems for luxury homes, hotels, and yachts across New York, New Jersey, and Europe — one partner from design through 24/7 support.

Start a Project

or explore our work