The wrong plan — why you cannot fool an ATS into selecting your car

The plan most people start with

The popular idea is neat on paper: buy a cheap 63 A automatic transfer switch (ATS), add a Wi-Fi smart relay, and use the relay to 'fool' the ATS into selecting the car on a timer — charge cheap overnight, then run a small board off the EV during expensive hours. It does not work, and it is worth understanding exactly why before you spend a penny. The short version: an ATS is the wrong *kind* of device for scheduled switching, the smart relay is far too small to switch the power, and forcing it would create two hazards at once.

Why it fails — three separate reasons

1. The ATS decides for itself — there is nothing to command

A cheap voltage-sensing ATS is self-sensing. It watches its normal (priority) input, transfers to the alternate source in a couple of seconds when that input *fails*, and auto-reverts when it returns. That behaviour is built around source loss, not a schedule. The status terminals on these units are indicators — they tell you which side is live — not a control input you can drive. There is simply no 'switch now' command to send. This is the behaviour the relevant product standard describes for transfer switching equipment.

Most of the affordable single-box ATS units on the market make this worse, not better. Their documented logic keeps normal-source priority in automatic mode — the grid wins whenever it is healthy — so even a unit that *does* expose a remote input will not freely 'pick grid or EV whenever I say so' in its ordinary auto behaviour. They are designed for backup, not for arbitrary scheduled source selection.

2. The smart relay is far too small to switch the power

A Wi-Fi smart relay (the kind people reach for) is a roughly 2 A dry contact intended for control-circuit signalling. Every power path in this arrangement is around 63 A — the consumer-unit scale. Asking that contact to switch or interrupt a 63 A supply is orders of magnitude beyond its rating; it would not make or break the load safely, and would likely weld or burn. A relay like this is fine for one job only: commanding a properly rated device, never carrying the load itself.

3. Even if you forced it — overload and no earth reference

Suppose you somehow bridged the switching. You would now connect a 63 A consumer unit to an EV's V2L output of roughly 10–15 A (around 2.3–3.6 kW at 230 V). That is a gross overload — the whole house demanding far more than the car can deliver. Worse, an EV's V2L output is commonly floating: there is no internal neutral-earth bond, so an RCD has nothing to operate against. The output would pass a plug-in socket tester yet offer no real shock protection. You would have islanded onto an over-stressed, unreferenced supply.

The correct device — command, don't fool

Scheduled switching between two *healthy* sources is a different job from an auto-ATS's source-loss reaction. The correct device is a timed relay commanding an interlocked two-pole contactor changeover: KM1 the grid contactor (the de-energised, fail-safe default) and KM2 the V2L contactor. The relay carries only the small coil current; the contactors carry the load. A mechanical interlock is the hard guarantee that the two can never close together — break-before-make, so grid and vehicle are never paralleled (an EV inverter cannot parallel the grid in any case).

Keeping grid and vehicle physically separate at all times is not just good practice — it is what keeps a domestic V2L island out of the grid-parallel regime, where a DNO and additional protection would come into play. A genuine interlocked, break-before-make changeover is the proof that the two sources can never bridge.

That is the whole picture on one sheet: the scheduled control, the interlocked two-pole power changeover, and — because the V2L output is floating — the earthing that gives the island a single neutral-earth reference. The full design (power, control and earthing) and the floating-output earthing decision are covered on the cornerstone and changeover pages; this page exists only to retire the ATS-fooling idea before you waste money on it.