"MultiPlus-II 48/6500 drops AC-OUT when grid comes back, anyone else?

Seen this exact nonsense with my cabin setup — MultiPlus-II 48/5000 (smaller than the 6500 but same headache), Cerbo GX, Fogstar Drift 48V battery, and whenever the grid decides to stagger back after a outage, AC-OUT drops for a good 30-40 seconds like it's having a think about life.

Load on mine is similarly pathetic — maybe 500W of lighting and a NAS box — so it's not like I'm asking it to restart a steel mill.

What I've tried so far:

• Tweaked the AC Input voltage range in VEConfigure — marginal improvement at best
• Adjusted Dynamic ESS assistant settings — still sulks on reconnection
• Firmware update to latest (3.14 on the Cerbo) — nope

My suspicion is it's a transfer switch delay being overly cautious about grid quality on reconnection, but I genuinely cannot pin it down. The VRM logs show no fault codes, which is almost more annoying than an actual error.

Anyone running an ESS setup in the UK hit this? Particularly curious whether it's worse on certain DNO areas — my supply quality here is frankly embarrassing and I wonder if the MultiPlus is seeing the returning grid as dodgy and doing an extended quality-check before handing over.

Also — is this a PowerAssist vs pure ESS assistant conflict? Running ESS assistant only currently.

Victron's documentation is, as ever, 40 pages long and answers approximately nothing.

WattAMess27 | 147 posts | Derbyshire

@MultiPlusNerd had almost identical grief with my 48/5000 last winter — turned out my grid code setting was wrong for the UK. Make sure you're on G98 or G99 as appropriate, because if VE Configure has it set to something else (I had mine on a generic European setting somehow), the reconnection timing after grid restoration goes completely haywire.

Also worth checking your AC input current limit — if it's set too low, the unit sometimes throws a wobble during that handover moment rather than smoothly transferring back.

Finally, what firmware are you running on the Cerbo? There was a known quirk around DVCC interaction that caused odd AC-out dropout behaviour on reconnect. Updating sorted mine completely. Check the Victron community forum for the specific changelog, it'll tell you exactly which version addressed it.

Check your AC-input current limit in VEConfigure — if it's set too low, the MultiPlus gets confused about whether the grid's actually stable enough to bother with and just sulks on inverter mode instead. Also worth looking at Grid code settings; wrong country profile causes exactly this nonsense because it's waiting for voltage/frequency to stabilise within tolerances that don't match what your DNO actually delivers post-outage. My 48/5000 in the static did this until I bumped the transfer switch sensitivity down a notch — ESS Assistant settings, Weak AC input option. Cerbo firmware also matters; anything pre-3.10 had known transfer bugs. Update everything via VRM and stop assuming Victron shipped it configured correctly, because they absolutely didn't.

Worth checking your transfer switch delay setting in VEConfigure too — there's a param for how long the MP-II waits before accepting grid input back. Mine was set to something daft like 0ms and it was causing all sorts of weirdness with my garden office loads dropping.

Also, what does your ESS assistant config look like? I had a phase where mine was set to "Keep batteries charged" and the handover logic just fell apart whenever grid quality was dodgy. Switching to "Optimised (with BatteryLife)" sorted it for me.

Check the VRM portal logs around the exact dropout timestamps — the AC-input voltage graph usually tells you straight away if it's a grid quality issue triggering a false disconnect rather than a config problem.

VivaroWanderer | 312 posts | Array

Worth adding — check your grid code settings in VEConfigure. In the UK we're supposed to be on G98/G99, and if yours ended up on a different regional standard (easy to happen after a firmware flash), the MultiPlus applies different voltage/frequency acceptance windows. A wobbly grid restoration often sits just outside those windows momentarily, so the unit legitimately rejects the incoming AC rather than transferring — looks identical to a bug but it's actually working as intended.

Also seconded on @Rob1963's current limit point — I run a Fogstar Drift 48V on my motorhome build and had to explicitly set AC input current limit via the Cerbo GX rather than leaving it to VEConfigure alone, as the two settings can conflict post-firmware update.

The VRM portal diagnostic logs will show exactly what the unit thought the incoming AC looked like at the moment of rejection — massively useful for narrowing this down.