MultiPlus-II ESS – short total blackout without any logged error

Had something weirdly similar on my narrowboat setup a few weeks back — single MultiPlus-II rather than a 3-phase rig, but the principle felt the same. Brief total dropout, everything killed for maybe 2-3 seconds, then back up as if nothing happened. VRM showed absolutely nothing. No warnings, no alarms, no blip in the SOC graph that would explain it.

Took me ages to even convince myself it had actually happened and wasn't just me imagining things.

My working theory at the time was a brief BMS disconnect — my Fogstar cells have a Daly BMS and I've read others saying a momentary BMS protection trip can cut the DC side faster than the MultiPlus can log anything meaningful. By the time it's recovered, there's nothing to write home about in the event log.

A few things worth checking if others are seeing this:

• DC cable connections — especially ring terminals under vibration (more relevant on a boat, fair enough)
• BMS communication timeout settings — if you're on VE.Can or even just the two-wire assistant, a hiccup there can cause a controlled shutdown
• AC input quality — even with AC-Out configurations, a weird grid blip can sometimes trigger a transfer that looks like a blackout from inside the property

Has anyone on a static cabin or house setup with a 3-phase arrangement seen this? Curious whether it's more likely to show up on larger installations where the phases have to stay in sync. Feels like there's something subtle going on that Victron's logging just doesn't catch at the right resolution.

@BoatEwan sounds like the classic "BMS hiccup pulls the CAN bus low for a split second and the Multi doesn't know what to do" scenario. Seen it on my motorhome setup with a Fogstar Drift — BMS momentarily freaks out under a heavy load spike, drops comms, Multi goes into protective shutdown before it even logs anything.

Worth checking if your BMS has a keep-alive issue. Some cheaper units do this under cold conditions too.

Are you running VE.Can or VE.Bus to the Multi? And what's the BMS? If it's not a Victron-native battery, the DVCC settings can cause exactly this — particularly SVS and SCS being enabled when they shouldn't be.

Interesting thread — has anyone actually scoped the CAN bus during one of these events? I've been meaning to do this on my cabin backup setup for a while but haven't got round to it.

What I'm wondering is whether it's worth adding a Cerbo GX with a local VRM log interval tightened right down — does that capture enough granularity to catch a sub-second BMS hiccup, or does even that miss the event entirely?

Also, which BMS are people running here? I've seen suggestions that some of the cheaper units (not Fogstar-grade stuff) can send malformed CAN frames that confuse the MultiPlus-II momentarily. Whether that's enough to trigger a full dropout rather than just a warning I genuinely don't know.

@DownsCruiser — when you say it pulls the CAN bus low, is that a hardware fault or a legitimate BMS protection trip being misinterpreted?

@PartnerNomad I actually did scope mine briefly after a similar dropout on my shepherd's hut setup — nothing conclusive but there was definitely some noise around the transition. The thing is, even if you capture something, correlating it to what the Multi was "thinking" at that moment is another matter entirely.

What I'd also look at alongside the CAN stuff is the DC bus voltage during the event. My Fogstar cells were fine but I had a dodgy cable lug that was dropping voltage just enough under load to confuse the BMS into a momentary protection trip. Completely silent in the logs.

Worth checking every connection physically before going deep on firmware or comms debugging — boring answer but it was my problem in the end.

@MarshLover interesting you got nothing conclusive — I had almost identical results when I scoped mine on the shepherd's hut. What I did eventually track down was a loose Anderson connector between the Fogstar cells and the BMS, creating just enough momentary resistance to spook the whole chain. Not CAN-related at all in my case. The Victron forums kept pointing me toward comms issues but the culprit was far more mundane. Worth working through your physical connections methodically before chasing ghosts in the data — torque every terminal, check your crimps under load. Pulled my hair out for three weeks before finding it.

@BoatEwan @PartnerNomad — the CAN scoping approach is worthwhile but you really need to be capturing at the right moment. Most analysers with passive triggering will miss a sub-100ms dropout entirely.

What I'd strongly recommend: set up VRM remote monitoring with high-frequency logging enabled, and simultaneously watch the DVCC SOC divergence between your BMS and Victron's own calculation. In my narrowboat install (Fogstar Drift 200Ah, Cerbo GX), I traced a near-identical blackout to the BMS momentarily asserting charge-inhibit and discharge-inhibit simultaneously — the MultiPlus sees that as a hard fault and shuts down cleanly with no alarm because it considers the shutdown commanded.

The logged silence is the giveaway — it's not a crash, it's an obedient shutdown.

Check your BMS event log independently of VRM. Daly and JK both store these locally even when CANbus communication drops.