Victron Multiplus keeps throwing a low battery alarm at night but voltages look fine

Had this exact faff with my Multiplus last winter in the van — turned out the alarm threshold was set stupidly high in VE.Configure and had nothing to do with actual battery state. 🙄

Worth checking a few things though, because "voltages look fine" can be a bit misleading depending on where you're measuring:

• Resting voltage vs load voltage — are you checking under load or when everything's quiet? A Fogstar Drift 100Ah can sag quite a bit at 2am when the heating kicks in
• Battery temperature compensation — if your Victron BMV or SmartShunt is reading cold temps, it might be adjusting thresholds downward and triggering the alarm prematurely
• Low battery disconnect settings — pop into VE.Configure and check what voltage the alarm is actually set to, sometimes it ships with daft defaults for LiFePO4 if it's still configured for lead acid

Also, are you on lithium or AGM? Makes a massive difference because LiFePO4 sits at a pretty flat voltage for ages then drops off a cliff, so the Multiplus can get a bit panicky right at the tail end.

My cabin setup does the same thing occasionally when I forget to update the battery profile after messing about with settings — essentially the Multiplus is crying wolf but occasionally it's right and I've just not noticed the actual draw.

What does your BMV say for state of charge overnight — anyone else had this with the newer Multiplus-II units specifically?

@VoltMax makes a fair point about VE.Configure thresholds, but don't overlook something more insidious — voltage sag under load.

"Voltages look fine" at rest means nothing if you're not measuring under load. A tired or undersized battery bank can sit at 12.8V idle then crater past the alarm threshold the moment the inverter draws serious current.

Things worth checking:

• State of charge vs resting voltage — are you on LiFePO4 or lead? Lead acid is notorious for misleading resting voltages
• Battery internal resistance — Fogstar and Victron SmartShunt can both give you this via Bluetooth
• Cable sizing and connections — voltage drop across undersized cables registers at the Multiplus terminals, not the battery

On my narrowboat setup I had a similar phantom alarm that traced back to a corroded negative busbar connection dropping 0.4V under load. Completely invisible at idle.

What's your battery chemistry and bank size?

@LH_Marine raises a good point on voltage sag but also worth checking your battery temp sensor if you're running lithium. Victron's low voltage alarms can shift based on temperature compensation settings that are tuned for lead-acid — had this exact mess in my cabin setup with a Fogstar Drift 100Ah. The Multiplus was effectively "seeing" a lower voltage than reality because the compensation curve was wrong for lithium chemistry.

Quick check: in VE.Configure look at the Battery temperature compensation value — should be 0mV/°C for LiFePO4, not the default lead-acid figure.

@GeoffRobinson good shout on the temp sensor. One thing nobody's mentioned yet — check your DC cable sizing and connections. Had this on my garden office build with a Multiplus-II; alarm was triggering because of voltage drop at the unit due to undersized cabling from the battery. Multiplus was seeing low voltage even though the battery itself was fine.

Grab a decent multimeter and measure voltage directly at the Multiplus DC terminals under load. If it's noticeably lower than at the battery terminals, that's your culprit.

• Connections corroded or loose? Clean and retorque
• Cables undersized? Victron's own sizing guide is worth following properly
• Fuse holder resistance adding up?

Sorted mine with proper tinned marine cable and decent crimps. Night alarms disappeared completely.

@CE_Builds makes a valid point on cabling — worth adding that in VE.Configure there's also the "DC input low shutdown" and "DC input low restart" values which are separate from the alarm threshold itself. Easy to have those misconfigured and get confusing behaviour that looks like a genuine battery issue but isn't.

On my static setup I found the default values were clearly written with flooded lead-acid in mind. Running LiFePO4 (Fogstar cells), the voltage curves are completely different and those defaults become meaningless at best, misleading at worst.

I'd suggest logging actual battery voltage via VRM overnight before changing anything — rule out real sag first, then interrogate the thresholds. Guessing at settings without data rarely ends well.