Multiplus II mit LiFePO4 NG

Anyone else noticed that Victron keep quietly shifting the goalposts on LiFePO4 setup with the Multiplus II? 🤔

Just been rewiring my whole setup after adding a couple of Fogstar Drift 200Ah cells to the bank, and I genuinely wasn't sure whether I still needed to load the LiFePO4 assistant in VEConfigure or whether the newer firmware handles it more gracefully out of the box.

Back when I first commissioned my original system, the assistant was absolutely non-negotiable — without it the Multiplus would happily try to charge like it was dealing with a lead-acid and your BMS would throw a proper wobbler. But with the NG (New Generation) units and recent GX firmware updates, there seems to be a bit of conflicting info floating about online, half of it in German which isn't ideal when you're knuckle-deep in cable ferrules at 11pm.

My current thinking:

• Old advice: Always load the LiFePO4 assistant, full stop
• New situation: Some NG units with updated DVCC settings via Cerbo GX might not strictly need it if your BMS communicates properly over CANbus

But "might not need it" isn't exactly confidence-inspiring when we're talking about a battery bank that cost more than my car.

Has anyone on here recently commissioned a Multiplus II GX or standard Multiplus II (50Hz UK spec obviously) with a communicating LiFePO4 battery and skipped the assistant entirely? Did it behave itself, or did you end up with a very expensive paperweight?

Would love to know what firmware versions people are running and whether the assistant is still considered best practice or becoming a bit legacy at this point.

Yeah @Brummie84, noticed the same thing when I expanded my own bank recently. Victron's ESS assistant parameters in particular seem to shift between firmware updates without much fanfare in the changelog. The sustain voltage settings especially can catch you out if you're not watching — what was fine for your original pack might be pushing the new Fogstar cells outside their comfort zone once you've paralleled them up.

Worth double-checking your DVCC settings too, since the Multiplus II handles shared current control differently once you're above a certain capacity threshold. Are you running a Cerbo GX in the loop? Makes life considerably easier for keeping everything talking properly across a mixed-capacity bank. What BMS are you using on the Fogstars? That'll make a big difference to how you approach the charge parameters.

@Brummie84 worth checking whether your Fogstar Drift units are presenting the correct cell chemistry flags over their BMS comms — particularly if you're using the VE.Bus BMS v2. Victron quietly updated the DVCC charge current limits in firmware 502 onwards, and if the BMS isn't advertising itself correctly, the Multiplus II will fall back to conservative defaults that look like misconfiguration but aren't.

On my narrowboat I had identical symptoms after adding capacity — absorption voltage kept creeping down unexpectedly. Turned out the Shared Voltage Sense was overriding my manually-set charge profile because DVCC was partially active without me realising.

Check in VictronConnect:

• DVCC — fully on or fully off, no halfway
• SVS/SCS/STS toggles underneath it
• Confirm BatteryLife is disabled if you're running straight self-consumption

The interaction between those three settings and a third-party BMS is where most of the confusion lives.

@Brummie84 oh, the joy of discovering that what worked perfectly last Tuesday has been quietly deprecated by a firmware update you didn't ask for 🙃

Had exactly this palaver when I wired up my shepherd's hut backup system — Multiplus II playing silly beggars because the charge profile I'd carefully dialled in suddenly had a new "recommended" default sitting on top of it like an uninvited guest.

The bit that catches people out specifically: Victron's absorption voltage recommendations for LiFePO4 crept down in recent VenusOS builds. If you're still running the older figures (28.8V on a 24V bank, say), you might find the BMS on your Fogstar Drift is technically correct to throw a strop.

Worth cross-referencing your current settings against the Victron LiFePO4 battery compatibility page — they do actually update it, annoyingly without fanfare.

@Brummie84 genuine question — are you running DVCC with an external BMS or letting the Multiplus handle charge control itself? I've got a similar Fogstar Drift setup in my static caravan and found the behaviour changed significantly depending on whether the BMS is publishing data over VE.Can or just sitting there doing its own thing silently.

Also worth checking — what firmware are you actually on? Victron's changelog isn't exactly light reading but there were some meaningful changes to the LiFePO4 charge algorithm around the 500-range builds that caught me out.

@JubileeClipHero that "deprecated by firmware" pain is real. Lost a weekend to that exact scenario last spring.

What does your current absorption/float voltage look like in VEConfigure? That'd help narrow down where the goalposts have actually moved for you.

@LisaStewart71 DVCC is non-negotiable with external BMS on LiFePO4, full stop. Had my Fogstar Drifts throwing fits until I stopped letting the Multiplus "help" with charge decisions.

@Brummie84 main thing people miss — SVS and STS need disabling when your BMS is talking CANBUS. Victron's defaults assume you're flying blind. You're not.

Also check your ESS assistant version. Mine silently reverted charge voltage limits after a firmware push last year, nearly cooked a cell group before I caught it on VRM.

Victron kit is solid but their changelog hygiene is absolutely appalling.

@GrumpyBuilder is right on DVCC, but worth adding — when you enable it, make sure SVS (Shared Voltage Sense) and SCS (Shared Current Sense) are actually doing what you think. In my shepherd's hut setup I had both ticked but the Multiplus was still ignoring the BMS current limit because the VE.Bus device was overriding via its own charge profile.

The fix was setting the DVCC max charge current explicitly in VictronConnect rather than leaving it to the BMS alone. Sounds counterintuitive but the Multiplus II needs that hard ceiling set at the DVCC level, not just communicated from the BMS.

@Brummie84 — which BMS are you running on the Fogstar cells? JK or something proprietary? That'll determine whether you need the VE.Can or VE.Direct path for proper DVCC handshaking.