2 × 24V LiFePO4 packs with Seplos BMS 3.0 – aggregated CCL/DCL via CAN to Cerbo GX?

Running two Seplos BMS 3.0 units into a single Cerbo GX via CAN — anyone actually done this on a live UK install, or are we all just theorising on the internet?

My situation: planning a 24 V LiFePO4 setup on the narrowboat, two packs wired in parallel with separate BMS units, and I'd love the Cerbo to see a combined CCL/DCL figure rather than just whichever pack decides to throw a wobbly first.

From what I've dug up, the Seplos 3.0 allegedly supports CAN aggregation, but the documentation is about as clear as the Trent in February. A few questions for the hive mind:

• Does the Cerbo actually aggregate CCL/DCL from two Seplos units, or does it just display them separately and take the lowest value?
• Is there a master/slave config within the Seplos app that handles this properly, or are you bodging it with a separate CAN hub?
• Anyone running Fogstar Drift cells inside a Seplos enclosure on a marine install? Keen to know how they're holding up with the vibration.

I've seen people do similar setups stateside but UK narrowboat installs have their own quirks — 240 V shore power, BSS compliance headaches, the usual joys. Would be useful to hear from anyone running Victron ESS with non-Victron batteries who's cracked the CAN comms side without it all going sideways during a balancing cycle.

Drop your configs below — actual wiring notes and firmware versions preferred over "it should work in theory" comments, cheers.

@EssexNomad done this on my boat install — two Seplos 3.0 packs, 24V, talking to Cerbo via CAN.

The short answer: Cerbo only sees one BMS as the "master" DVCC source. The aggregated CCL/DCL doesn't just magically sum — you need to configure one unit as master and manually set conservative limits that account for both packs.

Key things I learned the hard way:

• Make sure both packs are on different CAN node IDs (obvious but easy to miss)
• The Cerbo will take the lowest reported CCL/DCL from whichever BMS it's polling — which is actually quite sensible protection-wise
• Firmware matters — I had to update both units before the Seplos handshake with VenusOS was stable

What firmware version are you running on the BMS units? That'll tell us a lot about whether you'll hit the same gremlins I did.

@EssexNomad done exactly this in my motorhome last winter and nearly threw the Cerbo out the window before I realised Seplos uses a non-standard CAN ID that needs manually configuring — Victron's own docs are about as helpful as a chocolate teapot on that front.

Key thing nobody mentions: you'll need to set one BMS as the master for CCL/DCL aggregation, otherwise the Cerbo just sees whichever pack shouts loudest and ignores the other entirely, which is fantastic when one pack hits a cell limit and the system merrily carries on cooking the other.

Fogstar cells in mine, Seplos 3.0, works reliably now — but budget an afternoon of swearing at it first.

@EssexNomad the bit nobody mentions until it bites them: Seplos BMS 3.0 sends its own pack data over CAN, but the Cerbo sees one BMS as the authority. If you've got two units both broadcasting CCL/DCL, you need to be deliberate about which one "wins" — typically the more conservative value should govern, but that's not automatic.

What I did on my motorhome build was configure one Seplos as the master CAN talker and handle inter-pack balancing logic separately. Victron's DVCC then responds to that single CCL/DCL signal cleanly.

Also worth knowing: the Seplos 3.0 uses the Pylontech protocol variant over CAN — confirmed working with Cerbo on firmware v3.30+. Double-check your Cerbo firmware before pulling your hair out like @RetiredNurse49 apparently did.

@EssexNomad ran almost exactly this setup on my shepherd's hut last season — two 24V packs, Seplos 3.0, Cerbo GX doing the talking.

The piece worth adding to what @LiFePO4Nerd hinted at: your MPPT charge profile needs to respect whichever pack is crying loudest, not some averaged figure. I watched my Victron SmartSolar cheerfully ignoring a high-cell-voltage warning on pack one because the aggregated CCL still looked healthy from pack two's perspective.

Fixed it by deliberately derating the overall CCL in the Seplos config — pessimistic, yes, but the packs have been happier ever since. Think of it like running two different-tempered crew members on a boat: you set the pace for the grumpiest one.

Worth checking your CAN termination resistors too. Mine caused ghost data until I sorted the 120Ω properly at both ends.

@CliveKnight curious what you ended up doing for the DCL aggregation — Cerbo just picks the lowest reported limit or does it actually sum them?

Had a similar headache on my 48V setup (different BMS, mind) where the inverter was throttling badly because it was only seeing one pack's CCL rather than both combined. Ended up having to sort it through the DVCC settings.

From what I've read, Seplos 3.0 on CAN with Cerbo should report as separate batteries in the device list — think you then need to check whether your Multiplus firmware is new enough to handle aggregated limits properly. Some older builds definitely weren't.

Worth checking Venus OS version too — anything pre-3.x had dodgy multi-battery CAN handling in my experience.