Battery detection JKBMS & Pylontech

Been scratching my head over something similar on my narrowboat setup and wondered if anyone else has run into this.

I've got a mixed bank situation — a JKBMS-managed lithium pack alongside some older Pylontech units — and I'm noticing some odd behaviour in how Victron's new UI handles detection compared to the old interface.

On the old dashboard (and in VRM), both battery sets show up correctly as separate entities. The new UI seems to be collapsing them into one, or just picking one and ignoring the other entirely. Has anyone else seen this?

A few things I'm trying to work out:

• Is this a known bug in the new UI, or is it intentional behaviour that I'm misunderstanding?
• Does the JKBMS need specific firmware to play nicely alongside Pylontech units on the same Cerbo GX?
• Is there a workaround, or should I just stick with the old interface for now?

I realise mixing battery types isn't ideal — it's evolved over time as I've expanded the system — but both banks are managed independently with their own BMS, so in theory they should be visible as separate devices.

Would be interested to hear from anyone running a similar mixed setup, particularly on boats or cabins where you sometimes end up with a bit of a patchwork system rather than something clean and uniform from the start. Is this purely a display/UI issue or could there be something deeper going on with how the Cerbo is polling the BMS data?

@ExJoiner this exact scenario bit me on my shepherd's hut build last summer.

The root issue is usually that your inverter/charger is trying to auto-detect battery type and the JKBMS UART comms aren't speaking a protocol your system recognises — it's not Pylontech CAN, it's not Victron VE.Bus, it's somewhere in between.

What solved it for me was disabling auto-detection entirely on the Victron MultiPlus and setting battery parameters manually:

• Absorption/float voltages matched to your cell chemistry
• Tail current set conservatively
• Let the JKBMS handle cell-level protection independently

The two systems then coexist without either trying to assert authority over the other. Think of it like two shift commanders — they need clearly defined responsibilities or chaos ensues.

What inverter/charger are you running? That'll determine exactly which settings need overriding.

@ExJoiner had almost identical grief with my shepherd's hut when I first wired up my Victron Multiplus alongside a JKBMS pack.

The bit nobody tells you upfront — the JKBMS needs its CAN bus protocol set correctly to match what the Victron is expecting. Mine was defaulting to a Pylontech-style protocol but with slightly wrong cell count parameters, so the Cerbo GX was just... confused. Kept dropping detection intermittently.

Worth checking:

• JKBMS app → "Protocol" setting
• Cell count matches exactly what you've configured in VictronConnect
• Baud rate (usually 500k but double-check)

Once I got those three aligned it's been rock solid. The mixed bank side of things is a separate headache entirely though — are you trying to run them in parallel or just on the same bus with separate charge sources?

Running into something adjacent on my boat — has anyone checked whether the DVCC settings in Cerbo GX (if you're running one) are actually seeing the JKBMS as a managed battery?

Mine was stubbornly defaulting to fixed voltage charge curves even though the BMS was supposedly communicating over the VE.Can network. Turned out the JKBMS firmware needed updating before it would properly handshake with the Victron ecosystem.

Worth checking what firmware version you're on @ExJoiner — there were some significant compatibility updates pushed out last year specifically around Pylontech protocol emulation. The JKBMS can masquerade as Pylontech to Victron kit, but only reliably on more recent builds.

What's your comms cable situation — USB-RS485 or direct CAN?

@GeoffHenderson80 good shout on DVCC — that's actually what sorted a similar headache on my van build. Worth checking whether your Cerbo is seeing the JKBMS via BMS CAN or whether you've bodged it through a VE.Can adapter, because the two behave quite differently in terms of what data actually gets passed through to the Multiplus.

In my setup I had to explicitly set DVCC > SVS (Shared Voltage Sense) to enabled before the Multiplus would properly trust the BMS voltage readings rather than its own internal measurement. Made a noticeable difference to how the charge curve behaved.

Also — is your JKBMS firmware up to date? There were some known Pylontech protocol emulation quirks in older versions that caused detection failures with Victron kit.

@GeoffHenderson80 @Thommo9 DVCC sorted it for me on the boat too, but the bit that caught me out was the Shared Voltage Sense toggle — had it enabled when it shouldn't have been with a mixed bank. Victron's own docs are a bit vague on this edge case.

Also worth checking your JKBMS firmware — there was a version floating around last year that had dodgy Pylontech protocol emulation. Updating mine via the JK app cleared a phantom detection error I'd been chasing for weeks.

If you're on a Cerbo, the VRM portal diagnostics will usually show you exactly which device is winning the "who controls charging" argument. Saved me hours of head-scratching on my narrowboat.