Pylontech + MPII : Effective charge voltage qui dépasse 52.4V et atteint 54.4V

Anyone else seeing their Victron MultiPlus-II pushing charge voltage beyond what the Pylontech actually wants?

Running a MP-II 48/5000 with a stack of Pylontech US3000Cs here in Norfolk, and I've watched the effective charge voltage creep up to 54.4V when the Pylontechs are quite clearly screaming "52.4V is fine, thank you very much" via DVCC.

Config is fairly standard — AC-in only (no solar on this particular setup), ESS enabled, grid feed-in active. Two MPPT controllers sitting idle on another system nearby but not part of this loop.

The BMS comms look healthy in VRM, Pylontech app shows no alarms, but something in the charge algorithm seems to be ignoring the battery's own voltage ceiling under certain load conditions. Feels like ESS is being a bit enthusiastic.

Few things I'm suspicious of:

• DVCC SVS (Shared Voltage Sense) doing something odd
• A firmware mismatch between the MP-II and VenusOS
• Grid feed-in logic temporarily overriding battery limits during export ramp-up

Has anyone nailed down whether this is a known quirk with Pylontech + ESS, or am I chasing a ghost? Curious whether those running Fogstar or other BMS-controlled packs see similar drift, or if it's specifically a Pylontech CAN comms thing.

Would love to know what firmware versions people are running — VenusOS 3.x seems to have changed a few ESS behaviours quietly.

@NorfolkDweller had almost the exact same head-scratcher on my narrowboat last winter. The MP-II was cheerfully pushing past 52.4V and the Pylontech BMS was having a quiet panic about it.

Turned out my DVCC settings in VenusOS were the culprit — specifically, I hadn't properly enabled Use solar charger current limit alongside the strict voltage limits. Once DVCC was genuinely handing control over to the BMS via the VE.Bus, the voltage settled beautifully at what the Pylontechs actually requested.

Worth double-checking:

• DVCC fully enabled in Venus
• SVS (Shared Voltage Sense) ticked
• The Pylontech CANbus cable is actually the one Victron specifies — I once used a generic cable and it was silently corrupting the comms

What firmware are you running on the GX device? That made a measurable difference for me going from v2.87 to v3.x.

@NorfolkDweller worth checking whether you've got the Pylontech profile selected in VEConfigure and that the DVCC settings in VRM/Cerbo are actually overriding local settings. I had a similar thing on my shepherd's hut setup — turns out my MP-II was ignoring the BMS-controlled limits because DVCC wasn't properly enabled on the GX device.

The Pylontech comms should be telling the system to cap at 52.4V absorption. If it's hitting 54.4V something's broken in that communication chain.

Quick checklist:

• DVCC switched on in Cerbo/Venus settings
• BMS cable properly seated (sounds daft but...)
• VEConnect showing the Pylontech as "connected" not just "detected"

What firmware are you running on the US3000Cs? Some older versions had quirks with the MP-II handshake that a Pylontech firmware update sorted.

@NorfolkDweller had this exact issue when I first set up my tiny house system. Double-check your battery monitor source in DVCC — if it's set to the MP-II rather than the Pylontech BMS via VE.Can, the inverter basically ignores what the battery's actually asking for.

Once I let the Pylontech BMS take full control through DVCC (SVS, SCS, STS all enabled), the charge voltage sorted itself out immediately. The BMS tells the MP-II what it wants rather than the other way round.

Also worth grabbing the latest Pylontech firmware if you haven't — made a noticeable difference on mine.

@NorfolkDweller had this on my garden office setup when I first commissioned it. Turned out my BMS Connection setting in VEConfigure was set to "None" instead of "VE.Bus BMS" — once that was wrong, DVCC couldn't actually receive the CVL from the Pylontech and the MP-II just did whatever its own absorption setting said.

Pull up a GX Device → DVCC screen and check the "Active charge voltage limit" is showing something sensible like 52.4V. If it's blank or showing your manually-configured absorption voltage instead, the BMS data isn't getting through at all.

Also worth confirming firmware — Pylontech had some CAN comms quirks on older US3000C firmware that caused exactly this sort of drift upward. Their UK distributor (Bimble Solar have been helpful in the past) can advise on flashing if needed.