Staircase like increase in max/min battery voltages when SoC = 100%

Been noticing something similar with my setup and can't decide if it's normal behaviour or worth worrying about.

Running a single-phase system here — Victron Multiplus-II 3000 with a 3-unit Fogstar Drift 48V LiFePO4 stack and a Cerbo GX managing the whole lot. When the batteries hit 100% SoC, I'm seeing the reported max and min cell voltages creep upward in what looks like distinct steps rather than a smooth curve. Almost like the BMS is re-anchoring its reference points once it's confident the pack is genuinely full.

My gut says this is the BMS doing its balancing pass and recalibrating — essentially "staircase-ing" up as each cell group gets properly topped off. But I'm not 100% sure whether what I'm seeing in VRM is a genuine voltage event or just a reporting/logging artefact from how the Cerbo polls the data.

Anyone else seen this pattern, particularly with third-party batteries talking to Victron kit via CAN bus? I'm curious whether:

• It's specific to certain BMS firmware versions
• The Cerbo logging interval is just making a smooth ramp look stepped
• Or there's something actually happening at cell level worth keeping an eye on

Not losing sleep over it — voltages are all within normal range and the system's behaving fine day-to-day. Just one of those things that nags at you once you've noticed it.

Would be good to hear from anyone running Pylontechs or similar managed batteries who's spotted the same thing.

@SaltyHiker that staircase pattern is classic cell balancing behaviour — when your BMS tops off individual cells at 100% SoC, the pack voltage briefly "steps up" as each cell hits its ceiling before the balancer bleeds it back down, giving you that jagged voltage profile that looks alarming but is basically your Fogstar Drift doing exactly what it's supposed to. 🪜

Worth checking in your Victron VRM portal whether those voltage spikes are actually tripping any high-voltage alarms — if they're not, you're golden. If they are, a slight tweak to your Multiplus absorption voltage (drop it 0.1–0.2V) usually smooths the whole thing out nicely.

@BoatPaddy is right on the balancing front, but worth also checking if your Fogstar units are set to the same charge parameters. Had a nightmare with my cabin setup where two of my batteries had slightly different absorption voltages configured — caused all sorts of weird staggered behaviour that looked exactly like what you're describing.

Check your BMS app and make sure they're all singing from the same hymn sheet. Fogstar's app is decent enough for this.

Also — are all three units on the same firmware? Sounds daft but I've seen mismatched firmware cause oddities with how units report SoC to the Multiplus. Pull the logs from VRM if you're connected, the pattern in there will tell you a lot more than eyeballing the live display.

@BoatPaddy and @Curly38 have nailed the two most likely culprits, so I'll add a third possibility worth investigating: inter-unit communication latency.

With three Fogstar Drift units in parallel, if they're communicating via CAN bus to your Multiplus-II through a Cerbo GX, there can be a slight delay between units reporting their SoC. This creates a "staircase" effect as each unit sequentially reports hitting 100% and the system adjusts charge voltage accordingly.

Check your DVCC settings in Venus OS — specifically SVS (Shared Voltage Sense). If that's disabled, each unit may be reacting to slightly different voltage readings.

On my narrowboat I had near-identical behaviour with two units until I enabled SVS and set a single BMS as "the" controlling instance. Sorted it within a charge cycle.

Worth pulling the VRM logs to confirm the timing matches unit-by-unit balancing events.

Had the exact same thing with my shepherd's hut setup last summer. Stared at those voltage graphs for about three hours convinced something was dying dramatically 😅

Turned out to be completely fine — but while you're investigating @Curly38 and @DevonDweller's suggestions, worth double-checking your communication cables between the Fogstar units are actually seated