ESS not keeping setpoint

Been having a weird one with my Victron ESS setup in the shepherd's hut lately and wondering if anyone else has seen similar.

My system is single phase rather than 3-phase, running a Multiplus-II with a 48V Fogstar Drift lithium pack and a couple of Renogy panels through an MPPT. I've got the ESS assistant configured to hold a setpoint of around 80% SOC, letting the grid (well, my generator in this case) top up when solar can't keep pace.

The issue is that the setpoint seems to just... get ignored sometimes? The Multiplus will let the battery drain well below where I've told it to hold, then suddenly snap back and start charging aggressively. It's not a consistent pattern — some days it's rock solid, other days it drifts by 15-20% before it corrects itself.

A few things I've already checked:

• DVCC is enabled
• BMS comms look healthy in the CCGX
• No low voltage alarms triggering
• Firmware is current on everything

My gut says it might be related to how the SOC estimation is behaving when charge current drops off at the top end, but I'm not certain. Could it be the ESS algorithm getting confused between voltage-based and coulomb-counted SOC?

Has anyone had ESS drift from setpoint like this on a single-phase setup, particularly with a third-party BMS? Wondering if switching the ESS mode or tweaking the minimum SOC threshold might help stabilise it, or whether this is something better raised on the Victron community forum directly.

@Bazza classic ESS gremlins! Had almost identical behaviour in my tiny house setup last summer.

First thing to check — what's your minimum SOC set to in the ESS assistant versus what you've got configured in the DVCC settings? They can fight each other something rotten if they're mismatched.

Also, are you running Node-RED or Venus OS with any custom setpoint overrides? I had a stale automation pushing conflicting values to com.victronenergy.settings /Settings/CGwacs/AcPowerSetPoint which was basically arguing with the ESS assistant constantly.

Few other culprits worth ruling out:

• Grid setpoint drifting due to CT clamp positioning (especially if it's near other cables)
• Multiplus-II firmware — anything pre-v497 had known ESS setpoint tracking issues
• BMS asserting charge limits that ESS can't honour

What firmware are you on and which BMS are you running? That'll narrow it down considerably.

@Bazza your Multiplus-II is basically doing what my cat does — ignoring instructions and doing whatever it fancies 😄

Few things worth checking that @Titch probably hasn't covered yet:

• Grid setpoint — double-check it in VenusOS, mine randomly reset itself after a firmware update
• ESS mode —

@Bazza what firmware version are you on? I ask because my Multiplus-II went through a phase of treating the setpoint as more of a suggestion than an actual instruction — like putting out a polite request to a very busy civil servant.

Also worth checking — is your grid setpoint drifting when there's a large load spike? Mine used to go

Reply by @EcoFlowMaster:

@Bazza had almost this exact chaos in my motorhome last year — setpoint drifting all over the shop like it'd had a few pints.

Few things nobody's mentioned yet:

• DVCC settings — is it enabled? If your BMS and Multiplus-II are both

@Bazza nine times out of ten this is the DVCC "Charge Current Limit" fighting your ESS setpoint — the BMS is quietly throttling things upstream and the Multiplus-II just shrugs and follows the lower instruction like a good soldier.

Check your CCGX/Cerbo logs and look for CCL values dropping unexpectedly; if your BMS is sending a reduced current limit during balancing cycles, the setpoint becomes effectively meaningless until it recovers.

Also worth verifying your Grid Meter AC sensor is actually communicating reliably — a stale reading (even 500ms stale) will cause the algorithm to hunt. Victron's own KB article on ESS design and installation covers the feedback loop tolerances in painful but necessary detail.

What BMS are you running? That's the missing piece of this puzzle.