ESS: DC-coupled MPPT 250/100 repeatedly turned off (DeviceOffReason=8) during high PV - EasySolar-II

DIY
, ,
1

Hi everyone,

I’ve been struggling with this issue for about 2 years now and I’ve finally taken the time to do a thorough diagnostic. I’m hoping someone from the Victron team or the community can help, as I’ve exhausted all the settings I could find.

In short: my MPPT keeps getting shut down by hub4control during peak solar production, even though the battery has plenty of room and all feed-in settings are correctly configured. The same MPPT works perfectly fine in charge-only mode (no ESS) at full power.

Here’s the full diagnostic:

System

  • EasySolar-II GX 48/5000 (Multiplus FW v1362, MPPT SmartSolar 250/100 rev2 FW v202495)
  • Venus OS v3.72 (NanoPi, armhf)
  • Battery: JK-BMS via CAN, 2x16 LiFePO4 EVE 280Ah, CCL=171A, CVL=58.5V
  • Grid meter: Victron VM-3P75CT (Ethernet)
  • ESS mode: Optimized with BatteryLife, DVCC enabled

Problem

During high PV production (above ~5500W), the MPPT is repeatedly turned off by hub4control with DeviceOffReason=8 (Remote Input). The MPPT drops from 5600W to 0W in seconds, stays off for 15-30 seconds, restarts, ramps back up to 5500W+, and gets turned off again. This cycle repeats indefinitely during peak solar hours.

This problem has been present for approximately 2 years, across multiple Venus OS versions.

In “charge only” mode (ESS disabled), the MPPT happily produces 5600W and sends 100A into the battery with zero issues. The hardware is not the limitation – it is hub4control’s ESS logic that is causing the cutoff.

What we tried (none of these solved it)

Setting Value tested Result
PreventFeedback 0 Still cuts off
OvervoltageFeedIn 1 Still cuts off
MaxFeedInPower 10000 Still cuts off
AcPowerSetPoint +50W (was -20W) Still cuts off
SharedVoltageSense 0 (disabled) Still cuts off
RunWithoutGridMeter 0 (VM-3P75CT installed) Still cuts off
Hub4 /Overrides/FeedInExcess 1 Still cuts off
Hub4 /Pv/Disable stays at 0 during cutoff Not the mechanism
Hub/ChargeVoltage stays at 56.0V during cutoff Not the mechanism
Link/ChargeVoltage stays at 56.0V during cutoff Not the mechanism
MPPT /Mode watchdog (force to 1) hub4control overwrites immediately Cannot override

Diagnostic data

Typical cutoff sequence (captured with 1s polling)

Before cutoff – everything normal, MPPT at full power:

Time      |State|OffR|PV_W  |Grid_W|BatI |SOC
13:02:40  |252  |0   |5599.7|72.0  |69.9 |38%
13:02:43  |252  |0   |5603.0|71.0  |70.7 |38%
13:02:46  |252  |0   |5593.0|70.0  |69.9 |38%
13:02:49  |252  |0   |5604.0|82.0  |70.3 |38%

Cutoff – MPPT drops to 0W instantly:

13:02:55  |252  |0   |5597.0|93.0  |70.3 |38%  <-- last good reading
13:03:26  |0    |8   |477.2 |155.0 |9.8  |38%  <-- CUTOFF
13:03:29  |0    |8   |20.9  |67.0  |-30.4|38%
13:03:32  |0    |8   |0.0   |-221.0|-28.9|38%

Recovery then immediate cutoff again:

13:03:35  |245  |0   |0.1   |3.0   |-31.2|38%
13:03:41  |252  |0   |2231.0|44.0  |22.8 |38%
13:03:44  |252  |0   |5533.3|52.0  |35.4 |38%
13:03:50  |0    |8   |386.8 |89.0  |9.1  |38%  <-- CUTOFF AGAIN after 6s!

Key observations

  1. 156 cutoffs recorded in a single afternoon monitoring session
  2. Cutoffs occur exclusively when PV > ~5500W
  3. At the moment of cutoff, Grid_W is often positive (pulling from grid, NOT feeding back) – so it is not a feed-in prevention issue
  4. Hub/ChargeVoltage remains at 56.0V during cutoff – the voltage mechanism is not the trigger
  5. /Pv/Disable remains at 0 during cutoff – hub4control is not using this flag
  6. PvPowerLimiterActive = 1 during cutoff
  7. Battery voltage is 53-54V, well below CVL of 58.5V
  8. Battery current is 68-70A, well below CCL of 171A
  9. SOC is low (25-41%) – battery has plenty of room
  10. The MPPT can deliver 100A (5400W+) to the battery with zero issues in charge-only mode (no ESS)

Hub4control dbus state during cutoff

com.victronenergy.hub4:
  /PvPowerLimiterActive = 1
  /Pv/Disable = 0
  /Overrides/FeedInExcess = 1
  /MaxChargePower = []  (no limit)

Analysis

Hub4control appears to have an internal PV power limiter that activates around 5500W on DC-coupled systems (EasySolar-II with internal MPPT), regardless of:

  • Whether the battery can accept more current (CCL=171A, actual=70A)
  • Whether feed-in is allowed (OvervoltageFeedIn=1, PreventFeedback=0)
  • Whether the grid meter shows import or export

The limiter seems to use VE.Bus direct communication to turn off the MPPT, bypassing all dbus-level settings and overrides. A watchdog script that forces /Mode back to 1 is immediately overwritten by hub4control.

This issue has been reported by other users:

Expected behavior

With OvervoltageFeedIn=1 and PreventFeedback=0, the system should let the MPPT run at full capacity and export any excess PV power to the grid, rather than shutting down the MPPT entirely.

Request

Could the Victron team investigate the PvPowerLimiter logic in hub4control for DC-coupled EasySolar-II systems? The current behavior causes significant energy loss (5600W lost for 15-30 seconds, 156 times in an afternoon).

I’m happy to provide remote access, additional logs, or test any beta fix. Thanks in advance for any help!