I have a ESS system which has been running perfect since 6 months. No issues on V3.67 since 11/2025. On V3.70 however, the battery is drained and energy is sent to the grid while solar is available to charge the battery (picture 1). Or if even more solar is available, the battery is charged, but at a low power while excess solar is exported (picture 2). I have tried every possible setting to find out what causes this behaviour, without succes. When reverting back to 3.67 everything works as expected. Any ideas? Thanks!
Multiplus II 48/5000 GX + SmartSolar 150/45 r3 + VM-3P75CT + 400Ah BMS battery + 3.6Kw AC coupled solar
Grid setpoint: 0W - Charge control: CAN BMS (CCL:80A CVL:56.5V during screenshots)
I have tested again with 3.70 now the sun has set, system behaves as expected without solar input. It seems to only misbehave when there is solar input (DC + AC coupled):
I did some extra tests. 3.70 has been running during the night (no solar input) without issues. The 0W grid setpoint is maintained perfectly. But when the sun rises, not all of the PV energy is going to the battery without any obvious reason (and at higher SoC the battery is even discharged to the grid as seen yesterday). I have tried disabling DVCC and BMS control, disabling export DC, reboot, power cycle, but the behavior stays the same. Rolling back to 3.67 fixes the issue. DESS is not (and has never been) activated in this installation.
CVL: 56,5V CCL: 200A - Nothing else is limiting charging.
As i dont need any of the new functions in the new fw for now, this is not urgent for me. But if there is anything i can test on this setup please let me know, happy to help.
I had the same thing happen to me today. Excess solar was injected to the grid even when there was plenty of room in the battery. Tried with ESS, DESS, limit grid feed in of 0w, keep batteries charged, ⌠same behavior.
Only AC coupled PV (6kw) linked to 3x MP-II 5000VA. Should allow 3x70a. So the 86A that is charging in the image is only at 40% of full capacity. Also my battery bank is 3x 16kw so battery can also easily handle way over the 86A.
Prior to the update I was able to force charge well over 9000w from grid. and all 6kw of solar was able to be stored in the battery.
Now after the update solar seems to be limited to 5kw and even forced charge is not reaching the same peak as before
My guess: one cell has reached high voltage (depending on your BMS settings, i.e. 3.5 V), and the BMS is telling the MP2 to reduce charging, and so the MP2 is feeding the excess solar yield to the grid.
Max cell voltage at the time was 3.34v and max cell temp 25c.
Pretty sure it is the OS as this only started happening after the update. Will see tomorrow when there is solar if it happens again. And if so test to see if reverting firmware fixes it.
Below some screenshots where it is clear that all above the 5000w solar is injected without reason. Seems like there is a hard limit of 5000w of charging that is nowhere configured
I can confirm. In 3.67 when âKeep batteries chargedâ is selected the current drawn from the grid is capped at the âpeak shavingâ value (as it should). In 3.70 the battery is charged at a random grid current value, way below the configured peak shaving current.
With my guess I referenced the OP from user â555â. His picture shows a battery status with 98% SOC and 53.48 V. With this voltage the average cell voltage already is beyond your value. And with these values it is very likely that one of the cells already hit the set maximum so that the BMS hat to reduce charge power.
Hi, no the battery being almost full was not the reason for the export. CCL was still 80A at that time as stated. I have repeated the test at 62% SoC as you can see in the next post, with the same behavior. This is clearly something misbehaving in firmware 3.70 because everything returns back to normal immediately after rollback to 3.67.
As it is also present in Pigmasterâs setup (which has only AC coupled PV and no DC MPPT chargers which probably complicate DC-AC/infeed calculations) this issue should be easy to reproduce/fix.
Today I had the same issue when charging being limited to 4500-4700W for no obvious reason.
Iâve reverted back to 3.67 and issue instantly was resolved.
Dont know why/what is causing this, but for now the solution seems to be reverting back to the previous VenusOS.
Hi @555, Iâm curious if that name is inspired by the NE555 timer chip, but that aside.
I looked at the data of your site at the time you reported this. Your user defined charge voltage is/was set to 54.8V, and since the battery was at 54.8V at the time, all of the solar charger power was considered excess and was sent to the grid.
Interestingly the battery also sent CCL=0 (charge current limit = 0) at the same time, which means the Multi also wasnât allowed to charge, so all the PV inverter power also went to the grid.
When you later reboot back to 3.67, the charge voltage is also higher, could it be that you were experimenting with this?
Can you give me more details about your ESS settings. Do you have the ESS setting âSelf-consumption from batteryâ set to âOnly critical loadsâ?
I ask because there have been very few changes to the ESS control loop in the GX device. The few changes there were, were all related to peak shaving with the above setting enabled.
It never pays to be first to an update, more so for major updates. Let others fix the issues.
The more complex or modified a system is, the more conservative should the approach to updating be.
If you donât absolutely have to have the latest version, wait for the next minor update that inevitably follows.
FOMO and power electronics arenât a good mix
