From the firmware version 48x I've been seeing a strange behaviour in the modulation management of ESS. I've a dedicated question on it. Some of the effects are high feed-in rates, power imports of 2ths magnitude that take more than half an hour to come back to setpoint (+20w) and phantom consumption jumping from ac-out to ac-in (where nothing is connected).
As the power company doesn't allow me to feed-in, I've been looking for a way to be "Virtual Switched" while retaining the features of ESS, like ESS mode for battery maintain or grid setpoint (for the same purpose). First started with VS and slowly progressed from there.
So finally I found the way of archiving it and then I'm experiencing the above problems, but with a more consistent pattern. And I think that I can reproduce it almost at will.
Clarification: Contactor/Relay on grid feed left open during the low consumption hours. Night and early morning. The scenarios below happen on grid activated period. More than enough time has passed to allow charger activation.
My appreciation is that with time, something gets stuck at ESS management, and then the Multiplus2 stops doing "mode changes". For example ac-out2 toggle does not happen.
Other example is after being charging batteries during the morning while feeding a water heater, 2000-2500w steady load, someone starts cooking and the consumption reaches nominal max, two things happen:
It doesn't raise the mppt production to max (only smartsolar, no pv inverters), like still following some charging curve despite the increased load.
And does not compensate with grid. So overload happens.
The same situation in a Virtual switch configuration, ends with a VS that doesn't connect to the grid for supplement. So that's why I'm calling it "not doing working mode changes".
If you turn off and on the multiplus2, or reboot though the GX, it works as expected. Distributes energy flow and compensates with grid, but only for a time.
That sound me like if getting the same low readings (or zero) from ac.in for long periods of time, gets some variable so much sinked that it doesn't recover with small positive values or bursts. Just like using a PID controller and getting an integral windup due of so much time being deviated from setpoint. I documented that behaviour during grid blackouts.
I was going to share a youtube video with my findings on my ESS configuration, but felt that first I should talk with someone about this problem.
