I have SolarEdge SE9K inverter ac coupled on AC-OUT2, using Power limiting function.
Every day there is a restart of communication between Cerbo and SE (around 23:30, initiated by SE). After that READING recovers well, but WRITING fails. So Cerbo cannot write Dynamic power limit to SE.
I can see the data coming from SE to Cerbo, but there is no Dynamic power limit. In NODE-RED I can see NULL for dynamic power limit.
I have reached out SolarEdge support, they downgraded the firmware (to 4.22.44), they probably did some minor changes in settings, but there is no improvement.
The only thing that helps is restart of Cerbo. Most of the cases it helps immidately, once I had to do it twice.
I am using modbus tcp, ethernet cable. It worked well last year.
Hi @trpkosj , did this start recently or has it been going on for a while?
I looked at the log files on your system, and what the PV-inverter code does, and it seems that what happens is there is an error in reading the maximum power capacity of the SolarEdge inverter. It loses the connection at this critical point.
This is interpreted as “This device does not support limiting”, because there are older SolarEdge inverters that react in that way, they return an error if you try to read this particular register.
Unfortunately the probing for limiting support is done only once, when first detected, which is why it needs a reboot to do it again.
I’m not sure exactly what the solution is. Without the value it is reading here (the capacity of the system), it cannot do any limiting anyway, so the only solution would be some sort of scheduled retry.
it is difficult to answer. I know it started this year, 22.3.2026 I had an issue with send to much of the power to the grid. Later I found out the reason - loosing communication with SE.
But last year there was no fine from the grid maintaner, so probably it worked well all the time until March this year (or there was not so high surplus).
The scenario is probably more complex, isnt it?
system is running correctly, limiting SE
SE restarts communication
Cerbo looses the information about maximum power capacity (but it was known a minute, 5 mins before)
The solution… for me one from this would probably work:
if you know the maimum power, do not ask for it again (it does not change at all. If it would in the future, there would be a restart anyway)
if you do not know the maximum capacity, you still know maximum for Cerbo. If I am allowed to export 4100 kW, I am fully able to set the limit for SE
ask again in a loop. It happens during night (but with newer SEfirmware it was probably during day too, which is more critical)
There is only really two solutions to this. Either the SolarEdge side should not fall over at this critical point, or our side needs to be more resilient to this, and periodically retry (if limiting was previously working).
I’ll make an issue in our bug tracker to do the latter.