Since End of December I observe the same behavior of my SE-7K. It has now firmware 4.24.22. My last manual update of firmware was before over one, nearly two years. I assume it was an update initiated by SolarEdge themselves. Venus OS is still unchanged the latest official release 3.67. For years the modbus/sunspec communiation was fine. Now it’s broken in the same way like mentioned before by other users. Disconnects and connects every some minutes.
From Ethernet connecting point of view and connection to SolarEdge Monitoring everything seems to be fine.
As fas as I understood till now the only solution is to request an firmware downgrade to 4.22.44 by SolarEdge. Is this still the current situation or should I use the current beta of Venus OS?
There is also a last post from @mblaauw about a script for Node Red. But I don’t really know what he tries to say and what the script will do. Perhaps because of a language mix.
Can somebody suggest, which way I should go? Thank in advance!
”I did not update the firmware manual for this inverter at least for one year (nearly two years), because this installation worked fine and there was no need for an update. Since December 2025 I observe regular disconnects/drops of modbus/sunspec connection every some minutes to the Victron emergency/battery installation and main monitoring of the whole installation. Connection to SolarEdge Monitoring is fine and also the Ethernet connection within the network is fine. It seems that SolarEdge initiated an unrequested firmware update to 4.24.22. This issue is much known for many users of SolarEdge and Victron equipment for nearly one year now. At least the firmware 4.22.44 should work properly. Before your update there must be at least firmware 4.22.39 from November 2024 on my installation. Therefore please downgrade to 4.22.39 or at least 4.22.44. For future: Please also don’t update firmware by yourself. This will help to maintain stability. I use the Internet connection only for monitoring reasons. If there should be security reasons for a necessary update because of internet risks, please inform me before. Should this Modbus/Sunspec connection issue still not be fixed until this moment, nevertheless I will stay with the old firmware and disconnect the inverter completely from the external network (internet) via firewall (deny all) on this site.”
Hello Michel,
I have implemented your script and it is working except the part to transfer the powerdata to the virtual device in VRM.
(I can read the parameters in INV1 and INV2 Read (27 values) but in the Parse INV1 and INV2 I see errors, see pictures.)
Update : The errors were warnings, the Parse nodes are working correctly. Now I have to find out how to implement the virtual device in VRM.
I can see it in the main dashboard of VRM but I don’t know (yet) how to transfer the power data….
I’ve been trying (via my installer) to get SolarEdge to downgrade my firmware for over a month now. They’re simply not responding to our queries…
So I’ve decided, since I have enough room in my electrical cabinet, to just put a EM540 meter after the SolarEdge breaker and connect this with a rs485 - usb cable to the cerbo gx.
Can someone confirm that I just need to remove the current PV inverter on the cerbo, and replace it with a rs485 energy meter, configured as PV inverter?
Everything else (APS mode frequence shifting) will still work?
Ticket was successfully closed around 3.p.m. Downgrade to 4.22.44 was done by SE. I assume it happens around 2:15 p.m. Because of a short down in production during this sunny day. I was requested to observe the behavior. I will observe for some days. But currently it looks fine.
As you can see, the reaction of SE was very fast. At least in my case. I opened a ticket directly via the support ticket from the monitoring page from the installation directly. I got an ID and was able to follow the ticket status. Sure your installer really opened the ticket?
In my case I’m the official installer by myself.
Apparently you can create a case as non-installer as well, so I did so last week, and they performed the downgrade for me today.
I don’t know what my installer did wrong, they did have a case number, and I was CCed on one of the mails they sent to solaredge support, so it’s not like they did nothing.
I will still install the EM540 though. I have it now anyway (but still waiting on a DIN terminal block to wire up the N since the N terminal on the EM540 only takes up to 2.5mm2 wires), and you never know when solaredge might accidentally upgrade the firmware again…
After a downgrade by SE there is no automatic update possible, I had to call support to upgrade to 4.24.22 because customer ordered a SolarEdge One EV Charger. In may opinion its never a good Idea to freeze firmware versions.
As this was the first SE One EV Charger I installed in in my SE -Victron system, and with 4.24.22 the 5 min connection loose was bach which was running perfect with my downgraded SE9k(4.22.44). The problem is not only at SE side but also at Victron side the fix @victron side should be very easy, as a Cerbo reset solves the problem till the next SE reset which is performed 2 times a day, no glue why (that is second root cause of the problem). In my system I run a nodered flow which resets Cerbo after 3 connection losts, as this SE resets always occurs during night I do not loose data. Till now I could not get any feedback from Victron. SolarEdge will do NOTHING , they changed a timing in INIT and Victron can not synchronic ModbusTcp communication.
Had modbus TCP running with no problems before upgrading to 4.24.44
After upgrading to 4.24.44 I experienced intermittent loss of modbusTCP connectivity.
Also the restoration of the service was not structural in reproduction. ( friend GX reboot, TCP/IP disconnect, )
Ultimately rebooting the SE was the only structural reproducible solution for getting ModbusTCP back.. ( for a while ).
I had SE support roll back to 4.22.44 and all problems are gone.
Have not checked if autoupdate broke after this manual intervention, also not sure what the latest SE firmware version is.
But for sure it’s an SE issue !! So everybody please post your software versions which do not work.. or do work.
SE9k SE 4.24.22 & V3.70~83 : after SE restart/reset com lost every ~5 min – Reboot Venus –> o.k (sometimes 2 reboots needed)
SE9k SE 4.22.44 & V3.70~83 : no problems after SE resets (twice a day)
SE9k SE 4.24.22 & V3.62 : no problems after SE resets (twice a day)
The last test means it’s an Victron issue , but I would say its an issue of both , SolarEdge is making a hard reset twice a day (nobody knows why) AND Victron >3.62 can not deal with this resets, and this also can happen if SE is power cycled
update: SE9k SE 4.24.22 & V3.63: after SE restart/reset com lost every ~5 min – Reboot Venus –> o.k (Difference to V3.70~83: node <PV-inverter.status> is not set to “null” every retry, but <PV-inverter.totalPower> does , ~83 .status is set to null everytime the communicationis lost
Until December I used Home Assistant and the excellent SolarEdge Modbus Multi integration from WillCodeforCats, this was rock solid, even when reading every few seconds.
When I installed by Multiplus in December I switched to using the Venus Device integration, but this repeatedly times out, and I cannot monitor what is happening.
As my HA installation works correctly, it must be an issue with the Victron code base.
Can someone please see why the Home Assistant version works and theirs does not, I need to control my feed-in so require Venus to monitor the SolarEdge’s contribution.
I have, and I don’t want to regress my SolarEdge software, as it works with another solution the problem appears to be with the Victron software, not the SolarEdge.
Looks like I will have to create some Node-Red code and Virtual Devices to allow me to tell Venus the output from the SolarEdge via Home Assistant and likewise apply any throttling, which I can do, but don’t feel I should do, as other users seem to be having similar issues and not everyone runs Home Assistant and uses Node-Red.
Unfortunately that is incorrect. There is more to it.
This is an issue with sunspec that they are just unable to address.
Some of us have suggested victron drops sunspec support for them.
There are some creative solutions in the modifications section, including proxies, that have worked for some people.
, but I would say its an issue of both , SolarEdge is making a hard reset twice a day (nobody knows why) AND Victron >3.62 can not deal with this resets, and this also can happen if SE is power cycled