SolarEdge inverter restarting once or twice per day?

[erikgeurts]

In my logs, I've found log entries like this:

Type: Error
Logger: custom_components.solaredge_modbus_multi
Source: helpers/update_coordinator.py:506
Integration: SolarEdge Modbus Multi (documentation, issues)
First occurred: 26 July 2026 at 23:33:28 (2 occurrences)
Last logged: 27 July 2026 at 23:35:26

Error fetching SolarEdge Coordinator data: Modbus/TCP connect to 192.168.86.104:1502 failed.

It appears that my inverter does a kind of warm reboot every day, and sometimes even twice a day. These are just 2 entries in my logs right now, because I did a HA restart 2 days ago, but it happens every day.

At/around the same time of the log entry, the status text of the inverter has this kind of sequence: production -> unknown (at exactly the time of this log entry, lasts 44 seconds) -> production (11 seconds) -> Fault (37 seconds) -> production.

I know about these states because I've created a template sensor that reads the status text attribute of the status sensor in the integration: {{ state_attr('sensor.se2025_i1_status','status_text') }}.

I have a kWh meter that measures power in/out of the inverter to my electrical installation, and at the time of these log entries, I can see that the inverter stops providing power for the house from the batteries and instead draws a little power (I assume simply its own power consumption for the internal hardware).

Is this known behavior for SolarEdge inverters? Mine is a SE10K, currently with 0004.0025.0011 firmware. I also have Home Battery 48V - 2 modules and an inline energy meter.

Curious to learn if anyone in the community has also seen this.

[Robert-2669]

Hi Erik,

I also see round the same time a similair kind of error. Just cleard the logs, so tomorrow I can confirm if the same error.

SE5K / No batteries or inline meter / 0004.0024.0022

[WillCodeForCats]

Yes, all SolarEdge inverters do a restart/reset once every 24 hours. It will probably be sometime at night during non-production time. Mine reboot every night around 9:56PM.

[Robert-2669]

Hi Seth,

If it's the result of a standard Inverter operation, shouldn't be classified as Warning instead of Error?
Just a thought.

[UrbanOlsson]

FWIW My SE17K from 2019 restarts every 25 hours and I know for sure that I'm not alone having this annoying 25 hours cycle. Others with models from 2016-2020 has reported the same 25 hours cycle, not 24 hours.

[WillCodeForCats]

Coordinator failed to update is a HA "error" level, so that's coming out of HA rather than the integration.

What you can do here is increase the retry timing before it fails into error/unavailable state. I am going to end up increasing the retry timing for the 4.0.0 release, but you can do it now with some YAML settings in configuration.yaml:

solaredge_modbus_multi:
  retry:
    time: 900

The default base value for retry time is 800 milliseconds. This increases it to 900. If you're still seeing unavailable/error every night during the reset, try increasing it to 950 or 1000 milliseconds. By default it tries 5 times and multiplies the retry each time, so it's an increasing backoff to retry the inverter if it's busy (or rebooting itself).

In 4.0.0 the new default will be at least 900 milliseconds or higher based on what I'm observing with the new modbus-connection library with my inverters. Because this is a once a day kind of error to actually test and observe I have to make the adjustments on my system then wait to see it happen.

[erikgeurts]

Yes, all SolarEdge inverters do a restart/reset once every 24 hours. It will probably be sometime at night during non-production time. Mine reboot every night around 9:56PM.

OK, interesting, good to know. Since I also have batteries, there isn't really any non-production time for me. Mine tends to do it around 23:45 every night, it appears to have figured out what time the last person in the household usually goes to bed.