-
Notifications
You must be signed in to change notification settings - Fork 56
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Wechselrichter Neustart - Stunde fuer Neustart - funktioniert nicht? #300
Comments
Edit: The following is NOT true!This might be easy to explain, thanks to your detailed description: The restart was setup for 0:00 and the DPL calculated a timestamp at which it would reboot the inverter. In your case: midnight the next night. Before that event, you changed the restart time to 8:00. The DPL already had calculated the timestamp for the next restart, so your new setting was overlooked. The DPL will restart your inverter at midnight, then recalculate the next inverter restart timestamp, and the inverter will be restarted again at 8:00 tomorrow, eight hours later. You are asking for the DPL to make sure that a changed setting is respected immediately (without reboot). That's fair. Please observe your setup. If my predictions turn out to be true, then your issue is indeed that the setting is not picked up on change. |
Thanks @schlimmchen you were right! I had changed the restart for 13:00 during my tests, 24hrs later I see the restart happening at 13:00 👍 |
I have the same behavior since I've upgraded to one of the recent releases (344498d). I had the restart set to 04:00 ever since I started using this fork of OpenDTU and it has been working fine. |
Hi,
I assume @Manos1966 your observation was a single fault, due to whatever. |
I did not notice that
Meh. |
thanks for the follow up! I tried to set the restart time to 3:00 and observed the console log within OpenDTU GUI - no PowerLimiterClass::loop entries can be seen and the restart did not happen. The NTP time is synchronized. find below the relevant timeframe from the console log: 02:59:53.660 > PowerMeterClass: TotalPower: 212.00 |
Thanks for providing the log.
The DPL is dormant since the inverter cannot provide However, I see DevInfoSimple commands being sent and two answers being received. What inverter are you using? It is possibly different than the Hoymiles HM Series and may not provide its maximum power output capability. In that case the DPL must deal with that situation instead of insisting on the value to be available. |
@1-am-r00t It seems unlikely that your inverter does not provide |
HM-800
true, the ESP becomes unreachable from time to time and I think I have an explanation on hand. Back when I switched from ahoy to this fork of OpenDTU (about a month ago), I obviously made use of my Smartmeter values that are being sent every second via MQTT. It seems that OpenDTU onbattery is not capable of receiving it in such short interval (the application would become unresponsive and/or unpredictable up until the level where it disconnects from WiFi). I therefor started replicating the smartmeter broadcasting on a separate MQTT topic but with a limiter of max. 1 message per 5 seconds - that seems like the sweetspot for OpenDTU to run stable.
No, the first version I started using was before your changes of refactoring the powerlimiterclass (start/mid of June). However, a couple days ago when I upgraded OpenDTU to 344498d I've changed the MQTT power meter broadcast of messages to be less than the working 5 sec and that seems to have triggered the unpredictness of the application again. I noticed today in the morning that OpenDTU had crashed yet again and the ESP was disconnected from WiFi. I then proceeded and changed the broadcast interval of my smartmeter to 6 seconds and performed a power-reset of OpenDTU - since then it's working stable again. I will continue to monitor the restart behavior today in the night again. |
You should update to 2023.07.09. There was a fix in 23ff4ef "DPL: do not use nullptr when printing debug messages (#303)". But I observed other unintended restarts. 2 restarts in the last 24 hours. At the moment I don't know why. Guess I have to connect a notebook to catch the error in the logs, when it occurs. |
That's weird... My device processes nearly two power meter updates per second from three topics (phases). You seem to subscribe to only one and you suggest that more updates than one every five seconds is too much. Do you use MQTT with authentication? Retain flag enabled? TLS enabled? Home Assistent auto discovery enabled? @helgeerbe is correct, you should update to the most recent release, but I have doubts that this will fix your issue. Is your battery empty at 3 in the morning, when you expect the restart to happen? How many inverters do you watch with this OpenDTU? |
I've made this assumption based on my observations ever since I started using OpenDTU on battery. Since setting the publish interval of that specific topic more than 5 seconds fixed the unresponsiveness almost immediately, this confirmed back then my theory and it was in fact an easy "fix". However now that you mention you are processing two meter updates per second from three topics I have my doubts. I have just one topic subscribed and thats it.
No auth, no retain flags, no TLS, no HA auto disc.
I'm doubting it too, but for what its worth, I can give it a try. I definitely remember this issue was not existing before I flashed to the new version, i.e. the automatic restart was working by the time I initially setup OpenDTU in June.
Actually is not completely empty because in DPL I've specified a low voltage limit of 23.5V which forces the HM to stop producing. The cutoff voltage of my BMS is at around 20.6V.
Only my HM-800 with both channels attached to the battery. |
That's what I meant 😉 Okay, so your battery is not empty, but let's say it is depleted, and the inverter is stopped always or often at the time around which the inverter shall restart? If that's the case, then the nullptr access is probably the reason that your inverter is not restarted: shutdown() is called in every DPL loop that makes it to the calcPowerLimit() at all and triggers the bug every time. The ESP restarts every time before the DPL loop is executed a second time. Please upgrade. |
The battery usually is empty by midnight. I've chosen to set the restart of the inverter to 4:00 because it's a sweetspot where the battery will be definitely "empty" and sunrise is about to happen.
For a strange reason the inverter was no longer being polled at all when its not producing (i.e. lower voltage limit). I proceeded and removed the inverter from the polling list and re-added it using copy/paste of the SN followed by a quick restart. It was still not polling any data at all from the inverter. I then proceeded and manually disabled all polling options in the GUI followed again by a quick restart. It still didnt want to poll any data (like in my logfiles from yesterday). Since I was running out of ideas I then proceeded with the suggested upgrade and guess what; the SN suddenly did not match with what I previously entered/pasted (should be 114174206224, but was in fact something like 114174206116). Once I corrected the SN to the correct one, the polling immediately started working again and also the powerLimiterClass is now showing up in logging:
Honestly I dont know what was going on here, but I guess the error is gone now since the polling is working again. I will report back tomorrow morning/noon if the restart at 4:00 was a success (which I suppose it will be). Thanks for your help/input so far! |
Just a quick follow up: restart worked as intended. Somehow I'm still asking how that thing with the SN happened. I'll wait for the next update and open a new issue in case it'll happen again. Thanks for your contribution/help :) |
This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new discussion or issue for related concerns. |
What happened?
I have the suspicion, the Restart of the Inverter does not happen as planned.
Can somebody test whether this is a bug or not?
Initially I had the Inverter Restart programmed at 00:00
I decided to change it early this morning (at 01:00) to 08:00
My pictures show:
Restart planned for 08:00
Live Daten today (11:30) show a Daily Production of 16,43 kWh which is NOT POSSIBLE
Inverter log shows events as of 00:00 thus no restart happened at 08:00 (events before 08:00 would have been deleted).
To Reproduce Bug
set Inverter Restart at 08:00 and see if it happens.
(I presume that you could try also other times).
Expected Behavior
Restart of the Inverter should happen at the defined time.
Install Method
Pre-Compiled binary from GitHub
What git-hash/version of OpenDTU?
344498d
Relevant log/trace output
No response
Anything else?
No response
The text was updated successfully, but these errors were encountered: