Benni's SH register.yaml V2 #37
Conversation
InfluxDB was returning erros. So I adapted the way of building the query.
Hybrid inverter do not provide run_state. Therefor it need to determined by system_state. In a first step a list of gridonly and hybrid inverters was added to register.yaml setup.py was adapted in order to overwrite run_state by system_state if it is a hybrid inverter. ToDo: In future parsing of the states is nescarry.
There was a problem hiding this comment.
Can we leave 4990 and 5000 as level 3?
5000 is read on startup, and only needs to be read once.
If you want I can add the serial to script start as well. No need to read every cycle.
Why have you added the model lines? If there is no model line, then it is valid for all models. So only need to add that if we want to exclude a model.
|
Hi, SN I think it is good value to read and display the inverter serial number e.g. for the web service, as it helps to identify the devices when viewing the data in browser. The SN is a true unique identifier. That's why I think they should be level two information. And that's why I think that information should be broadcasted with all other measurements. If you change it to level 3 because of coding logic as it is only read once on startup and it does not affect any other feature I am fine with that. But following that logic, we could also move other basic information to that part such as: protocol number and version, arm_software_version, dsp_software_version and so on. On the other hand it would mean that we need a restart if this data changes (upgrading inverter firmware). Long story short: If we really think about splitting data in to "scraped everytime" and "scraped fewer" we need to think about a small and big scraping cycle. I am happy to help with that ;) Model lines I guess you are referring to l 582, l 1580 and so on. According to my Modbus Communication documentation they are not part of SH-Inverters. E.g. register address 5006 is "reserved" same for the others. For SH inverters those addresses are not supported, so I added the whole bundle of non hybrid inverters. I hoped I didn't miss any and I didn't meant to break an existing config. Best [Addition:] I am going to set SN scraping to level 3 under the above named assumption. |
|
level updated |
|
In 13222f6 I have additionally adapted the register names to the standard naming. I have adapted this locally already long time ago. |
|
@benni336 can you have a look at fixing the conflicts? |
| @@ -1140,14 +1141,14 @@ registers: | |||
| accuracy: 0.1 | |||
| unit: "kWh" | |||
| models: ["SH5K-20","SH3K6","SH4K6","SH5K-V13","SH5K-30","SH3K6-30","SH4K6-30","SH5.0RS","SH3.6RS","SH4.6RS","SH6.0RS","SH10RT","SH8.0RT","SH6.0RT","SH5.0RT"] | |||
| - name: "load_power_hybrid" | |||
| - name: "load_power" | |||
There was a problem hiding this comment.
changing the name will break any existing queries/dashboards etc. that rely on it
|
@benni336 could you have a look at this and resolve the conflicts when you get a chance? Trying to get through the backlog of the last year sorry. |
Hi @bohdan-s ,
this is my first pull request according to registers.yaml. You were right, that 90% has been merged.
I don't get why github showed me so many changes.
I want to point out line 217, which I think is a typo. At least to my knowledge and also referring to the Modbus documentation it should be "3P3L"
I also updated file version. I hope you are happy with that. I personally think that versions should not be written into files but only marked in git tags, but I guess that is also some kind of personal taste.
Last but not least I wanted to point out that some of the write registers are not applicable for SH series.
Please comment. I will create a new pull request for some bugfixes within the webserver.