Closed verisgit closed 2 years ago
see if the new commit (https://github.com/basking-in-the-sun2000/solar-logger/commit/b5743971ce9543e4c7fd5e4d3af6d168de47d3ef) solve your issues
Import the dashboards as described below rather than importing the json code https://github.com/basking-in-the-sun2000/solar-logger#grafana-files
Thanks trying to set this up now. A few questions regarding units.
tzone - your local timezone (what is the timezone format?) solcast_adj - if you need to adjust the forecast, 1.0 doesn't change it (what is the unit? is this a factor i.e. 1.2*x) peak_I - the maximum current the inverter produces (is this in kW? i.e. is a Sun2000-5KTL-L1 5kW Inverter = 5?
Edit: Only timezone I could get to load was "Australia/Perth" - UTC+8, UTC+08:00, etc. didn't work. So I assume "Australia/Perth" is correct; however, solcast is still 8 hours out.
These other dashboards look really cool by the way! Didn't realise you had these.
Here's what the latest dashboard looks like with the changes.
tzone - your local timezone (what is the timezone format?)
Should follow the same format as the one described in the config.py fil
That is bewildering. All the db calls in grafana have a timezone clause. I will need to look further into this.
What value do you have for time_zone in the config.py file?
solcast_adj - if you need to adjust the forecast, 1.0 doesn't change it (what is the unit? is this a factor i.e. 1.2*x)
It is just a factor that multiplies the solcast forecast for display only
peak_I - the maximum current the inverter produces (is this in kW? i.e. is a Sun2000-5KTL-L1 5kW Inverter = 5?
Current in Amperes. Should be in the datasheet for your inverter. The inverter injects current onto the grid, and there is a maximum value. This current can produce work based on the voltage, so your 5kW are really 5kVA. Since you entered 5, you have a line in the power panel around 1.2kW
If your voltage is low you won't reach 5kVA, and if the voltage is high you will surpass this "limit". It uses peak_I to draw a max line in the power panel. This to show you how close of the inverter maximum you are reaching, and it will be the point at which the inverter clips the output
Looking at your inverter, Huawei might be limiting the current to produce a fixed power?!?! Mine only produces 10 kVA at 240V and less at lower voltages. Supposedly the maximum current for yours is 22.7A
config.py = time_zone = "Australia/Perth" from https://en.wikipedia.org/wiki/List_of_tz_database_time_zones I think Huawei limit the inverters to 5kw as that's the max you are allowed to export here.
I had a play around with the influx queries in Grafana, weird thing is the "tz" property in Grafana didn't seem to be doing anything even if I made all the datasets different timezones for testing if they would align (or to see if one was wrong).
this made me wonder what's in the actual database and what could cause them to be offset by this amount. I did find that my RPI OS timezone was set to UTC time. I have changed it to Australia/Perth. But not sure if that's used anywhere or what the impact would be of suddenly changing it. Hopefully, it doesn't start writing the rest of the data offset.
the timezone clause allows queries to be aligned to your time. This prevents getting multiple values for what should be only one, since you day is probably split among two different days in relation to UTC
Hope your setting the rpi time solved your issues
After I set the correct time on the RPI via raspi-config it seems to have fixed it. Thanks for the help with that.
Hi there,
Thanks for the updates earlier, I've checked the Grafana settings and Grafana itself is set to the correct timezone.
The dashboard variable is set to "Australia/Perth" but I wasn't sure what format it's supposed to be in? Australia/Perth, GMT+8, UTC+08:00, etc. I tried a few and some caused errors but none moved the solcast forecast to the correct spot.
I also changed the other occurrences in the Dashboard JSON from UTC to Australia/Perth but solcast is still in the same spot. I'm wondering if it's wrong in the database but I assume all of that is stored in UTC so offsets can always be applied to correct it?
The main problem is for some reason the Solcast data is offset by 8hrs. So the charts aren't lined up properly. It's like everything is using the timezone value except for solcast.