Yes that is is excatly true. It fetches the data directly form the device. You can configure the poll interval. The default is 35s.
I also want to use my new HMS-800W-2T completely local, like @e-raser - and I would like to share my experiences:
I connected my HMS-800W-2T to a sub-net that is not connected to the internet at all. So the inverter is connected to my local network, but it does not have any internet access.
I am using GitHub - suaveolent/ha-hoymiles-wifi: Home Assistant custom component for Hoymiles DTUs and the HMS-XXXXW-2T microinverters
At first I chose the default update interval of 35 seconds.
I quickly noticed that the inverter seemed to loose its connection every few minutes, maybe due to the fact that it cannot reach its cloud server. (see image below)
As a wrong update interval seem to prevent the inverter from communication with its cloud I decided to change the update interval a bit:
At 13:30 I change the update interval to 15 seconds. Since that I get continuous valid values from the inverter. (again, see image below)
(It was a short cloudy winter day, without any sun…)
DTU is still on HW-version H00.01.00 and SW-version V00.01.05.
Inverter reports HW-version H00.04.00 and SW-version V01.00.08.
I do not dare to update, I am afraid it might brake the way I grab data from the inverter without any cloud or internet access.
I am keen to get experiences from other people, especially with current firmware versions.
Also I would like to find more information about the meaning to the error codes. I did not find anything on google yet. My error codes are:
- port 1: 50921472 = 0x3090000
- port 2: 50331648 = 0x3000000
Maybe they indicate that there is no internet connection?
Hi,
I am using the same setup from suaveolent with an HMS-800W-2T and it works fine with a 35 sec. update interval. I also use the Hoymiles cloud apart.
I also get the same errorcodes and cannot find any information - every day the same codes at these ports.
- port 1: 50921472
- port 2: 50331648
→ would be nice to know what it means. THANK YOU.
Inverter:
HW-Version H00.04.00
SW-Version V01.02.01
DTU:
HW-Version H00.01.00
SW-Version V00.01.11
I am still trying to find out what those mean. I did not yet find anything useful in the application but this on the todo list.
Hi,
my Inverter:
Model: HMS-800W-2T (Gen3) | Grid Profile: 2.0.0 (DE_VDE4105_2018)
Hardware Version: H00.04.00 | Firmware Version: V01.00.08
my DTU:
Hardware Version: H00.01.00 | Firmware Version: V00.01.11
I think, i recognize the same (much more serious) error in your graph, that i see in my inverter.
Since the “Port 1 error code 50921472” show up in the HA add-on,
my inverter has been capped at 19.x W (both ports together).
Today I experimented a little:
If I leave HA active, the power curve looks like the one on the left.
Max 10W - all day, everyday.
Today around 12:00 I initially only restarted the DTU via the S-Miles Cloud website.
That didn’t help and A few minutes later I restarted the inverter.
Suddenly the port then supplied over 250W instead of 10W!
After 1-2 days, the inverter will hang again. (hoymiles cloud shows no errors, btw.)
I will now leave the addon switched off for a few days and see if the inverter runs fine again.
Hi @abecedarian, have you ever tried it again with the zero-feed-in in a newer version? I also want to realise the “Nulleinspeisung” and started my whole HA project with this reason 
Thx
First of all, thank you @abecedarian for this component. I have been using it for a few weeks and it is helping me a lot. I have even used the graphs to point out a heating problem related to microinverter installation.
Regarding codes, same here with the HMS-2000DW-4T. Only port 1 gets a different code. It does not seem to affect the generation, though.
Any updates on the meaning of it?
-
Has anyone noticed any problems with the S-Miles app related to data that is sent to the cloud?
I have been monitoring 3 microinverters with both S-Miles and HA. Both of them do not present any errors/alarms, but S-Miles seems to loose some data points, showing zero energy from some microinverters at some periods during the day. This even affects the amount of daily energy generated reported by S-Miles, which ends up being smaller than the one reported by HA.
I was wondering if the integration might be affecting somehow the data report to the Hoymiles cloud. I have it configured to pool data every 60 seconds. -
Another issue: does anyone else have this problem related to the reset of daily sensors at midnight? Not sure if this is a microinverter timesync misconfiguration or just some delay in HA processing, but it affects my history graph, since I am considering the max value of these accumulators each day. I get same value for two subsequent days since the previous value “leaks” to the new day.
There might be some different way to plot this graph, maybe use a specific time each day instead of max value, but the whole HA thing is new to me. Any
tip is welcome.