search

britkat1980 / giv_tcp

TCP connection (from inverter) and MQTT implementation
73 stars 34 forks source link

GivTCP Charge/Discharge Time Remaining appear wrong #77

Closed Tau512 closed 1 year ago

Tau512 commented 1 year ago

I've recently moved to givTCP from givenergy_local HA addon, and migrating my dashboards over. I've noticed that the 'GivTCP Charge Time Remaining' & 'GivTCP Discharge Time Remaining' entity values appear to be wrong.

Specifically these entities in question: sensor.givtcp_ea########_charge_time_remaining sensor.givtcp_ea########_discharge_time_remaining

Around 8:20 my system went from discharge to charging the battery. battery was 15% full, and battery charging rate around 1.3kw, so that should mean approximately 4hours to 100%. image Based on how the graph reports historic values, it appears there's an incorrect calculation somewhere rather than no data.

the related timestamp entities also seem wrong. sensor.givtcp_ea########_charge_completion_time = 2023-07-07T08:08:43+00:00 sensor.givtcp_ea########_discharge_completion_time = 2023-07-07T08:27:24+00:00

The current real time is 09:08 (UK timezone), so it puts the charge time as NOW, and discharge in about 15mins time, even battery is in a charging state.

I should be on the latest & greatest versions of software: Home Assistant 2023.7.1 Supervisor 2023.07.1 Operating System 10.3 Frontend 20230705.1 - latest

giv_tcp 2.2.3

britkat1980 commented 1 year ago

Do you have the timezone set correctly? I'm in UK and as its BST the timestamp should have +01:00 at the end. Perhaps this is part of the problem? Secondly, what is your target SOC set to? It calculates time to charge from current SOC up to target.

Tau512 commented 1 year ago

configuration.yaml didnt have some settings (only country), but now looks like this:

homeassistant:
  ...
  country: GB
  currency: GBP
  time_zone: "Europe/London"
  unit_system: metric
  ...

after a full restart sensor.givtcp_ea########_discharge_completion_time = 2023-07-20T09:58:50+00:00, so much more believable based on the current battery consumption rate even if not BST adjusted. I'll be able to get a better feel for it this evening with sustained draw.

fwiw number.givtcp_ea########_target_soc = 20 (default)

Tau512 commented 1 year ago

it looks like discharge was working as expected last night. however charging sensor still seems wrong. 5.2kw battery, so charge time should somewhere around +3.8hours.

image

since you mention the timestamps should account for BST, any more tips on where to confirm thats setup right? if i've read HA wiki correctly, the time_zone should account for this so nothing else i need to define.

Tau512 commented 1 year ago

still having the issue so figured i'd take a look at code to see if i could figure out how it was calculated.

https://github.com/britkat1980/giv_tcp/blob/a9d357cd739d5e3a8d564031f4ceb4afbbac733a/GivTCP/read.py#L425

my Target_SOC was wrong (number.givtcp_ea########_target_soc); was set to 20%. Changed to 100 and the charge_completion_time is now in line with my own calculations!

Looks like it's been set like this since 8th July which, if memory serves correctly, is when i moved to givtcp; Target_SOC has been 20% since the start.