search

masaccio / ha-kingspan-watchman-sensit

Kingspan Connect Sensor integration for Home Assistant
MIT License
11 stars 0 forks source link

Integration keeps rejecting my Kingspan username and password #32

Open Topolino65 opened 1 month ago

Topolino65 commented 1 month ago

Installed the integration and despite several attempts to enter my username and password (which I have checked are correct and also let me login into the Kingspan Connect mobile app), the intergration keeps saying incorrect username and password, so I can't set it up. Any ideas?

masaccio commented 1 month ago

You will need to enable debug for th integration and add the log here (you probably need to delete your email).

hilburn commented 1 month ago

It happens when you haven't got any readings from the tank in your account yet, it fixed itself for me after a day with the first reading

Topolino65 commented 1 month ago

OK thanks. All sorted now. One question though, there seems to be quite a difference in the litres per day consumed (30 day rolling average) in the Kingspan app and what the integration shows for the same 30 day consumption average in home assistant. For example in HA it shows today a figure of 3.2 litres, whereas in the mobile app it shows 3.89. I have changed the HA Integration usage period to consider, to 30 days to mirror the app. Also the integration predicts I wil run out in 231 days but the app says 6th May 2025 which is 196 days from now, so quite a difference. Any idea why it is not pulling the correct data from Kingspan's server, after each update is sent from the transmitter?

masaccio commented 1 month ago

This is a personal choice. I find the Kingspan forecast empty data completely useless in a climate that has very variable weather (the UK). So I fit a linear trend to the consumption numbers in the consumption window configured in the integration. I consider this to be a better representation of when you're likely to run out of oil.

The Kingspan server does provide run-out date and consumption data, but my experience is that it's just garbage data.

masaccio commented 1 month ago

It happens when you haven't got any readings from the tank in your account yet, it fixed itself for me after a day with the first reading

I'll modify the errors in the future to distinguish between a successful login with data, a successful login without data and an unsuccessful login. I'd expected that people using HA would have installed the sensor at least 24 hours before adding it to HA. I guess not!

Topolino65 commented 1 month ago

This is a personal choice. I find the Kingspan forecast empty data completely useless in a climate that has very variable weather (the UK). So I fit a linear trend to the consumption numbers in the consumption window configured in the integration. I consider this to be a better representation of when you're likely to run out of oil.

The Kingspan server does provide run-out date and consumption data, but my experience is that it's just garbage data.

OK thanks for the update. Happy with the integration. just wondered about the differences in 30 day readings between the app and HA that's all !!

hilburn commented 1 month ago

This is a personal choice. I find the Kingspan forecast empty data completely useless in a climate that has very variable weather (the UK). So I fit a linear trend to the consumption numbers in the consumption window configured in the integration. I consider this to be a better representation of when you're likely to run out of oil.

The Kingspan server does provide run-out date and consumption data, but my experience is that it's just garbage data.

It's also just utterly innacurate when you have <30 days of data as instead of comparing the average of 2-31 days vs 1-30 days to get the consumption, it does something... different I currently have data from when I installed it (a bit over a week ago) of: 482, 482, 469, 469, 482, 469, 482, 469, 469, 469, 469 The integration sensibly works out approx 1 l/day from this using the best fit data - about the same as what you get from (first-last)/days (approx 1.2 l/day), but kingspan is currently telling me in excess of 4!

I do wonder whether with enough data you could do am estimate correction for temperature/season - forecasting forward 3+ months is fairly useless when the current consumption is based on autumn and barely putting the heating on vs winter being on significantly more