Integration of an external BLE proxy into the Tesla template

[wimaha]

Add the possibility to define ble proxy and use it instead of Tesla Fleet Api to control the vehicle.

vehicles:
  - name: tesla
    type: template
    template: tesla
    accessToken: TOKEN
    refreshToken: TOKEN
    bleHost: IP
    blePort: 8080

The BLE Proxy has to use the same API scheme like the Fleet Api.

• Use BLE Proxy if defined
• Don't fetch current when using BLE Proxy (because of API Limits)
• Build locally and tested the control part
• Updated Docs
• Maybe: Add more explanation in docs ?

Helps for #14252 and #14226

[FraBoCH]

I would be cautious with that as, IMHO, it will bring back issue #13007 where going below 5A is ignored by the Tesla the first time. This feedback loop allows evcc to detect the current mismatch and send the command again.

[wimaha]

We don't have this issue with ble. I've used it for multiples days now. And I always charge with less then 5A.

[FraBoCH]

Ok, good to read.

I’m also using your ble proxy with TeslaMate and I did face this issue however. Maybe I’m a special case as I also charge a non-Tesla vehicle on my TWC3 ;-)

Anyway I fixed it by fetching the actual current from TeslaMate. I added that in my custom vehicle definition and then evcc started to detect and automatically fix the current mismatch:

GetMaxCurrent:
    source: mqtt
    topic: teslamate/cars/2/charger_actual_current
    timeout: 8h

[wimaha]

That's annoying. However, if we query the Tesla server every time there is a change in charging speed, we will hit the API limit. Therefore, I think this query must be removed if we want to avoid the API limits.

[FraBoCH]

It’s funny, I proposed yesterday almost the same PR (#14616) as I wanted to push progress on that topic. @andig you can really chose the one you prefer, no feelings on my side, just happy to see things progressing !

[wimaha]

I close this, because I think #14616 is more advanced.