Possible to automatically disable fast charging at night?

[vertigo]

I also have a WiFi switch...in the car

How does that work? Do you mean a WiFi car charger? Does it use WiFi direct?

Also, I checked WiFi plugs, but they're almost all cheap junk from random Chinese companies. The only semi-reputable ones are saw are from TP-Link, but those ones are pretty large. Any recommendations? And how do you control them with AM? Do they require their own app which you control by sending broadcasts?

[Wibbly]

I created an solution myself using an ESP8266 based board. It creates a WiFi AP in the car (no internet), and receives http requests via AM to turn it's relay on or off depending on the phone's battery level. The relay controls power between the cigarette lighter socket and the phone. It less elegantly does what this does:

https://chargie.org/

The Chargie is controlled by Bluetooth, not Wifi and so far offers no API that AM can use. I have not tried a Chargie myself. My guess is this would be a good generic solution which avoids "hacking" but it's outside of AM. Their app does pretty much what my AM solution does and can be used in a car or at home (assuming it works)

At home I use a Wifi plug from Sonoff, and have used one of the many alternative firmwares for it and I also control it directly from local http requests from AM. So it does not use the manufacturer's app to control it any more, nor connect to the internet/a cloud service to be able to work.

Although my phone supports fast charging, I don't use a fast charger at all, as I'm never in that much of a hurry to charge it, and a normal charger is kinder to the battery. If you need to use a fast charger, you should check if a solution like any of those above would work with a fast charger.

[vertigo]

What's the difficulty level on that? Is soldering or programming required, or is it just a matter of putting the chip in a box and connecting a power supply? At the very least, you must have to connect to it somehow to give it an SSID. I'm interested in something like this, both for the charging and for a better way to let my phone know I'm in the car, since neither BT nor activity recognition are very reliable.

[Wibbly]

> for a better way to let my phone know I'm in the car, since neither BT nor activity recognition are very reliable.
Bluetooth is very reliable for me.

Basically, the in-car charging solution uses a bare board which has a header you can use to be able to flash it. I edited some code someone gave to me.
https://www.itead.cc/wiki/Sonoff_SV

For the Sonoff plug, you need to to solder in order to get connections for the header to be able to flash it. There are many IOT firmwares to choose from that will support HTTP requests.

https://esphome.io/devices/sonoff_s20.html
https://github.com/arendst/Sonoff-Tasmo ... art-Socket
https://sites.google.com/site/mppsuite/ ... 66download

In both cases lots of research to be done!

[Desmanto]

@Micky @vertigo : I recommend sonoff, it is based on ESP8266. You need to tinker it a while, but it is fun. No soldering needed, you just need to buy the cable and the plug. Since you can flash Tasmota and now you can have full http/MQTT control. We can use htttp request to trigger on/off. Or if you buy the POW version, you can use periodic http request to query the power usage. The flow is in my to do list. I can then turn the combination of both to monitor power usage of any electric appliances with my phone. Currently I only use it to track my monthly PC power usage. I have the sonoff basic to control standing fan, which is actually only 1 meter from me (too lazy to reach it ). I just use keyboard shortcut in my PC to turn it on/off when needed.


@vertigo @Wibbly : The main enemy of the battery is heat. From my experience, charging it to full and kept full for long time is not as bad as stated in that website. Maybe because the newer battery tech and charging has all kinds of mechanism to protect the battery. My RN5 whyred usually charged several times a day, from around 50-70% to 100% and sometimes kept in 100% plugged in to PC while scrcpy mirroring. It is 2+ years already, maximum SOT as I tested maybe weeks ago, still can achieve 10+ hours (compared to max 15 hours at the first several months, average 12 hours).

Using quick charger (QC 2.0/3.0) surely will create more heat and mine has charging throttling at around 70% if the phone is above 43 C. I use automagic flow to test this several times in 2018, and still can replicate the result early this year. I planned to post my finding at that time, but already lost the mood to do that, since I already gain the knowledge.

There should be a balance between convinient and maximizing battery life. I typically charge whenever it is convinient, using PC or normal slow charger, doesn't care if the battery is close to full, just left it there fully charged. But when travelling, where quick charging is important, then I prefer to bring my QC charger.

[Wibbly]

> still can achieve 10+ hours (compared to max 15 hours at the first several months, average 12 hours).

So about 20% battery capacity loss, right?

[Desmanto]

Yes, that's about after 2 years, which I think is quite impressive. Consider that some of the users in our local forum already changed their batteries.

[vertigo]

@Desmanto

My battery has only lost ~6% after 3.5 years. So while that's only one case, and does not make a trend, I would argue that it definitely works in favor of not fully charging. And while I used slow-charging much more than fast-charging for much of the first 2.5 years, this past year has been probably 80-90% fast-charging, and still very little loss.

Either way, there's no doubt heat is bad, and slow-charging is better, though I wonder just how significant that is. And I definitely believe, based on everything I've read there and elsewhere about batteries and my own experience, that charging to full is quite bad.

[Desmanto]

@vertigo : ~6% only? Do you measure the time used from 100% to 1%? Because when I have the chance to do so, I will purposely drain from 100% until close to 0% to get the most accurate measurement. If the data is only from interpolation, let say you use from 75% - 50% and get 3 hours SOT, then interpolated to become 12 hours; that is not as accurate as 100% used to 1%.

I also want to maximize the battery life. But I still want convinient. As the battery university says that the best cycle only if you keep your battery at 65-75%, which is almost close to impossible to keep on. I would trade some battery aging with more convinient charging cycle.

[vertigo]

Mainly based off of Accubattery, which monitors how much of the battery is discharged in % and charged in mAh over time to calculate the capacity. Also, when I'm using it, it loses very close to 10%/h, even when driving, running GPS, and playing music over BT. So I get roughly 10h of SOT like that (can get a bit more if doing less intensive stuff, but that's a pretty good baseline), and that hasn't really changed enough to even notice in the past few years.