Re: [PATCH 1/2] power: supply: ab8500: Respect charge_restart_voltage_uv

From: Linus Walleij <linus.walleij@linaro.org>
To: "Vaittinen, Matti" <Matti.Vaittinen@fi.rohmeurope.com>
Cc: Sebastian Reichel <sre@kernel.org>,
	Marcus Cooper <codekipper@gmail.com>,
	"linux-pm@vger.kernel.org" <linux-pm@vger.kernel.org>
Subject: Re: [PATCH 1/2] power: supply: ab8500: Respect charge_restart_voltage_uv
Date: Thu, 5 May 2022 22:17:09 +0200	[thread overview]
Message-ID: <CACRpkdYCWYgUAUJVVKy0jU4yF=d95e0Ead=coeMaA=uD3WV35Q@mail.gmail.com> (raw)
In-Reply-To: <a1ad97a5-4048-1f8e-272b-6d1e99bd20cf@fi.rohmeurope.com>

Hi Matti,

sorry for slow replies!

On Tue, Apr 19, 2022 at 10:51 AM Vaittinen, Matti
<Matti.Vaittinen@fi.rohmeurope.com> wrote:

> Hi again Linus,
>
> For some reason I am always slightly terrified when charging is
> controlled by the software ;)

That is a normal reaction... The AB8500 has a few hardware safeguards
to make sure it is somewhat safe. Such as multiple thermal sensors
and over-voltage protection. But it also has a pretty elaborate state
machine.

> On 4/15/22 23:36, Linus Walleij wrote:

> > +     /* Some batteries tell us at which voltage we should restart charging */
>
> Is utilizing this limit something that has already existed for these
> batteries?

Yes, look for example at the Kyle battery, Samsung SDI EB425161LA:
https://github.com/linusw/u8500/blob/Samsung-SGH-I407-Kyle/arch/arm/mach-ux500/board-kyle-bm.c
line 312 called "recharge_vol" is set to 4.3 V
Then in the charging algorithm:
https://github.com/linusw/u8500/blob/Samsung-SGH-I407-Kyle/drivers/power/ab8500_chargalg.c
line 2229 you can see how it is used the same way.

> I am just slightly worrying if this can cause problems at low
> temperatures? I am by no means an expert on this area (as I've told
> earlier :]) so I may be completely off here. Anyways, I think I've seen
> voltage curves for batteries at different temparetures - and AFAIR, the
> voltage of a battery with near 100% SOC at -20  Ccan be close to the
> voltage of a nearly depleted battery at +40 C.

Probably true, because the batteries have operational conditions for
low/high temperatures which change the behaviour of the charging,
so that is how they choose to deal with this.

> Hence I am just asking if this is not causing my phone to keep charging
> even when the battery is full. I mean, when I am at next autumn spending
> the night in a tent at Enontekiö Finland - and forget to disconnect my
> phone from charger before the campfire fades away? :]

The battery in my example EB425161LA is defined in
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/power/supply/samsung-sdi-battery.c
lines 677-719, there you find .temp_alert_min = 0, and if you look in
the charging algorithm:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/power/supply/ab8500_chargalg.c
below that temperature the charging algorithm will switch to a lower recharging
current by going into the state TEMP_UNDEROVER where this
recharging voltage does not apply, instead a much lower voltage
alert_low_temp_charge_voltage_uv = 4000000 will apply, 4.0V
instead of 4.3V.

Further you see that temp_min = -30, so if the temperature goes
below -30 degrees, the algorithm will shut down charging altogether.

Yours,
Linus Walleij