Re: [PATCH v2 2/4] ASoC: dt-bindings: realtek, rt5682s: Add AVDD and MICVDD supplies

[Chen-Yu Tsai]

On Thu, Oct 27, 2022 at 12:20 PM Nícolas F. R. A. Prado
<nfraprado@xxxxxxxxxxxxx> wrote:
>
> On Thu, Oct 27, 2022 at 11:11:08AM -0700, Chen-Yu Tsai wrote:
> > On Thu, Oct 27, 2022 at 10:48 AM Chen-Yu Tsai <wenst@xxxxxxxxxxxx> wrote:
> > >
> > > On Thu, Oct 27, 2022 at 7:36 AM Nícolas F. R. A. Prado
> > > <nfraprado@xxxxxxxxxxxxx> wrote:
> > > >
> > > > On Tue, Oct 25, 2022 at 01:12:49PM -0700, Chen-Yu Tsai wrote:
> > > > > On Mon, Oct 24, 2022 at 3:01 PM Nícolas F. R. A. Prado
> > > > > <nfraprado@xxxxxxxxxxxxx> wrote:
> > > > > >
> > > > > > The rt5682s codec can have two supplies: AVDD and MICVDD. They are
> > > > >
> > > > > The actual chip also has LDO1_IN (for digital core and charge pump)
> > > > > and DBVDD (for I/O). However in the Chromebook designs these two
> > > > > and AVDD are all provided from the same power rail, through separate
> > > > > filter banks.
> > > >
> > > > What about rt5682 (no s), does that chip also have these same supplies?
> >
> > (Missed this question)
> >
> > The RT5682 has the same supplies, plus the VBAT one.
> >
> > ChenYu
> >
> > > > Also, since you already gave the purpose of these other supplies, could you also
> > > > tell the purpose of AVDD, MICVDD and (for rt5682) VBAT? That way I could add
> > > > some description for them in the binding.
> > >
> > > As Mark mentioned in his reply, these are quite standard names.
> > >
> > > AVDD is for the analog bits. MICVDD is for the microphone bias.
> > > VBAT is called battery power in the datasheet. The block diagram
> > > shows it going through an internal controllable LDO whose output
> > > then powers MICVDD. This could be used in designs that don't
> > > include a suitable external supply for MICVDD. If MICVDD is provided,
> > > then one would turn the internal LDO off.
> > >
> > > So either VBAT or MICVDD has to be provided.
>
> So if I get this right we should be making the following supplies required:
> AVDD
> VBAT or MICVDD (on rt5682s there's no VBAT so just MICVDD)

That's right.

> And for LDO1_ON and DBVDD what would be the best way to model? Should we make
> them required in the binding as well and add them pointing to the same regulator
> from AVDD in the chromebook DT or just leave them as optional?

They should be required. And we then describe them to match the board.

ChenYu

> Thanks,
> Nícolas
>
> > >
> > > ChenYu
> > >
> > > > Thanks,
> > > > Nícolas
> > > >
> > > > >
> > > > > Neither does the datasheet specify the ordering of AVDD, DBVDD, and
> > > > > LDO1_IN for power sequencing, just that three should be toggled together.
> > > > >
> > > > > Should we model these? Or wait until some design actually splits these?
> > > > [..]