Re: [PATCH] iio: potentiometer: add driver for Maxim Integrated DS1807

[Jonathan Cameron <jic23@kernel.org>]

On Sat, 12 May 2018 11:50:23 +0200
Michael Nazzareno Trimarchi <michael@amarulasolutions.com> wrote:

> Hi
> 
> On Sat, May 12, 2018 at 11:45 AM, Jonathan Cameron <jic23@kernel.org> wrote:
> > On Wed, 9 May 2018 11:19:51 +0200
> > Michael Nazzareno Trimarchi <michael@amarulasolutions.com> wrote:
> >  
> >> Hi Jonathan
> >>
> >>
> >> On Wed, May 9, 2018 at 11:01 AM, Michael Nazzareno Trimarchi
> >> <michael@amarulasolutions.com> wrote:  
> >> > Hi
> >> >
> >> > On Mon, May 7, 2018 at 7:17 PM, Jonathan Cameron <jic23@kernel.org> wrote:  
> >> >> On Mon, 7 May 2018 18:55:16 +0200
> >> >> Lars-Peter Clausen <lars@metafoo.de> wrote:
> >> >>  
> >> >>> On 05/07/2018 06:44 PM, Lars-Peter Clausen wrote:  
> >> >>> > On 05/06/2018 07:37 PM, Jonathan Cameron wrote:  
> >> >>> >> On Sun,  6 May 2018 15:30:47 +0200
> >> >>> >> Michael Trimarchi <michael@amarulasolutions.com> wrote:
> >> >>> >>  
> >> >>> >>> The following functions are supported:
> >> >>> >>> - write, read potentiometer value
> >> >>> >>>
> >> >>> >>> Value are exported in DBm because it's used as an audio
> >> >>> >>> attenuator  
> >> >>> >>
> >> >>> >> This is interesting.  The problem is that there is no way for
> >> >>> >> userspace to know that it is reporting in DBm rather than
> >> >>> >> reporting a linear gain or a straight forward resistance.
> >> >>> >>
> >> >>> >> This is rather closer in operation to the analog front end
> >> >>> >> driver I took the other day than to the other potentiometers
> >> >>> >> we have drivers for.
> >> >>> >>
> >> >>> >> Anyhow, how to solve this?  Two options come to mind.
> >> >>> >> 1) Look up table mapping to linear gain as per current ABI
> >> >>> >> 2) Add a new channel type to represent the fact we are
> >> >>> >> looking at a logarithmic value, letting us handle it as DB.  
> >> >>> >
> >> >>> > Yeah, I guess it is a bit difficult. I don't think this should be a separate
> >> >>> > type since we are still describing the same thing, just the scale is
> >> >>> > logarithmic rather than linear. Translation table doesn't work either since
> >> >>> > your values would get ridiculous small/large. We could add a db suffix to
> >> >>> > the type, but that's just terrible. I guess the best we can do is have a
> >> >>> > scale attribute that says 1dB.  
> >> >>>
> >> >>> The other problem of course is that dB is a relative unit. The ratio of one
> >> >>> value to another. Whereas our normal scale refers to an absolute value.  
> >> >> I'm really not keen on this.  We have done the separate types
> >> >> for humidity already, where we had relative (which is a ratio) and absolute
> >> >> (which isn't).  It's not pretty though.
> >> >>
> >> >> Potentially we could define a new attribute that says this one is
> >> >> is db or linear but that's ugly too.
> >> >>
> >> >> As you asked, are we looking at a part that gets used for anything other
> >> >> than audio or not?  If just audio, alsa driver does indeed make more sense.
> >> >>  
> >> >
> >> > This can be used in audio but even in other field. It's just a potentiometer.
> >> > Can I know what is wrong to use the same approch of audio ampliefier that we
> >> > have already in the iio tree?
> >> >  
> >>
> >> cat /sys/bus/iio/devices/iio:device1/out_altvoltage0_hardwaregain
> >> -10.000000 dB
> >> echo -10 > /sys/bus/iio/devices/iio:device1/out_altvoltage0_hardwaregain  
> >
> > Wow, somehow that entire thing had slipped my mind.  I guess we went
> > through the whole question of how to support dB scales years ago
> > and it has just been very little used.
> >
> > Hmm. Sorry for my absent mindedness! Anyhow, there are a few additional
> > comments that need cleaning up.
> >
> > It is going to be a little odd as the only potentiometer (I think) that
> > is acting as a scale free attenuator, rather that being controlled on the basis
> > of resistance, but for the part that seems to make sense so fair enough.
> >
> > I'm slightly curious to know what you have this wired up to though?  
> 
> I'm design GIOTTO ;) an audio module that use those to control the
> volume. It's a dsd
> native sound card that demultiplex i2s to L and R dsd on a pcm1795a.
> Everything already
> run under linux. The idea is to create an audio card and connect iio
> device to the volume
> to change dsd volume
> 
> > Are the inputs and outputs invisible to the kernel or is this feeding
> > into another device?  
> 
> I think a reply above. Anyway we don't want to have driver duplication
> and I think should be land
> there
> 
> >
> > If we are feeding another device then the work recently done on a
> > generic AFE driver may be useful.  At somepoint we'll need a version  
> 
> Can you point to it? I need to read about ;)

https://patchwork.kernel.org/patch/10358131/

Should be in linux-next by now ready for the next merge window.
As it turns out, probably not relevant in your case you will
probably want to have the sound card as a consumer so that
the volume control maps nicely via usual interface etc.

Perhaps cc the relevant sound lists and maintainers on next version.
I don't want to tread on anyone's toes if they are of the view that
it shouldn't be done this way (should be fine from previous conversations
with a few of them!)

Jonathan

> 
> Michael
> 
> > of that which deals with standalone amplifiers and attenuators anyway,
> > but I don't know if it is useful to you.
> >
> > Jonathan
> >
> > Jonathan
> >
> >  
> >>
> >> uname -a
> >> Linux linaro-alip 4.4.93 #7 SMP Sun May 6 13:23:08 CEST 2018 armv7l GNU/Linux
> >>
> >> Michael
> >>  
> >> > Michael
> >> >  
> >> >> Jonathan  
> >> >
> >> >
> >> >
> >> > --
> >> > | Michael Nazzareno Trimarchi                     Amarula Solutions BV |
> >> > | COO  -  Founder                                      Cruquiuskade 47 |
> >> > | +31(0)851119172                                 Amsterdam 1018 AM NL |
> >> > |                  [`as] http://www.amarulasolutions.com               |  
> >>
> >>
> >>  
> >  
> 
> 
>