Re: [PATCH 3/5] iio: adc: ad7606: add offset and phase calibration support

[Nuno Sá]

On Wed, 2025-04-30 at 13:33 -0500, David Lechner wrote:
> On 4/30/25 11:14 AM, David Lechner wrote:
> > On 4/30/25 10:36 AM, Nuno Sá wrote:
> > > On Tue, 2025-04-29 at 15:06 +0200, Angelo Dureghello wrote:
> > > > From: Angelo Dureghello <adureghello@xxxxxxxxxxxx>
> > > > 
> > > > 
> 
> ...
> 
> > > > +
> > > > +	val += start_val;
> > > 
> > > Shouldn't this be val -= start_val?
> > > 
> > > I also don't think we have any strict rules in the ABI for units for these
> > > kind
> > > of interfaces so using "raw" values is easier. But FWIW, I think we could
> > > have
> > > this in mv (would naturally depend on scale) 
> > > 
> > > - Nuno Sá
> > > 
> > 
> > From testing, it seems to be working as expected for me, so I think this is
> > correct. The register value is not signed. 0x80 is no offset.
> > 
> 
> Heh, you are actually quite right. Even though it working correctly, it is
> because the value that gets written to the register is val & 0xFF, so adding
> or subtracting here basically has the same effect. But subtracting is the more
> logical way to do it. (I tested it that way too just to be 100% sure.)

Yeps, when testing it i realized that the current form just gives the correct
value in the 2 LSB so I assumed we were doing something to cast way the invalid
bits.

To be more pedantic, I think subtracting is the *correct* way :)

- Nuno Sá