Re: [PATCH] iio: humidity: hdc100x: fix IIO_TEMP channel reporting

[Alison Schofield <amsfield22@gmail.com>]

On Sun, May 29, 2016 at 07:46:47PM +0100, Jonathan Cameron wrote:
> On 27/05/16 03:55, Matt Ranostay wrote:
> > IIO_TEMP channel was being incorrectly reported back as Celsius when it
> > should have been milliCelsius.
> > 
> > Signed-off-by: Matt Ranostay <mranostay@gmail.com>
> Applied to fixes-togreg-post-rc1 and marked for stable.
> 
> Thanks,
> 
> Jonathan


Hi Matt, I tried testing this one out and it looks good, but led me
to more questions ;)

(fyi: this is one of the first hw sensors I got, and as you can guess
by the other work, we didn't realize it already had a driver, hence,
using it as a learning experience.) 

Summary of this long msg is:
- this fix correcting the temp scale works fine 
- questioning what driver is providing as "raw" data for both temp & humid
- humidity calc seems off

Here's my path: I read the raw temp, look at the data sheet calculation,
and expect to be able to plug the raw reading sysfs provides into that
equation and come up with the actual temp.  I can't do that because the 
raw data the driver is provided is not so 'raw'.  The driver is stripping
away the extraneous bits and providing 14bits of actual temp data from the
16bit temp register as 'raw'. I get that and can see how that might be the
right thing to do, but not so sure, since I just got bit by it.  Should the
driver be providing the real raw register data and doing all the munging via
scale and offset. Do some users want to get that raw number, look at the
datasheet, and do the work themselves? 

The temp calcs seem close enough:
16 bit raw: 24929, 14 bit raw: 6232

$ cat in_temp_raw in_temp_offset in_temp_scale
6232
-3971.879096
10.070800781

Grabbed both - 16 bit raw: 24929, 14 bit raw: 6232

14Bit way using scale & offset:
(6232 - 3971.879096) * 10.070800781 / 1000 = 22.761227365

16Bit way using datasheet equation:
(24929/65536 * 165) -40 = 22.763748163


Now, onto the humidity - not close, am I doing this right?
Since my trusty Bionaire humidity sensor says 32% I'm
thinking the datasheet calc is correct.

$ cat in_humidityrelative_raw in_humidityrelative_scale
5075
0.010000

Again, grabbed both - 16 bit raw: 20303, 14 bit raw: 5075

14Bit way using scale:
5075 * 0.01 = 50.75

16Bit way using datasheet equation:
20303/65536 * 100% = 30.98

That's all I have.  Let me know if I can do anything more with it.
Thanks,
alisons

> > ---
> >  drivers/iio/humidity/hdc100x.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/iio/humidity/hdc100x.c b/drivers/iio/humidity/hdc100x.c
> > index fa47676..7e73c56 100644
> > --- a/drivers/iio/humidity/hdc100x.c
> > +++ b/drivers/iio/humidity/hdc100x.c
> > @@ -211,7 +211,7 @@ static int hdc100x_read_raw(struct iio_dev *indio_dev,
> >  		return IIO_VAL_INT_PLUS_MICRO;
> >  	case IIO_CHAN_INFO_SCALE:
> >  		if (chan->type == IIO_TEMP) {
> > -			*val = 165;
> > +			*val = 165000;
> >  			*val2 = 65536 >> 2;
> >  			return IIO_VAL_FRACTIONAL;
> >  		} else {
> > 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-iio" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html