Re: [PATCH v11 1/2] dt-bindings: adc: add AD7173

[Jonathan Cameron <Jonathan.Cameron@xxxxxxxxxx>]

On Thu, 18 Jan 2024 10:10:46 +0200
Ceclan Dumitru <mitrutzceclan@xxxxxxxxx> wrote:

> On 1/17/24 18:37, Jonathan Cameron wrote:
> > On Wed, 17 Jan 2024 14:43:21 +0200
> > Ceclan Dumitru <mitrutzceclan@xxxxxxxxx> wrote:  
> >> On 1/16/24 18:30, Jonathan Cameron wrote:  
> >>> On Mon, 15 Jan 2024 15:53:39 -0600
> >>> David Lechner <dlechner@xxxxxxxxxxxx> wrote:  
> >>>> On Wed, Dec 20, 2023 at 4:48 AM Dumitru Ceclan <mitrutzceclan@xxxxxxxxx> wrote:    
> 
> ...
> 
> >>>>> +
> >>>>> +  refin-supply:
> >>>>> +    description: external reference supply, can be used as reference for conversion.      
> >>>> If I'm understanding correctly, this represents both voltage inputs
> >>>> REF+ and REF-, correct? The datasheet says "Reference Input Negative
> >>>> Terminal. REF− can span from AVSS to AVDD1 − 1 V". It seems like they
> >>>> should be separate supplies in case REF- is non-zero. Otherwise, how
> >>>> can we know what voltage it is? (same comment applies to refin2.)    
> >>> Agreed, in this case these are directly used as references (we recently
> >>> had another driver that could take a wide range of negative and positive
> >>> inputs but in that case an internal reference was generated that didn't
> >>> made it not matter exactly what was being supplied.  Not true here though!
> >>>     
> >> Wouldn't it be alright to specify that the voltage specified here should
> >> be the actual difference (REF+)-(REF-)?  
> > How do you establish the offset to apply to single ended channels if you don't
> > know the value of REF- (relative to local ground)?
> > 
> > So no - as the device supports single ended channels the difference isn't
> > enough information.  It would probably be fine to do as you say if it
> > were a device with only differential channels where all that matters is
> > the scaling.  
> 
> 	I suppose that you are referring to the first page presentation: "Cross
> point multiplexer; 8 full differential or 16 single-ended channels". I
> consider this to be a bit misleading as all channels are actually fully
> differential (must select positive and negative source, AVSS is not one
> of them).
> 
> 
> 	Even more, the datasheet specifies that when using "single-ended"
> inputs you need to select which of the pins is the common one and
> connect it to the desired GND (be it AVSS, REF-):
> 
> "Because there is a cross point mux, the user can set any of the analog
> inputs as the common pin. An example of such a scenario is to connect
> the AIN16 pin to AVSS or to the REFOUT voltage (that is, AVSS +
> 2.5 V)" (ad7173-8 page 27)
> 
> 	For me this is 100% the case that this is a fully differential ADC in
> which the datasheet presents a way to use it single-ended. Let's say
> that we are using EXT_REF, and REF- is non zero. If someone connects
> AVSS to the desired common pin, the ADC will still measure correctly the
> difference of voltage between AIN_POS and AIN_NEG and compare it to the
> EXT_REF.

Thanks for the explanation. I indeed was mislead by the introduction!
Seems that the difference in the reference inputs is sufficient for this
device.

Jonathan