Re: [PATCH v2 1/7] dt-bindings: iio: adc: Add TI ADS126x ADC family

On 7/26/26 2:49 PM, Kurt Borja wrote:
> On Tue Jun 30, 2026 at 7:28 PM -05, Jonathan Cameron wrote:
>> On Tue, 30 Jun 2026 13:38:30 -0500
>> David Lechner <[email protected]> wrote:
>>
>>> On 6/30/26 12:14 PM, Kurt Borja wrote:
>>>> On Mon Jun 29, 2026 at 11:43 AM -05, David Lechner wrote:  
>>>>> On 6/29/26 11:27 AM, Kurt Borja wrote:  
>>>>>> On Mon Jun 29, 2026 at 9:21 AM -05, David Lechner wrote:  
>>>>>>> On 6/28/26 2:12 PM, Kurt Borja wrote:  
>>>>>>>> On Sun Jun 28, 2026 at 10:45 AM -05, David Lechner wrote:  
>>>>>>>>> On 6/28/26 12:36 AM, Kurt Borja wrote:  
>>>>>>>>>> The ADS1262 and ADS1263 are 32-bit, 38.4-kSPS delta-sigma ADCs with an
>>>>>>>>>> integrated PGA, internal reference, excitation and burn-out current
>>>>>>>>>> sources for sensor biasing and diagnostics. The ADS1263 adds a second,
>>>>>>>>>> 24-bit delta-sigma ADC (ADC2) for background measurements.
>>>>>>>>>>  
>>>>>>>
>>>>>>> ...
>>>>>>>  
>>>>>>>>>> +
>>>>>>>>>> +patternProperties:
>>>>>>>>>> +  "^channel@[0-9]+$":
>>>>>>>>>> +    $ref: /schemas/iio/adc/adc.yaml#
>>>>>>>>>> +    unevaluatedProperties: false
>>>>>>>>>> +
>>>>>>>>>> +    properties:
>>>>>>>>>> +      reg:
>>>>>>>>>> +        maxItems: 1
>>>>>>>>>> +  
>>>>>>>>>
>>>>>>>>> If we want to allow single-ended/pseudo-differential inputs, then we should
>>>>>>>>> also allow single-channel (positive pin) and common-mode-channel (negative
>>>>>>>>> pin) properties.
>>>>>>>>>
>>>>>>>>> This will also require additional common-mode-<N>-supply properties to allow
>>>>>>>>> for the negative pin connected to something other than GND.  
>>>>>>>>
>>>>>>>> Ah interesting. Why the N though? wouldn't a single supply connected to
>>>>>>>> AINCOM be enough here?  
>>>>>>>
>>>>>>> In theory, any AINx could be a common mode input by connecting it to
>>>>>>> a constant voltage supply.  
>>>>>>
>>>>>> Technically yes, but there is a pin named AINCOM intended for this
>>>>>> purpose. Can we do a simplification here? I propose having
>>>>>>
>>>>>>   common-mode-channel:
>>>>>>     /* AINCOM pin */
>>>>>>     const: 10
>>>>>>
>>>>>> That way we would only need one common-mode-supply. Would that be okay?
>>>>>>  
>>>>>
>>>>> Ideally, we shouldn't limit how the chip can be wired up in the devicetree
>>>>> bindings. It doesn't mean that we have to implement everything in the driver
>>>>> though.  
>>
>> I'm a bit confused on this.  Arguably you could wire up any side of a differential
>> pair to a common mode signal, but we haven't had this binding applied more generally
>> than cases where their is a setup where the intent is a shared common (shielding
>> on cable in some cases IIRC). 
>>
>> So why is this case special? I see AIN_COM can be wired up to an internal vbias
>> but other than that it looks like any other AINX input.
> 
> Hi, 
> 
> I forgot to respond to this email!
> 
> As you said, there is nothing special about AINCOM, it's just like any
> other input. That's why David suggested to have a common-mode-N-supply
> for each input.
> 
> I implemented this of course, but I still have doubts because there is
> no precedent for something like this in IIO. I understand the logic, but

FWIW, we have 4 ainx-supply properties on adi,ad7380.

> I'll personally be inclined to drop it. If there is no further
> discussion I'll keep the 10 common-mode supplies.

To make the bindings less verbose, we could use:

  patternProperties:
    ain[0-9]-supply:
      description: fixed voltage input when AIN pin is used as common mode

> 
> I'm finishing v3 and I'll submit it in a couple of days.
> 
>>
>> Jonathan
>>
>>>>
>>>> Makes sense.
>>>>
>>>> I just have a doubt: In the ADS112c14 bindings, is there a specific
>>>> reason why you didn't describe common-mode-<N>-supply?
>>>>   
>>>
>>> For singled-ended inputs on ADS112c14, there is an internal
>>> connection to GND, so it isn't possible to have a pseudo-
>>> differential input like that. I guess technically, it could
>>> still be possible, but wasn't a typical wiring described in
>>> the datasheet like it is on ADS1263.
>>>
>