[PATCH v10 00/11] iio: adc: Add support for AVIA HX710B ADC

Piyush Patle posted 11 patches 2 weeks ago
There is a newer version of this series
.../bindings/iio/adc/avia-hx711.yaml          |  79 +++-
drivers/iio/adc/Kconfig                       |   8 +-
drivers/iio/adc/hx711.c                       | 351 +++++++++++++++---
3 files changed, 363 insertions(+), 75 deletions(-)
[PATCH v10 00/11] iio: adc: Add support for AVIA HX710B ADC
Posted by Piyush Patle 2 weeks ago
The HX710B is a 24-bit ADC from AVIA Semiconductor, related to the
HX711 already supported in this driver.  Unlike the HX711 (which has
selectable gain and two input channels), the HX710B has a fixed gain
of 128 and two operating modes selected by the trailing PD_SCK pulse
count after each conversion:

  25 pulses (1 trailing): differential input at 10 SPS
  26 pulses (2 trailing): DVDD-AVDD supply monitor at 40 SPS
  27 pulses (3 trailing): differential input at 40 SPS

This series first extends the existing HX711 binding and driver with
preparatory refactoring, then adds HX710B support on top.

The differential input is exposed as a single IIO channel with
IIO_CHAN_INFO_SAMP_FREQ, allowing userspace to select 10 or 40 SPS
without needing two separate channels for the same physical input.
The supply monitor is a second channel at fixed 40 SPS.

Patches 01/11, 02/11, 03/11, and 05/11 have already been applied to
the iio.git testing branch.  They are re-included here so the series
applies cleanly as a unit.

Changes from v9:
  - 02/11: Added dvdd-supply alongside vsup-supply. Per the v8
    discussion, DVDD is a board-supplied digital rail on the HX711
    (figure 1 of the datasheet shows it as a separate input supply,
    not merely a bypass capacitor point). It was missing from the
    original binding. Moved its declaration from 04/11 to this earlier
    shared-supply patch.
  - 04/11: Removed dvdd-supply declaration (now in 02/11). Updated
    commit message to reflect the patch now adds only vref-supply and
    the allOf constraint.
  - 06/11: No code change. Removed inaccurate below-the-line note
    about __counted_by_ptr() commit reference.
  - 08/11: Remove blank line between val = hx711_wait_for_ready() and
    if (val) check in hx711_reset().
  - 11/11: Revert IIO_DECLARE_BUFFER_WITH_TS() to explicit named
    struct (both variants have exactly 2 channels). Add
    iio_device_claim_direct_mode() in hx710b_write_raw() to prevent
    concurrent hardware changes during triggered buffer capture; drop
    channel_set = 0 reset (hx711_set_hx710b_channel() re-programs the
    chip automatically on the next read). MODULE_DESCRIPTION: change
    'compatible' to 'similar'.

Piyush Patle (11):
  dt-bindings: iio: adc: hx711: clean up existing binding text
  dt-bindings: iio: adc: hx711: add VSUP supply property
  dt-bindings: iio: adc: hx711: add RATE GPIO property
  dt-bindings: iio: adc: hx711: add HX710B support
  iio: adc: hx711: move scale computation to per-device storage
  iio: adc: hx711: introduce hx711_chip_info structure
  iio: adc: hx711: pass trailing pulse count into hx711_read
  iio: adc: hx711: split variable assignments in hx711_read and
    hx711_reset
  iio: adc: hx711: localize loop iterators in hx711_read
  iio: adc: hx711: pass iio_chan_spec to hx711_reset_read
  iio: adc: hx711: add support for HX710B

 .../bindings/iio/adc/avia-hx711.yaml          |  79 +++-
 drivers/iio/adc/Kconfig                       |   8 +-
 drivers/iio/adc/hx711.c                       | 351 +++++++++++++++---
 3 files changed, 363 insertions(+), 75 deletions(-)

-- 
2.43.0
Re: [PATCH v10 00/11] iio: adc: Add support for AVIA HX710B ADC
Posted by Jonathan Cameron 1 week, 5 days ago
On Mon, 25 May 2026 16:21:16 +0530
Piyush Patle <piyushpatle228@gmail.com> wrote:

> The HX710B is a 24-bit ADC from AVIA Semiconductor, related to the
> HX711 already supported in this driver.  Unlike the HX711 (which has
> selectable gain and two input channels), the HX710B has a fixed gain
> of 128 and two operating modes selected by the trailing PD_SCK pulse
> count after each conversion:
> 
>   25 pulses (1 trailing): differential input at 10 SPS
>   26 pulses (2 trailing): DVDD-AVDD supply monitor at 40 SPS
>   27 pulses (3 trailing): differential input at 40 SPS
> 
> This series first extends the existing HX711 binding and driver with
> preparatory refactoring, then adds HX710B support on top.
> 
> The differential input is exposed as a single IIO channel with
> IIO_CHAN_INFO_SAMP_FREQ, allowing userspace to select 10 or 40 SPS
> without needing two separate channels for the same physical input.
> The supply monitor is a second channel at fixed 40 SPS.
> 
> Patches 01/11, 02/11, 03/11, and 05/11 have already been applied to
> the iio.git testing branch.  They are re-included here so the series
> applies cleanly as a unit.

Given the version I had of patch 2 was rather different from current
I've rebased my tree to drop it and patch 3.
Thankfully that didn't result in an sha hashes changing for the few
fixes I have queued up.  Hopefully not too painful for anyone
based on my togreg tree. Briefly that will be out of sync with testing
as I have other stuff being tested on there.

> 
> Changes from v9:
>   - 02/11: Added dvdd-supply alongside vsup-supply. Per the v8
>     discussion, DVDD is a board-supplied digital rail on the HX711
>     (figure 1 of the datasheet shows it as a separate input supply,
>     not merely a bypass capacitor point). It was missing from the
>     original binding. Moved its declaration from 04/11 to this earlier
>     shared-supply patch.
>   - 04/11: Removed dvdd-supply declaration (now in 02/11). Updated
>     commit message to reflect the patch now adds only vref-supply and
>     the allOf constraint.
>   - 06/11: No code change. Removed inaccurate below-the-line note
>     about __counted_by_ptr() commit reference.
>   - 08/11: Remove blank line between val = hx711_wait_for_ready() and
>     if (val) check in hx711_reset().
>   - 11/11: Revert IIO_DECLARE_BUFFER_WITH_TS() to explicit named
>     struct (both variants have exactly 2 channels). Add
>     iio_device_claim_direct_mode() in hx710b_write_raw() to prevent
>     concurrent hardware changes during triggered buffer capture; drop
>     channel_set = 0 reset (hx711_set_hx710b_channel() re-programs the
>     chip automatically on the next read). MODULE_DESCRIPTION: change
>     'compatible' to 'similar'.
> 
> Piyush Patle (11):
>   dt-bindings: iio: adc: hx711: clean up existing binding text
>   dt-bindings: iio: adc: hx711: add VSUP supply property
>   dt-bindings: iio: adc: hx711: add RATE GPIO property
>   dt-bindings: iio: adc: hx711: add HX710B support
>   iio: adc: hx711: move scale computation to per-device storage
>   iio: adc: hx711: introduce hx711_chip_info structure
>   iio: adc: hx711: pass trailing pulse count into hx711_read
>   iio: adc: hx711: split variable assignments in hx711_read and
>     hx711_reset
>   iio: adc: hx711: localize loop iterators in hx711_read
>   iio: adc: hx711: pass iio_chan_spec to hx711_reset_read
>   iio: adc: hx711: add support for HX710B
> 
>  .../bindings/iio/adc/avia-hx711.yaml          |  79 +++-
>  drivers/iio/adc/Kconfig                       |   8 +-
>  drivers/iio/adc/hx711.c                       | 351 +++++++++++++++---
>  3 files changed, 363 insertions(+), 75 deletions(-)
>
Re: [PATCH v10 00/11] iio: adc: Add support for AVIA HX710B ADC
Posted by Jonathan Cameron 1 week, 5 days ago
On Mon, 25 May 2026 16:21:16 +0530
Piyush Patle <piyushpatle228@gmail.com> wrote:

> The HX710B is a 24-bit ADC from AVIA Semiconductor, related to the
> HX711 already supported in this driver.  Unlike the HX711 (which has
> selectable gain and two input channels), the HX710B has a fixed gain
> of 128 and two operating modes selected by the trailing PD_SCK pulse
> count after each conversion:
> 
>   25 pulses (1 trailing): differential input at 10 SPS
>   26 pulses (2 trailing): DVDD-AVDD supply monitor at 40 SPS
>   27 pulses (3 trailing): differential input at 40 SPS
> 
> This series first extends the existing HX711 binding and driver with
> preparatory refactoring, then adds HX710B support on top.
> 
> The differential input is exposed as a single IIO channel with
> IIO_CHAN_INFO_SAMP_FREQ, allowing userspace to select 10 or 40 SPS
> without needing two separate channels for the same physical input.
> The supply monitor is a second channel at fixed 40 SPS.
> 
> Patches 01/11, 02/11, 03/11, and 05/11 have already been applied to
> the iio.git testing branch.  They are re-included here so the series
> applies cleanly as a unit.

FWIW I took a look at Sashiko's feedback and other than patch descriptions
and the dt-binding dicussion that is ongoing they were either
- wrong, though that sometimes was for subtle reasons - like the fact we
  can't just guess the device if the driver is forced to bind.  We should
  fail in that case.
- pre existing issues.  If you have time to look at those, then great
  but I'm not going to insist on that.

Jonathan