2022-03-17 22:05:20

by Jagath Jog J

[permalink] [raw]
Subject: [RFC] iio: accel: bma400: patches for todos

Hello,

To continue the contribution I chose the bma400 accelerometer device driver,
there are multiple todos, so thought of adding support for the same.
Currently, I am able to get the trigger buffer with data ready interrupt,
step counter and event for step change with bma400 accelero breakout board.

Below are the changes I have implemented for the bma400 driver in separate
patches

1. Conversion to device-managed function and remove the bma400_remove()
function.
2. Added trigger buffer support with data ready interrupt.
3. Added a separate channel for the step counter.
4. Added event support for step change.

Can I send the above patches for review?
Sorry for all the mistakes that I have made in the previous ds1803 patch
series and I will make sure to avoid all those mistakes.

In the bma400 driver, I am stuck on one point and your advice will be very
helpful.

https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-bus-iio
From IIO documentation for accelerometer device the calculation for
final value from the sysfs attributes of the accelerometer device for
the Z-axis will be
(in_accel_z_raw + in_accel_z_offset) * in_accel_scale = 9.8 m/s2 approximately.

In the mainline kernel bma400 device driver, there are multiple-scale
values for different acceleration ranges +-2g, +-4g, etc.

root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
cat in_accel_scale_available
0.038357 0.076714 0.153428 0.306856

With +-4g range
root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
cat in_accel_scale
0.076714
root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
cat in_accel_z_raw
519
root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
echo "0.076714 * 506" | bc
39.277568

With +- 2g range
root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
echo 0.038357 > in_accel_scale
root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
cat in_accel_scale
0.038357
root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
cat in_accel_z_raw
1012
root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
echo "0.038357 * 1012" | bc
38.817284

The _raw value from the sensor is as per the datasheet but the final value for the
Z axis after applying scale for all ranges is around 38.8 which is not expected.
I tried by placing the sensor in a different position where the X-axis value should
be approximately 9.8 m/s2 but In this case also the value is around 38.8 m/s2

Is the above method to calculate the acceleration in the Z-axis correct or
am I missing something?

For confirmation I also tested the mpu6050 IMU.

root@raspberrypi:/sys/bus/iio/devices/iio:device1# cat in_accel_scale_available
0.000598 0.001196 0.002392 0.004785
root@raspberrypi:/sys/bus/iio/devices/iio:device1# cat in_accel_scale
0.000598
root@raspberrypi:/sys/bus/iio/devices/iio:device1# cat in_accel_z_raw
16706
root@raspberrypi:/sys/bus/iio/devices/iio:device1# echo "0.000598 * 16706" | bc
9.9801
With MPU6050 I am getting 9.9801 m/s2 which is near to earth's acceleration due
to gravity 9.8 m/s2

For BMA400 I recalculated the scale values from sensitivity for different
acceleration ranges
https://www.bosch-sensortec.com/media/boschsensortec/downloads/datasheets/bst-bma400-ds000.pdf

For +-2g - 1 LSB = 0.0095768 m/s2
For +-4g - 1 LSB = 0.0191536 m/s2
For +-4g - 1 LSB = 0.0383072 m/s2
Is the above method to calculate the scale values correct or am I missing
something?

Thank you
Jagath


2022-03-18 03:50:59

by Dan Robertson

[permalink] [raw]
Subject: Re: [RFC] iio: accel: bma400: patches for todos

On Fri, Mar 18, 2022 at 03:22:27AM +0530, Jagath Jog J wrote:
> Hello,
>
> To continue the contribution I chose the bma400 accelerometer device driver,
> there are multiple todos, so thought of adding support for the same.
> Currently, I am able to get the trigger buffer with data ready interrupt,
> step counter and event for step change with bma400 accelero breakout board.

This is awesome! Thanks so much for checking out the driver. I enjoyed
writing it. It seems like quite a capable device with lots of features I've
yet to implement, and I had far less time free for development on the device
than I had hoped. Any help would be _greatly_ appreciated.

> Below are the changes I have implemented for the bma400 driver in separate
> patches
>
> 1. Conversion to device-managed function and remove the bma400_remove()
> function.
> 2. Added trigger buffer support with data ready interrupt.
> 3. Added a separate channel for the step counter.
> 4. Added event support for step change.
> Can I send the above patches for review?

I used the driver for this device as a way to learn, so I by no means claim to
be an expert, but I'll do my best to review and test patches sent. At a bare
minimum I can be a shoulder to cry on and fellow to commiserate with :-)

> Sorry for all the mistakes that I have made in the previous ds1803 patch
> series and I will make sure to avoid all those mistakes.
>
> In the bma400 driver, I am stuck on one point and your advice will be very
> helpful.
>
> https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-bus-iio
> From IIO documentation for accelerometer device the calculation for
> final value from the sysfs attributes of the accelerometer device for
> the Z-axis will be
> (in_accel_z_raw + in_accel_z_offset) * in_accel_scale = 9.8 m/s2 approximately.
>
> In the mainline kernel bma400 device driver, there are multiple-scale
> values for different acceleration ranges +-2g, +-4g, etc.
>
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> cat in_accel_scale_available
> 0.038357 0.076714 0.153428 0.306856
>
> With +-4g range
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> cat in_accel_scale
> 0.076714
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> cat in_accel_z_raw
> 519
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> echo "0.076714 * 506" | bc
> 39.277568
>
> With +- 2g range
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> echo 0.038357 > in_accel_scale
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> cat in_accel_scale
> 0.038357
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> cat in_accel_z_raw
> 1012
> root@raspberrypi:/sys/devices/platform/soc/3f804000.i2c/i2c-1/1-0014/iio:device0#
> echo "0.038357 * 1012" | bc
> 38.817284
>
> The _raw value from the sensor is as per the datasheet but the final value for the
> Z axis after applying scale for all ranges is around 38.8 which is not expected.
> I tried by placing the sensor in a different position where the X-axis value should
> be approximately 9.8 m/s2 but In this case also the value is around 38.8 m/s2
>
> Is the above method to calculate the acceleration in the Z-axis correct or
> am I missing something?

Based on your tests, a review of the code, and a re-read of the datasheet, I
definitely calculated the BMA400_SCALE_MIN and BMA400_SCALE_MAX incorrectly.

> For confirmation I also tested the mpu6050 IMU.
>
> root@raspberrypi:/sys/bus/iio/devices/iio:device1# cat in_accel_scale_available
> 0.000598 0.001196 0.002392 0.004785
> root@raspberrypi:/sys/bus/iio/devices/iio:device1# cat in_accel_scale
> 0.000598
> root@raspberrypi:/sys/bus/iio/devices/iio:device1# cat in_accel_z_raw
> 16706
> root@raspberrypi:/sys/bus/iio/devices/iio:device1# echo "0.000598 * 16706" | bc
> 9.9801
> With MPU6050 I am getting 9.9801 m/s2 which is near to earth's acceleration due
> to gravity 9.8 m/s2
>
> For BMA400 I recalculated the scale values from sensitivity for different
> acceleration ranges
> https://www.bosch-sensortec.com/media/boschsensortec/downloads/datasheets/bst-bma400-ds000.pdf
>
> For +-2g - 1 LSB = 0.0095768 m/s2
> For +-4g - 1 LSB = 0.0191536 m/s2
> For +-4g - 1 LSB = 0.0383072 m/s2

Assuming you mean 8g for the last one here.

> Is the above method to calculate the scale values correct or am I missing
> something?

Yeah, the scale is definitely off currently.

I think we need to change the following, if I'm reading the datasheet correctly.

- BMA400_SCALE_MIN to 9582
- BMA400_SCALE_MAX to 76659

Great catch and thanks for working on the device!

> Thank you
> Jagath

Cheers,

- Dan