LSM6DSL Sample Fails to Reboot LSM6DS3TR-C (Pin-for-pin compatible with LSM6DSL) on XIAO BLE Sense

[MattMacGregor]

New to Zephyr so apologies if there are any violations of contribution rules/standards in this issue.

I'm trying to get the IMU on the XIAO BLE Sense up and running with the lsm6dsl sample but it seems to fail during the initialization section of the LSM6DSL driver. As the LSM6DS3TR-C is pretty much a pin-for-pin replacement, I believe this should be working, so something seems amiss (but I acknowledge this sample hasn't been officially tested on the XIAO BLE Sense yet).

It seems to fail at these lines

zephyr/drivers/i2c/i2c_nrfx_twim.c

Lines 155 to 178 in b0ea0c8

	ret = k_sem_take(&dev_data->completion_sync,
	I2C_TRANSFER_TIMEOUT_MSEC);
	if (ret != 0) {
	/* Whatever the frequency, completion_sync should have
	* been given by the event handler.
	*
	* If it hasn't, it's probably due to an hardware issue
	* on the I2C line, for example a short between SDA and
	* GND.
	* This is issue has also been when trying to use the
	* I2C bus during MCU internal flash erase.
	*
	* In many situation, a retry is sufficient.
	* However, some time the I2C device get stuck and need
	* help to recover.
	* Therefore we always call i2c_nrfx_twim_recover_bus()
	* to make sure everything has been done to restore the
	* bus from this error.
	*/
	LOG_ERR("Error on I2C line occurred for message %d", i);
	(void)i2c_nrfx_twim_recover_bus(dev);
	ret = -EIO;
	break;
	}

in the nrfx_twim driver after calling lsm6dsl_reboot here:

zephyr/drivers/sensor/lsm6dsl/lsm6dsl.c

Line 716 in 9a63f9d

if (lsm6dsl_reboot(dev) < 0) {

I2C debug log messages show I2C messages repeatedly being dropped when running the sample, and since the reboot pin process seems to be the same between the two chips, my gut is pointing me towards the I2C being the issue.

I added a log message to the event_handler for the nrfx_twim and didn't see any messages from it. Add that it fails to take the completion_sync mutex, which is released by the event_handler on NRFX_TWIM_EVT_DONE, seems to indicate that the event_handler is not being successfully called. FYI, k_sem_take always fails with -EAGAIN, but I don't believe that is much help in this case. Maybe it's my lack of experience but I cannot seem to find why this would be the case.

To reproduce simply build the lsm6dsl sample (samples/sensor/lsm6dsl) for the board xiao_ble_sense. It is using all the default configuration settings.

[MattMacGregor]

This was rightfully converted to a discussion, as the problem seems to be more focused in the I2C realm then this sensor. Here is a link to the issue with a bit more detail: #56012

[tarikgraba]

@MattMacGregor with PR #56640 applied, I can access the xiao sens imu using i2c commands from a Zephyr shell (starting with samples/basic/minimal), still the the lsm6dsl sample does not work.
Any hints?

[tarikgraba]

I have seen from the PR discussion that the regulator boot delay has also to be changed (from 900 to 3000) which makes I2C communication and the example work correctly.

[tarikgraba]

PR #57875 changes the default boot delay of the regulator.

[purpl3F0x]

Hello, is this supposed to be working now ? Cause I still get [00:00:00.500,762] <err> i2c_nrfx_twim: Error on I2C line occurred for message 0 [00:00:00.500,915] <err> LSM6DSL: Failed to initialize chip, using ncs 2.4.0 with a modified xiao ble sense dts

[syamusivan]

Increase "startup-delay-us" to some higher value (10000 us worked for me) within xiao_ble_sense.dts. The delay was around 900 us initially. I have not tested the optimal delay (higher than 900 us).

[purpl3F0x]

Wasn't the delay. I was falsly enabling pull-up.

[syamusivan]

Could you please send me the code snippet that was falsly enabling pull-up. This issue is reappearing for me.
Thank you

[mentarus]

Still seems to be an issue, posted about it on the Nordic forum here. Note that the Arduino libraries work fine, I confirmed that this is not a hardware issue as I can access the IMU fine through the SEEED provided library.