Using the LIS2DH12 driver of the SDK v15.0.0 to configure the LIS2DH12 sensor

Question

Hi everyone, I'm trying to get the LIS2DH12 driver (SDK v15.0.0) to configure a LIS2DH12 sensor connected to my nRF52 Dev Kit. I've wired the chip up as follows (using the I2C interface) nRF52 -> chip P0.26 -> SDA P0.27 -> SCL GND -> GND VDD -> VDD, VDD_IO, CS, SDO I've managed to read out the WHO_AM_I register using the driver and poll some samples. The samples do change when I drop my hand on the table on which the sensor is placed. Now I'm trying to configure the FIFO and an interrupt (for now FIFO watermark) using this function: 
 #define TWI_INSTANCE_ID 0

#define MAX_PENDING_TRANSACTIONS 33

#define LIS2DH12_MIN_QUEUE_SIZE 32

NRF_TWI_MNGR_DEF(m_nrf_twi_mngr, MAX_PENDING_TRANSACTIONS, TWI_INSTANCE_ID);

NRF_TWI_SENSOR_DEF(m_nrf_twi_sensor, &m_nrf_twi_mngr, LIS2DH12_MIN_QUEUE_SIZE);

LIS2DH12_INSTANCE_DEF(m_lis2dh12, &m_nrf_twi_sensor, LIS2DH12_BASE_ADDRESS_HIGH);

...

/**
 * initialize and configure LIS2DH12 sensor
 */
void acc_init()
{
	ret_code_t err_code;

	nrf_drv_twi_config_t const config = {
 .scl = 27,
 .sda = 26,
 .frequency = NRF_DRV_TWI_FREQ_100K,
 .interrupt_priority = APP_IRQ_PRIORITY_LOWEST,
 .clear_bus_init = false
	};

	err_code = nrf_twi_mngr_init(&m_nrf_twi_mngr, &config);
	APP_ERROR_CHECK(err_code);

	err_code = nrf_twi_sensor_init(&m_nrf_twi_sensor);
	APP_ERROR_CHECK(err_code);

	// FIFO configuration
	LIS2DH12_FIFO_CFG(m_lis2dh12, true, LIS2DH12_FIFO, false, 16);

	// data acquisition configuration
	LIS2DH12_DATA_CFG(m_lis2dh12, LIS2DH12_ODR_1HZ, false, true, true, true, LIS2DH12_SCALE_2G, false);

	// interrupt configuration
	LIS2DH12_INT1_PIN_CFG(m_lis2dh12, false, true, false, false, false, true, false, false);

	err_code = lis2dh12_init(&m_lis2dh12);
	APP_ERROR_CHECK(err_code);

	err_code = lis2dh12_who_am_i_read(&m_lis2dh12, print_register, &m_data);
	APP_ERROR_CHECK(err_code);

	//LIS2DH12_INT1_CFG(m_lis2dh12, 100, 1, false, true, false, false, false, false, false, false, false);
	//err_code = lis2dh12_cfg_commit(&m_lis2dh12);
	//APP_ERROR_CHECK(err_code);
} 
 However I didn't manage to detect any interrupt (even with a oscilloscope) so far and the FIFO seems to be stuck at 32 available values while polling does not change the corresponding register value. I am at a loss here - from what I've read in the LIS2DH12 data sheet and driver code, the driver seems to do the interfacing with the chip right. I'm suspecting a wrong order in my configuration function or wrong configuration parameters. 
 
 Any pointers on what I am doing wrong?

lhochstetter · Answer

Hi everyone, I managed to figure it out on my own! Configuration: NRF_TWI_MNGR_DEF(m_nrf_twi_mngr, MAX_PENDING_TRANSACTIONS, TWI_INSTANCE_ID);

NRF_TWI_SENSOR_DEF(m_nrf_twi_sensor, &m_nrf_twi_mngr, LIS2DH12_MIN_QUEUE_SIZE);

LIS2DH12_INSTANCE_DEF(m_lis2dh12, &m_nrf_twi_sensor, LIS2DH12_BASE_ADDRESS_HIGH);

void acc_init()
{
	ret_code_t err_code;

	nrf_drv_twi_config_t const config = {
 .scl = 27,
 .sda = 26,
 .frequency = NRF_DRV_TWI_FREQ_100K,
 .interrupt_priority = APP_IRQ_PRIORITY_LOWEST,
 .clear_bus_init = false
	};

	err_code = nrf_twi_mngr_init(&m_nrf_twi_mngr, &config);
	APP_ERROR_CHECK(err_code);

	err_code = nrf_twi_sensor_init(&m_nrf_twi_sensor);
	APP_ERROR_CHECK(err_code);

	err_code = lis2dh12_init(&m_lis2dh12);
	APP_ERROR_CHECK(err_code);

	// data acquisition configuration: 10HZ sample rate, low power mode, x axis, y axis, z axis, values in range -2g <-> +2g, disable high resolution mode (in this case step size is 256)
	LIS2DH12_DATA_CFG(m_lis2dh12, LIS2DH12_ODR_10HZ, true, true, true, true, LIS2DH12_SCALE_2G, 0);
	err_code = lis2dh12_cfg_commit(&m_lis2dh12);
	APP_ERROR_CHECK(err_code);

	// FIFO configuration: enable FIFO, streaming mode, int1 pin used, 32 values watermark
	LIS2DH12_FIFO_CFG(m_lis2dh12, 1, LIS2DH12_STREAM, 0, LIS2DH12_MIN_QUEUE_SIZE);
	err_code = lis2dh12_cfg_commit(&m_lis2dh12);
	APP_ERROR_CHECK(err_code);

	// interrupt configurataion: disable everything except the FIFO watermark interrupt
	LIS2DH12_INT1_PIN_CFG(m_lis2dh12, 0, 0, 0, 0, 1, 0, 1, 0);
	err_code = lis2dh12_cfg_commit(&m_lis2dh12);
	APP_ERROR_CHECK(err_code);
} Interrupt (using P004 and a LED to indicate that an interrupt has happend): void in_pin_handler(nrf_drv_gpiote_pin_t pin, nrf_gpiote_polarity_t action)
{
	nrf_drv_gpiote_out_toggle(PIN_OUT);
	data_available = 1;
}

static void gpio_init(void)
{
	ret_code_t err_code;

	err_code = nrf_drv_gpiote_init();
	APP_ERROR_CHECK(err_code);

	nrf_drv_gpiote_out_config_t out_config = GPIOTE_CONFIG_OUT_SIMPLE(false);

	err_code = nrf_drv_gpiote_out_init(PIN_OUT, &out_config);
	APP_ERROR_CHECK(err_code);

	nrf_drv_gpiote_in_config_t in_config = GPIOTE_CONFIG_IN_SENSE_HITOLO(false);

	err_code = nrf_drv_gpiote_in_init(PIN_IN, &in_config, in_pin_handler);
	APP_ERROR_CHECK(err_code);

	nrf_drv_gpiote_in_event_enable(PIN_IN, true);
}
 And finally: retrieving and printing the values: void acc_print_data()
{
	ret_code_t err_code;

	err_code = lis2dh12_data_read(&m_lis2dh12, 0x00, m_sample, LIS2DH12_MIN_QUEUE_SIZE);
	APP_ERROR_CHECK(err_code);

	for (uint8_t i = 0; i < LIS2DH12_MIN_QUEUE_SIZE; i++)
 NRF_LOG_INFO("%d %d %d", m_sample[i].x, m_sample[i].y, m_sample[i].z);

}

[...]

	while (true) {
 // avoid accessing the registers to fast, value arbitrarily chosen
 nrf_delay_ms(200);
 // debug: check FIFO_SRC register to see the FSS value rise
 err_code = lis2dh12_fifo_src_read(&m_lis2dh12, 0x00, &m_data);
 APP_ERROR_CHECK(err_code);
 NRF_LOG_INFO("WTM: %d OVN: %d EMPTY: %d FSS: %2d ", (m_data & 0b10000000), (m_data & 0b01000000), (m_data & 0b00100000), (m_data & 0b0001111));

 // do not try to access FIFO values if watermark was not reached yet
 if (!data_available)
 continue;

 // collect and print the data
 acc_print_data();
 // "reset" interrupt
 data_available = 0;
 nrf_drv_gpiote_out_toggle(PIN_OUT);

 NRF_LOG_FLUSH();
	} I've encountered following "issues": * fatal system error on the nRF52 side if the watermark was not reached but data is read * accessing the sensor's registers to fast I don't know whether or not I'm still doing something wrong which then leads to the issues ...