nrf52 adc spike problem

[ENENER]

Hi,

I'm using nrf52 with s132(v17.0.0).

When I connect nrf52 with my phone, there is error like this picture.

And, the code is

#define SAADC_SAMPLES_IN_BUFFER 240

void saadc_callback(nrf_drv_saadc_evt_t const * p_event)
{
    if (p_event->type == NRF_DRV_SAADC_EVT_DONE)
    {
        ret_code_t err_code;
        uint16_t bytes_to_send = (uint16_t) SAADC_SAMPLES_IN_BUFFER;
        uint8_t DATA[SAADC_SAMPLES_IN_BUFFER];
        // set buffers
        err_code = nrf_drv_saadc_buffer_convert(p_event->data.done.p_buffer, SAADC_SAMPLES_IN_BUFFER);
        APP_ERROR_CHECK(err_code);

        // print samples on hardware UART and parse data for BLE transmission

        for (int i = 0; i < SAADC_SAMPLES_IN_BUFFER; i++)
        {
          DATA[i] = p_event->data.done.p_buffer[i];
        }
       err_code = ble_nus_data_send(&m_nus, DATA, &bytes_to_send, m_conn_handle);
         
        if ((err_code != NRF_ERROR_INVALID_STATE) && (err_code != NRF_ERROR_NOT_FOUND))
        {
            APP_ERROR_CHECK(err_code);
        }
        
    }
}

When the buffer is filled, then next buffer is filled by ADC value.

I guess the problem occurs at this time.

Also, when I use printf function to see value in PC, same problem occurs.

How can I solve this?

Thank you.

[Karl Ylvisaker]

Hello,

I'm using nrf52 with s132(v17.0.0).

Thank you for clarifying. I recommend that you upgrade to SDK v17.0.2 as it is API compatible and contains some bugfixes - all you need to do is download the v.17.0.2 SDK and move your project over to it instead.

When I connect nrf52 with my phone, there is error like this picture.

What is the axis labels here? Please include more of the image. Without the axis' labels and magnitudes it is very hard to say anything about what this is depicting.

And, the code is

Could you possibly share your entire main.c file, or at least all the SAADC related code from your application?
For example, how are you triggering sampling - are doing periodic sampling through PPI ( like in the SAADC example ), or are you calling the samples manually?

I guess the problem occurs at this time.

Also, when I use printf function to see value in PC, same problem occurs.

I assume these values are a list of returned SAADC samples? If so, what were you sampling, and what were you expecting?
It seems you have a 10x value halfway through this list. Is this expected?
Since I know nothing about what you are sampling, I also do not know if this 10 x spike is as expected. If it is, I have no idea what you would like me to read from this list - please elaborate. 

How can I solve this?

I do not yet know enough about the issue to help you resolve it, but I look forward to resolving this issue together!

Best regards,
Karl

[ENENER]

Hi.

First of all, I attached screenshot picture which means Xaxis label(1st buffer, 2nd buffer).

I applied sine graph. 

In one buffer([0, ~ , 239]), there are linearity. And I can get sine graph so exactly.

But when the buffer is changed, the problem occurs.

There are no linearity.

I guess the reason happens in changing buffer ( [0, ~ ,239] -> [0, ~ 239] ).

And I'm using PPI to use SAADC function.

I attached the code.

Thanks

void saadc_sampling_event_init(void)
{
    ret_code_t err_code;
    err_code = nrf_drv_ppi_init();
    APP_ERROR_CHECK(err_code);
    
    nrf_drv_timer_config_t timer_config = NRF_DRV_TIMER_DEFAULT_CONFIG;
    timer_config.frequency = NRF_TIMER_FREQ_31250Hz;

    err_code = nrf_drv_timer_init(&m_timer, &timer_config, timer_handler);
    APP_ERROR_CHECK(err_code);

    /* setup m_timer for compare event */
    uint32_t ticks = nrf_drv_timer_us_to_ticks(&m_timer,SAADC_SAMPLE_RATE);
    nrf_drv_timer_extended_compare(&m_timer, NRF_TIMER_CC_CHANNEL0, ticks, NRF_TIMER_SHORT_COMPARE0_CLEAR_MASK, false);
    nrf_drv_timer_enable(&m_timer);

    uint32_t timer_compare_event_addr = nrf_drv_timer_compare_event_address_get(&m_timer, NRF_TIMER_CC_CHANNEL0);
    uint32_t saadc_sample_event_addr = nrf_drv_saadc_sample_task_get();

    /* setup ppi channel so that timer compare event is triggering sample task in SAADC */
    err_code = nrf_drv_ppi_channel_alloc(&m_ppi_channel);
    APP_ERROR_CHECK(err_code);
    
    err_code = nrf_drv_ppi_channel_assign(m_ppi_channel, timer_compare_event_addr, saadc_sample_event_addr);
    APP_ERROR_CHECK(err_code);
}


void saadc_sampling_event_enable(void)
{
    ret_code_t err_code = nrf_drv_ppi_channel_enable(m_ppi_channel);
    APP_ERROR_CHECK(err_code);
}


void saadc_callback(nrf_drv_saadc_evt_t const * p_event)
{
    if (p_event->type == NRF_DRV_SAADC_EVT_DONE)
    {
        ret_code_t err_code;
        uint16_t bytes_to_send = (uint16_t) SAADC_SAMPLES_IN_BUFFER;
        uint8_t DATA[SAADC_SAMPLES_IN_BUFFER];
        // set buffers
        err_code = nrf_drv_saadc_buffer_convert(p_event->data.done.p_buffer, SAADC_SAMPLES_IN_BUFFER);
        APP_ERROR_CHECK(err_code);

        // print samples on hardware UART and parse data for BLE transmission

        for (int i = 0; i < SAADC_SAMPLES_IN_BUFFER; i++)
        {
          DATA[i] = p_event->data.done.p_buffer[i];
          printf("%d\r\n", DATA[i]);
        }
       err_code = ble_nus_data_send(&m_nus, DATA, &bytes_to_send, m_conn_handle);
         
        if ((err_code != NRF_ERROR_INVALID_STATE) 
        && (err_code != NRF_ERROR_NOT_FOUND)
        && (err_code != BLE_ERROR_GATTS_SYS_ATTR_MISSING))
        {
            APP_ERROR_CHECK(err_code);
        }
        
    }
}


void saadc_init(void)
{
    ret_code_t err_code;
	
    nrf_drv_saadc_config_t saadc_config = NRF_DRV_SAADC_DEFAULT_CONFIG;
    saadc_config.resolution = NRF_SAADC_RESOLUTION_8BIT;
	
    nrf_saadc_channel_config_t channel_0_config =
    NRF_DRV_SAADC_DEFAULT_CHANNEL_CONFIG_SE(NRF_SAADC_INPUT_AIN0);

    channel_0_config.gain = NRF_SAADC_GAIN1_6;
    channel_0_config.reference = SAADC_CH_CONFIG_REFSEL_Internal;
    channel_0_config.acq_time = NRF_SAADC_ACQTIME_40US;
    
    err_code = nrf_drv_saadc_init(&saadc_config, saadc_callback);
    APP_ERROR_CHECK(err_code);

    err_code = nrf_drv_saadc_channel_init(0, &channel_0_config);
    APP_ERROR_CHECK(err_code);

    err_code = nrf_drv_saadc_buffer_convert(m_buffer_pool[0],SAADC_SAMPLES_IN_BUFFER);
    APP_ERROR_CHECK(err_code);   
    err_code = nrf_drv_saadc_buffer_convert(m_buffer_pool[1],SAADC_SAMPLES_IN_BUFFER);
    APP_ERROR_CHECK(err_code);
}

/**@brief Application main function.
 */
int main(void)
{
    bool erase_bonds;

    // Initialize.
      uart_init();
    log_init();
    timers_init();
    buttons_leds_init(&erase_bonds);
    power_management_init();
    ble_stack_init();
    gap_params_init();
    gatt_init();
    services_init();
    advertising_init();
    conn_params_init();
    
    saadc_sampling_event_init();
    saadc_init();
    saadc_sampling_event_enable();

    // Start execution.
    printf("\r\nUART started.\r\n");
    NRF_LOG_INFO("Debug logging for UART over RTT started.");
    advertising_start();

    // Enter main loop.
    for (;;)
    {
        idle_state_handle();
    }
}   

[Karl Ylvisaker]

ENENER said:

I applied sine graph. 

In one buffer([0, ~ , 239]), there are linearity. And I can get sine graph so exactly.

But when the buffer is changed, the problem occurs.

There are no linearity.

I guess the reason happens in changing buffer ( [0, ~ ,239] -> [0, ~ 239] ).

And I'm using PPI to use SAADC function.

Thank you for clarifying.

Is the graph here from the values that the phone is receiving - i.e the phone is creating this graph to display the values it is getting, or is this a graph generated to illustrate the issue from the outputted SAADC values directly(no phone involved)? If there is a phone involved - could you tell me which BLE connection parameters are you using?
How is this graph generated, and could you show me the raw output of the SAADC instead?
I suppose the resolution of your graph is a little low, but it seems that the first value of the second buffer somehow is equal to the second last value of the first buffer.

Since you are using double buffering the SAADC will immediately continue to sample into the next buffer once the first buffer is filled - so there will be no computational time necessary here before sampling resumes. This way, you will not miss a single sample.

Best regards,
Karl

[ENENER]

Thank you

I think the reason is not nrf52 but application.

Thanks!

[Karl Ylvisaker]

ENENER said:

Thanks!

No problem at all, I am happy to help!

ENENER said:

I think the reason is not nrf52 but application.

I too think this might be likely. I am glad to hear that you think you've found the source of the issue!

Please do not hesitate to open a new ticket if you should encounter any issues or questions.

Good luck with your development!

Best regards,
Karl