nrf52810 PWM issue

[WrightLin]

Hi All:

When I simulated PWM by GPIO, the output current noise was shown like the picture below (red circle area):

 

Cause it will lead to a loading issue in the backend, so I'm wondering the root cause and if there is any possible solution for this.

Any suggestion will be great. Thanks!!

Update working details: 

SDK: 14.2

52810: QFAAB0 - 1740AC

  

  

   

Update code of generating PWM :

We use GPIOTE to simulate the PWM.

 

    #define  IO_KEY_PAIRING     15

    #define  IO_ZERO_X               31

 

     APP_ERROR_CHECK(nrf_drv_gpiote_init());

    nrf_drv_gpiote_in_config_t gpiote_config =

    {

        .sense = NRF_GPIOTE_POLARITY_TOGGLE,

        .pull = NRF_GPIO_PIN_PULLUP,

        .is_watcher = false,

        .hi_accuracy = false

    };

    APP_ERROR_CHECK(nrf_drv_gpiote_in_init(IO_KEY_PAIRING, &gpiote_config, gpiote_event_handler));

    nrf_drv_gpiote_in_event_enable(IO_KEY_PAIRING, true);

    gpiote_config.pull = NRF_GPIO_PIN_NOPULL;

    APP_ERROR_CHECK(nrf_drv_gpiote_in_init(IO_ZERO_X, &gpiote_config, gpiote_event_handler));

    nrf_drv_gpiote_in_event_enable(IO_ZERO_X, true);

 

In  gpiote_event_handler(), we generate PWM pulse according to IO_ZERO_X.

  

    

  

Update code of generating PWM  --- gpiote_event_handler()- :

static void gpiote_event_handler(nrf_drv_gpiote_pin_t pin, nrf_gpiote_polarity_t action)

{

    NRF_LOG_INFO("gpiote_event_handler");

    if (pin == IO_KEY_PAIRING)

    {

        uint32_t read_value = nrf_gpio_pin_read(IO_KEY_PAIRING);

        if (!read_value)

        {

            nrf_delay_ms(1);

            if (read_value == nrf_gpio_pin_read(IO_KEY_PAIRING))

            {

                m_pairing_key_latch_for_on_off = 0x01;

                m_pairing_key_latch = 0x01;

            }

        }

    }

 

    if (pin == IO_ZERO_X)

    {

        controls_zero_x_trigger();//generate PWM

    }

}

[WrightLin]

Hi Mahesh & Hung:

I have updated the code of generating PWM , please kindly check it. Thank you!!! 

[Hung Bui]

Hi, could you explain a little bit more ? You mean you flip the GPIO pin based on the input from IO_ZERO_X ? Like cloning the signal ? And there is no timer ? 

[Hung Bui]

I'm not sure what you are trying to do. Could you please explain ? What is generating the duty cycle ? 

[WrightLin]

Hi Bui:

we are trying to make PWM pulse , and it's controlled by an external signal "ZeroX".

 

 
To be more precisely, when we detect there is a rising/falling edge of ZeroX , we trigger nrf52 to generate PWM pulse.

It's a known issue that Softdevice will interfere the process of  PWM generating.

So, when generating PWM pulse , we should  use PWM driver to avoid interference , right?

  

 

But the controlling signal -- "ZeroX" which is also an external signal .

We're wondering if it will also be interfered by  softdevice. 

If it will be interfered , then how should we do to avoid this situation ?  

[Mahesh Venkitachalam]

I think your GPIOTE idea is correct, but you should just use the external pin trigger to toggle PWM on/off in the handler. The PWM itself should be configured separately. You also have a 1ms delay in your gpiote handler, which might be a bad idea.

[WrightLin]

Hi  Mahesh:

Got it , thanks for your suggestion. 

Do you have any idea about whether the external signal be interfered by softdevice or not?

[Mahesh Venkitachalam]

I don't think you'll have a problem with softdevice. You can see the softdevice documentation to see which peripherals are in use. As long as you use others (and there are plenty left) you should not have a problem.  

"The Programmable peripheral interconnect (PPI) enables peripherals to interact autonomously with each other using tasks and events independent of the CPU."

So this is designed exactly for what you're doing.  An external signal can control PWM with minimal CPU involvement.

[WrightLin]

Hi Mahesh and Bui : 

Thanks for your help. 

our customer is trying to use PPI methods to solve their problem.