Asynchronous uart rx issues on nrf52832 using Zephyr. Constantly receive UART_RX_DISABLED event

[JoshK]

I am using an nrf528232 based BLE module, the Taiyo Yuden EYSHSNZWZ: https://www.digikey.com/en/products/detail/kaga-fei-america-inc/EYSHSNZWZ/7104534

Running BLE and uart using Zephyr. I am using an example that has worked on other hardware, including the nrf52840 development kit, which is why I thought my code was solid

I am having issues with the asynchronous uart api in Zephyr: https://docs.zephyrproject.org/latest/hardware/peripherals/uart.html#uart-async-api.

In my uart interrupt callback function I observe that the interrupt is being toggled repeatedly about every 560us even though there is no data being sent on the UART RX line. Debugging shows I receive the same sequence of events in the uart callback function - events of type UART_RX_BUF_RELEASED, UART_RX_BUF_REQUEST, UART_RX_DISABLED repeatedly in that order. Using the same code on the nrf52840 dev kit I only receive a single UART_RX_BUF_REQUEST event when the asynchronous rx is started with uart_rx_enable(), and then no events until data is actually sent on the line.

Any ideas of what is wrong here? What is causing these repeated events and UART interrupts?

My callback function code is below:

/**

 * @brief UART callback function. Format specified by zephyr uart driver

 * @param dev UART device instance.

 * @param evt Pointer to uart_event instance.

 * @param user_data Pointer to data specified by user.

*/

static void uart_callback(const struct device *dev, struct uart_event *evt, void *user_data)

{

    //debug code

    gpio_pin_toggle_dt(&dbg0_pin);

    if(dbg_events_idx < DBG_NUM_EVENTS)

    {

        dbg_events[dbg_events_idx++] = *evt;

    }

    else

        dbg_events_idx = DBG_NUM_EVENTS;

    //handle interrupt event

    switch (evt->type)

    {

    case UART_RX_RDY:

        if ((evt->data.rx.len != 0) && (evt->data.rx.len < sizeof(rx_buf))) // Do nothing if buffer is full.

        {

            memset(&(rx_buf[0]), 0, sizeof(rx_buf));

            rx_buf_size = evt->data.rx.len;

            memcpy(&(rx_buf[0]), &(rx_dma_buf[0]), rx_buf_size);

            new_rx = true;

        }

        uart_rx_disable(dev); // Use this to reset DMA input to the RX buffer

        break;

    case UART_RX_DISABLED:

        memset(&(rx_dma_buf[0]), 0, sizeof(rx_dma_buf));

        uart_rx_enable(dev, &(rx_dma_buf[0]), sizeof(rx_dma_buf), RECEIVE_TIMEOUT_US);

        break;

    case UART_TX_DONE:

        tx_in_progress = false;

        break;

    case UART_TX_ABORTED:

        tx_in_progress = false;

        break;

    default:

        break;

    }

}

[JoshK]

Changing the device tree files helped solve this problem.

I changed the uart configuration to be 

compatible = "nordic,nrf-uart";

instead of

compatible = "nordic,nrf-uarte";

&uart0 {

    compatible = "nordic,nrf-uart";

    status = "okay";

    current-speed = <115200>;

    pinctrl-0 = <&uart0_default>;

    pinctrl-1 = <&uart0_sleep>;

    pinctrl-names = "default", "sleep";

};

[Kenneth]

Glad to hear you have made it work, but I don't see how that should solve the problem, I would have looked at adding a pull-up on the RXD line or at least ensured it is high level and not floating, else it will cause a lot of errors on the RXD line when not used/connected.

[JoshK]

I did verify the RX line with a scope. The line was set with a pullup and did actually remain high. This seemed like a software problem. I am not sure why treating it as a uart instead of uarte solved the problem.