ws2812b can not work on ncs

[kenyon]

application : ncs\v3.0.1\zephyr\samples\drivers\led\led_strip

Hi
Recently, I have been working on development in NCS. I attempted to drive the WS2812B LED lights, but they failed to light up so far. I would like to ask you to check if there are any issues with the parameters and configurations. I successfully implemented the normal operation of the nrf52832 development board and the ws2812b LED driver in SDK 17.1.Therefore, I can confirm that my hardware is functioning properly.

Here is the ws2812 pwm params on sdk17.1

#define WS2812_PIN  NRF_GPIO_PIN_MAP(0,23)

// In each value, the most significant bit (15)determines the polarity of the output and theothers (14-0) compose the 15-bit value
#define WS2812_T1H                  14 | 0x8000     //14ticks
#define WS2812_T0H                  6  | 0x8000     //6ticks
#define LED_MATRIX_WIDTH  14
#define LED_MATRIX_HEIGHT 1

void pwm_init(void){
    nrfx_pwm_config_t pwm_config = NRFX_PWM_DEFAULT_CONFIG;
    pwm_config.output_pins[0] = NRFX_PWM_PIN_NOT_USED; 
    pwm_config.output_pins[1] = WS2812_PIN; 
    pwm_config.output_pins[2] = NRFX_PWM_PIN_NOT_USED;
    pwm_config.output_pins[3] = NRFX_PWM_PIN_NOT_USED;
    pwm_config.load_mode    = NRF_PWM_LOAD_INDIVIDUAL;
    // WS2812 protocol requires a 800 kHz PWM frequency. PWM Top value = 20 and Base Clock = 16 MHz achieves this
    pwm_config.top_value    = 20; 
    pwm_config.base_clock   = NRF_PWM_CLK_16MHz;
}

Here is th spi config

nrf52dk_nrf52832.overlay
#define SPI_FREQ    2400000    //2.4Mhz = 0.42us
#define ZERO_FRAME  0x04       //100  means 0.42us"1" + 0.84"0"
#define ONE_FRAME   0x06       //110  means 0.84us"1" + 0.42"0"
&arduino_spi { /* MOSI on D11 / P0.23 */
	compatible = "nordic,nrf-spim";
	led_strip: ws2812@0 {
		compatible = "worldsemi,ws2812-spi";

		/* SPI */
		reg = <0>; /* ignored, but necessary for SPI bindings */
		spi-max-frequency = <SPI_FREQ>;

		/* WS2812 */
		chain-length = <16>; /* arbitrary; change at will */
		color-mapping = <LED_COLOR_ID_GREEN
				 LED_COLOR_ID_RED
				 LED_COLOR_ID_BLUE>;
		spi-one-frame = <ONE_FRAME>;
		spi-zero-frame = <ZERO_FRAME>;
	};
};
/ {
	aliases {
		led-strip = &led_strip;
	};
};

&pinctrl {
	spi2_default: spi2_default {
		group1 {
			psels = <NRF_PSEL(SPIM_MOSI, 0, 23)>;
			};
	};

	spi2_sleep: spi2_sleep {
		group1 {
			psels = <NRF_PSEL(SPIM_MOSI, 0 ,23)>;
			low-power-enable;
		};
	};
}

prj.conf
CONFIG_LOG=y
CONFIG_LED_STRIP=y
CONFIG_LED_STRIP_LOG_LEVEL_DBG=y

# clock setting
CONFIG_CLOCK_CONTROL_NRF=y
CONFIG_CLOCK_CONTROL_NRF_K32SRC_RC=y
CONFIG_CLOCK_CONTROL_NRF_K32SRC_XTAL=n
CONFIG_CLOCK_CONTROL_NRF_K32SRC_RC_CALIBRATION=y

CONFIG_GPIO=y


main:
#define STRIP_NODE		DT_ALIAS(led_strip)

#if DT_NODE_HAS_PROP(DT_ALIAS(led_strip), chain_length)
#define STRIP_NUM_PIXELS	DT_PROP(DT_ALIAS(led_strip), chain_length)
#else
#error Unable to determine length of LED strip
#endif

#define DELAY_TIME K_MSEC(CONFIG_SAMPLE_LED_UPDATE_DELAY)

#define RGB(_r, _g, _b) { .r = (_r), .g = (_g), .b = (_b) }

static const struct led_rgb colors[] = {
	RGB(0x0f, 0x00, 0x00), /* red */
	RGB(0x00, 0x0f, 0x00), /* green */
	RGB(0x00, 0x00, 0x0f), /* blue */
};

static struct led_rgb pixels[STRIP_NUM_PIXELS];

static const struct device *const strip = DEVICE_DT_GET(STRIP_NODE);
#define LED0_NODE DT_ALIAS(led0)
static const struct gpio_dt_spec led = GPIO_DT_SPEC_GET(LED0_NODE, gpios);

int main(void)
{
	size_t color = 0;
	int rc;

	if (device_is_ready(strip)) {
		LOG_INF("Found LED strip device %s", strip->name);
	} else {
		LOG_ERR("LED strip device %s is not ready", strip->name);
		return 0;
	}

	if (!gpio_is_ready_dt(&led)) {
		return 0;
	}

	rc = gpio_pin_configure_dt(&led, GPIO_OUTPUT_ACTIVE);
	if (rc < 0) {
		return 0;
	}

	LOG_INF("Displaying pattern on strip");
	while (1) {
		for (size_t cursor = 0; cursor < ARRAY_SIZE(pixels); cursor++) {
			memset(&pixels, 0x00, sizeof(pixels));
			memcpy(&pixels[cursor], &colors[color], sizeof(struct led_rgb));

			rc = led_strip_update_rgb(strip, pixels, STRIP_NUM_PIXELS);
			if (rc) {
				LOG_ERR("couldn't update strip: %d", rc);
			}

			k_sleep(DELAY_TIME);
		}

		color = (color + 1) % ARRAY_SIZE(colors);
	}

	return 0;
}

best regard

[Vidar Berg]

Hi,

I'm not able to see anything obviously wrong from the code and configurations you posted. Do you have a oscilloscope or logic analyzer you can use to compare the waveforms with and confirm that the "one frame" and "zero frame" have the same length? 

Best regards,

Vidar

[kenyon]

Hi Vidar

static const struct led_rgb colors[] = {

    RGB(0xFF, 0x00, 0x00), /* red */

};

#define SPI_FREQ    8000000 //  1 freq : 125ns  

#define ZERO_FRAME  0x03    //  0000 0011   (ws2812b code1: 250ns h + 750us l)

#define ONE_FRAME   0x3F    // 0011 1111   (ws2812b code0: 750ns h + 250us l)  

I captured the waveform of MOSI using the logic analyzer. It corresponds to colors.
But why are not all the light strips connected to my development board red? What's the reason for this?
I have attached the photos of the abnormal light strips and the waveforms traced by the logic analyzer.


Best regard

[kenyon]

Hi Vidar

#define SPI_FREQ    8000000 //  1 freq : 125ns  

#define ZERO_FRAME  0xC0    //  1100 0000   (ws2812b code1: 250ns h + 750us l)

#define ONE_FRAME   0xFC    //  1111 1100   (ws2812b code0: 750ns h + 250us l)  

&arduino_spi { /* P0.25 */

    compatible = "nordic,nrf-spim";

    led_strip: ws2812@0 {

        compatible = "worldsemi,ws2812-spi";

        /* SPI */

        reg = <0>; /* ignored, but necessary for SPI bindings */

        spi-max-frequency = <SPI_FREQ>;

        /* WS2812 */

        chain-length = <2>; /* arbitrary; change at will */

        color-mapping = <LED_COLOR_ID_GREEN

                 LED_COLOR_ID_RED

                 LED_COLOR_ID_BLUE>;

        spi-one-frame = <ONE_FRAME>;

        spi-zero-frame = <ZERO_FRAME>;

    };

};

I attempted to light only two of the light bulbs, and made some modifications to ZERO_FRAME and ONE_FRAME. I discovered that the color of the first light bulb was incorrect. I checked the waveform and it seemed that it did not follow the budgeted plan.Two strange waveforms appeared at the very beginning and the very end.
Best regard

[Vidar Berg]

Hi,

It's been a while since I tried to use neopixels, so I don't remember the details of how it really worked and if the pause at the beginning and end is a problem or not. Could you please compare the difference between the wafeform generated from your SDK 17 project to see what the differences are?

[kenyon]

Hi Vidar
PWM Specific parameters in SDK17.1 are

#define WS2812_T1H                  14 | 0x8000   //code1: 875ns H + 375ns L 
#define WS2812_T0H                  6  | 0x8000   //code0: 375ns H + 875ns L
static uint32_t pwm_init(void)
{
    nrfx_pwm_config_t pwm_config = NRFX_PWM_DEFAULT_CONFIG;
    pwm_config.output_pins[0] = NRFX_PWM_PIN_NOT_USED; 
    pwm_config.output_pins[1] = WS2812_PIN; 
    pwm_config.output_pins[2] = NRFX_PWM_PIN_NOT_USED;
    pwm_config.output_pins[3] = NRFX_PWM_PIN_NOT_USED;
    pwm_config.load_mode    = NRF_PWM_LOAD_INDIVIDUAL;
    // WS2812 protocol requires a 800 kHz PWM frequency. PWM Top value = 20 and Base Clock = 16 MHz achieves this
    pwm_config.top_value    = 20; 
    pwm_config.base_clock   = NRF_PWM_CLK_16MHz;
    
    return nrfx_pwm_init(&m_pwm0, &pwm_config, pwm_handler);
}

Below is the waveform I debugged using SDK17's PWM. It involves two light bulbs, and the output is 0XFF0000.A very standard waveform!

[Vidar Berg]

Hello,

I tried the LED strip sample (zephyr/samples/drivers/led/led_strip/) from SDK v3.1.1 with nrf52dk/nrf52832 as the build target. I did not make any changes to the sample, but I do not see the inverted signal at the end of the transfer as you did.

4064.led_strip_nrf52832.hex

Did you see the inverted signal before you made your changes?