STM32G0B1RE output clock

[undacmic]

Hello everyone and thank you for taking the time to read my post.

The problem I'm encountering is regarding the configuration of MCO.

I configured MCOSEL and MCOPRE to use PLLRCLK with a 1 prescaler value. The PLLR clock is configured with a frequency of 64 MHz.

I also activated GPIOAEN in RCC registers and configured PA8 for alternate function 0, push&pull, no push, no pull, low speed and i dont get the clock at all. If i connect an oscilloscope to a pin direclty touching SB17 on the board, i can see the rectangular signal and its generated well, but when I try using PA8 it doesn't work. I have connected a GND referencec to the oscilloscope as well.

For reference, I will leave here the snippet of code:

static void MX_GPIO_Init(void)
{
  GPIO_InitTypeDef GPIO_InitStruct = {0};
/* USER CODE BEGIN MX_GPIO_Init_1 */
/* USER CODE END MX_GPIO_Init_1 */

  /* GPIO Ports Clock Enable */
  __HAL_RCC_GPIOA_CLK_ENABLE();

  /*Configure GPIO pin : PA8 */
  GPIO_InitStruct.Pin = GPIO_PIN_8;
  GPIO_InitStruct.Mode = GPIO_MODE_AF_PP;
  GPIO_InitStruct.Pull = GPIO_NOPULL;
  GPIO_InitStruct.Speed = GPIO_SPEED_FREQ_LOW;
  GPIO_InitStruct.Alternate = GPIO_AF0_MCO;
  HAL_GPIO_Init(GPIOA, &GPIO_InitStruct);

/* USER CODE BEGIN MX_GPIO_Init_2 */
/* USER CODE END MX_GPIO_Init_2 */
}

[Andrew Neil]

---

@undacmic wrote:

low speed

---

Try high?

[Mike_ST]

Hello,

- SB17 is connected the MCO coming from the ST-Link V2, this 8MHz clock can be used as input clock for the STM32G0 when SB17 is closed.
it has almost nothing to do with MCO of the STM32G0.

- If you're using the HAL library, you can call the
HAL_RCC_MCOConfig(uint32_t RCC_MCOx, uint32_t RCC_MCOSource, uint32_t RCC_MCODiv) function. In theory, it configures the registers automatically.
Example: HAL_RCC_MCOConfig(RCC_MCO_PA8, RCC_MCO1SOURCE_SYSCLK, RCC_MCODIV_1);

- Check you are probing the right pins:
PA8 is on CN9/PIN8 or CN10/PIN23

[undacmic]

---

@Mike_ST wrote:

- SB17 is connected the MCO coming from the ST-Link V2, this 8MHz clock can be used as input clock for the STM32G0 when SB17 is closed.
it has almost nothing to do with MCO of the STM32G0.

---

Oh, I was not aware of this, this really misguided me as I was basing my measurements on this. Thank you.

---

@Mike_ST wrote:

- If you're using the HAL library, you can call the
HAL_RCC_MCOConfig(uint32_t RCC_MCOx, uint32_t RCC_MCOSource, uint32_t RCC_MCODiv) function.

---

I am just using HAL library for benchmarking what I do in my own setup which looks something like this:

static const io_config io_default_configs[IO_PORT_CNT * IO_PIN_CNT] = {
    [0 ... (IO_PORT_CNT * IO_PIN_CNT -1)] = { IO_ANALOG, 0, 0, 0, 0, 0 },
    [IO_USART2_TX] = { IO_ALTERNATE_FUNCTION,  IO_AF1, IO_PUSH_PULL, IO_VERY_HIGH_S, IO_NONE, IO_LOW },
    [IO_USART2_RX] = { IO_ALTERNATE_FUNCTION,  IO_AF1, IO_PUSH_PULL, IO_VERY_HIGH_S, IO_NONE, IO_LOW },
    [IO_USER_LED] = { IO_GPIO_OUTPUT, 0, IO_PUSH_PULL, IO_VERY_HIGH_S, IO_NONE, IO_HIGH },
    [IO_MCO] = { IO_ALTERNATE_FUNCTION, IO_AF0, IO_PUSH_PULL, IO_VERY_HIGH_S, IO_NONE, IO_LOW }
};

The results obtained using the HAL library and my own wrapper are the same.

I was probing the right pins according to the schematics, but what looked strange to me is the signal I was getting on my probe.

 I expected a noisy rectangular signal, and this looks a bit more like a sine wave to me.

I have the probe connected to CN9/PIN8 and I get the GND ref from CN6/PIN7.

[undacmic]

I tried different combinations, it seems like HIGH and VERY HIGH speed are the ones that output my clock.

[undacmic]

UPDATE

I also managed to get the standard deviation of the period cycle of 103 ps for a 64MHz frequency and will follow the same process described in the FPGA implementation here (An embedded true random number generator for FPGAs ) for my RNG implementation and then test it against NIST SP-800 statistical tests.