RTC with LSE not keeping time

[gerrygralton]

Hi,
I'm having a bit of trouble with my STM32F411RE not keeping time. I have an 32.768kHz output from a DS3231M chip being using as the LSE source. I am finding that the internal stm32 RTC and the DS3231M times are drifting substantially (approx 1 minute per day). This was not something I expected as they should be sharing the same clock signal.

My current setup for the STM32 RTC is very simple:

    STM32RTC & rtc = STM32RTC::getInstance();
    rtc.setClockSource(STM32RTC::LSE_CLOCK);
    rtc.begin();

I'm wondering if I need to set up the STM32 RTC in bypass mode. Is this possible with this library? Am I even on the right track?
Thanks!

[Frederic Pillon]

Hi @gerrygralton
I have a STM32F407VE with LSE crystal and do not have this issue. Looking at the ref man:

So I guess you should configure it as bypass. To do this simply redefine the SystemClock_Config

[gerrygralton]

Thank you! It's good to know that we're at least on the right track. Could you please give me an example of the configuration code? Or where to find the current configuration from this library?

[Frederic Pillon]

You can use STM32CubeMx to generate the configuration and follow the Wiki link I gave in my previous comment.

Anyway, here the modified one of the Generic_STM32F411RE to add in your sketch which will be used instead of the default one:

extern "C" void SystemClock_Config(void)
{
  RCC_OscInitTypeDef RCC_OscInitStruct = {};
  RCC_ClkInitTypeDef RCC_ClkInitStruct = {};

  /** Configure the main internal regulator output voltage
  */
  __HAL_RCC_PWR_CLK_ENABLE();
  __HAL_PWR_VOLTAGESCALING_CONFIG(PWR_REGULATOR_VOLTAGE_SCALE1);

  /** Initializes the RCC Oscillators according to the specified parameters
  * in the RCC_OscInitTypeDef structure.
  */
  RCC_OscInitStruct.OscillatorType = RCC_OSCILLATORTYPE_HSI | RCC_OSCILLATORTYPE_LSE;
  RCC_OscInitStruct.LSEState = RCC_LSE_BYPASS;
  RCC_OscInitStruct.HSIState = RCC_HSI_ON;
  RCC_OscInitStruct.HSICalibrationValue = RCC_HSICALIBRATION_DEFAULT;
  RCC_OscInitStruct.PLL.PLLState = RCC_PLL_ON;
  RCC_OscInitStruct.PLL.PLLSource = RCC_PLLSOURCE_HSI;
  RCC_OscInitStruct.PLL.PLLM = 16;
  RCC_OscInitStruct.PLL.PLLN = 192;
  RCC_OscInitStruct.PLL.PLLP = RCC_PLLP_DIV2;
  RCC_OscInitStruct.PLL.PLLQ = 4;
  if (HAL_RCC_OscConfig(&RCC_OscInitStruct) != HAL_OK) {
    Error_Handler();
  }

  /** Initializes the CPU, AHB and APB buses clocks
  */
  RCC_ClkInitStruct.ClockType = RCC_CLOCKTYPE_HCLK | RCC_CLOCKTYPE_SYSCLK
                                | RCC_CLOCKTYPE_PCLK1 | RCC_CLOCKTYPE_PCLK2;
  RCC_ClkInitStruct.SYSCLKSource = RCC_SYSCLKSOURCE_PLLCLK;
  RCC_ClkInitStruct.AHBCLKDivider = RCC_SYSCLK_DIV1;
  RCC_ClkInitStruct.APB1CLKDivider = RCC_HCLK_DIV2;
  RCC_ClkInitStruct.APB2CLKDivider = RCC_HCLK_DIV1;

  if (HAL_RCC_ClockConfig(&RCC_ClkInitStruct, FLASH_LATENCY_3) != HAL_OK) {
    Error_Handler();
  }
}

[gerrygralton]

Thank you!

[Frederic Pillon]

Let me know if it works.

[gerrygralton]

No, this hasn't worked unfortunately. Over three hours, the external DS3231M chip gained 12 seconds on the stm32 RTC.

[Frederic Pillon]

Unfortunately, I have no more input.
I will test longer with my F407 board but I'm confident it will be OK.
Maybe due to your wiring? or something bad with the hardware.

[gerrygralton]

Thanks for your help anyway Frederic. I'll update this if we establish the root cause.

[gerrygralton]

I think we have established the cause. The DS3231M chip has a 32.768 kHz accuracy of +- 2.5 % which easily explains our 1 minutes per day drift.

[Frederic Pillon]

Thanks for the feedback