Posted on May 17, 2011 at 12:51I have been having an issue where my processor will freeze after quite a few cycles of returning from STOP mode.
After looking through some of the files I think I found the problem and it relates to the STM32 library.
Every time the processor comes back from being asleep it goes through the ''SYSCLKConfig_STOP()'' function that is in the example code.
In this function the processor reconfigures the system clock after wake-up by enabling the HSE and PLL. The first thing that is done after the HSE is configured ON is:
/* Wait till HSE is ready */
HSEStartUpStatus = RCC_WaitForHSEStartUp();
In this function we have the following code from the fwlib:
/*******************************************************************************
* Function Name : RCC_WaitForHSEStartUp
* Description : Waits for HSE start-up.
* Input : None
* Output : None
* Return : An ErrorStatus enumuration value:
* - SUCCESS: HSE oscillator is stable and ready to use
* - ERROR: HSE oscillator not yet ready
*******************************************************************************/
ErrorStatus RCC_WaitForHSEStartUp(void)
{
ErrorStatus status = ERROR;
/* Wait till HSE is ready and if Time out is reached exit */
do
{
HSEStatus = RCC_GetFlagStatus(RCC_FLAG_HSERDY);
StartUpCounter++;
} while((HSEStatus == RESET) && (StartUpCounter != HSEStartUp_TimeOut));
if (RCC_GetFlagStatus(RCC_FLAG_HSERDY) != RESET)
{
status = SUCCESS;
}
else
{
status = ERROR;
}
return (status);
}
The problem seems to be that ''StartUpCounter'' is not reset. It is initialized at the top of the stm32f10x_rcc.c file as:
static vu32 StartUpCounter = 0;
Because it is static the value remains and is incremented every time the processor comes back from sleep. After enough cycles, the StartUpCounter value is high (equal to the HSEStartUp_TimeOut value) and will result in a fail.
Please confirm if this is correct. To fix this I initialize StartUpCounter to zero within the RCC_WaitForHSEStartUp() function.
