Posted on April 12, 2018 at 01:41
STM32L151R8 - I'm using the Injection channels with a 4 channel sequence. The conversion is handled by software and is repeated immediately (on a different set of data). This works fine for a while. Eventually, the second polling loop fails with a timeout. After some debugging, I found it always fails on the second polling loop. Also, when it fails, no amount of waiting will ever make it pass. (TheJEOC bit never goes high).
I am using the regular ADC channels as well, but I don't believe this problem is coupled with the regular channels. I disabled the trigger to the ADC channels, so no conversions are occurringduring this code. This made no difference.
Injection code:
assert(HAL_ADCEx_InjectedStart(&hadc) == HAL_OK);
NOP();
assert(HAL_ADCEx_InjectedPollForConversion(&hadc,5) == HAL_OK);
for(i=0; i<ADC_INJ_NUMCONV; i++) {
data1[i] = HAL_ADCEx_InjectedGetValue(&hadc,ADC_INJECTED_RANK_1+i);
}
// Change external MUX to put new data into injection inputs
assert(HAL_ADCEx_InjectedStart(&hadc) == HAL_OK);
NOP();
assert(HAL_ADCEx_InjectedPollForConversion(&hadc,5) == HAL_OK); //Fails!
for(i=0 ; i<ADC_INJ_NUMCONV; i++) {
data2[i] = HAL_ADCEx_InjectedGetValue(&hadc,ADC_INJECTED_RANK_1+i);
}�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?
Note, as a workaround, I tried replacing the polling all with the following:
void inject_wait(void) {
while(HAL_ADCEx_InjectedPollForConversion(&hadc,2) != HAL_OK) {
assert(HAL_ADCEx_InjectedStart(&hadc) == HAL_OK);
}
}�?�?�?�?�?�?�?�?�?�?
This basicallykeeps trying to start the injection conversion, until it succeeds. This never causes an assertion, but the data conversion is not done correctly. What I see is data from the previousconversion. It's like the injection converter needs a break between executions (?).
