ADC reading correction per VREF_INT

[ECost]

Using STM32F030 Discovery board.

I read the ADC channels including the internal temperature sensor and vrefint.

#define TEMP30_CAL_ADDR ((uint16_t*) ((uint32_t) 0x1FFFF7B8))
#define VREF_INT_ADDR ((uint16_t*) ((uint32_t) 0x1FFFF7BA))

The calibration value of the sensor I get from VREF_INT_ADDR is 0x05F4 or 1524. Calculating the value for a 3.30V value on VDDA I get 1.228V, which is consistent with internal reference voltage per data sheet.

When I read the vrefint channel I get a 0x06A2 or 1698. Calculating the VDDA considering 1.228V I get 2.96V, which is consistent with the VDD voltage I am reading on my Discovery board.

So, to correct the ADC readings taking VDDA into account I have to do the following calculation:

adc_val_corr = (*VREF_INT_ADDR * adc_val)/vrefint;

Is that correct? Am I missing something?

Thanks in advance.

[TDK]

HAL provides some routines to help with this. It sounds like you're doing the same as the routines:

#define __LL_ADC_CALC_DATA_TO_VOLTAGE(__VREFANALOG_VOLTAGE__,\
 __ADC_DATA__,\
 __ADC_RESOLUTION__) \
 ((__ADC_DATA__) * (__VREFANALOG_VOLTAGE__) \
 / __LL_ADC_DIGITAL_SCALE(__ADC_RESOLUTION__) \
 )
 
#define __LL_ADC_CALC_VREFANALOG_VOLTAGE(__VREFINT_ADC_DATA__,\
 __ADC_RESOLUTION__) \
 (((uint32_t)(*VREFINT_CAL_ADDR) * VREFINT_CAL_VREF) \
 / __LL_ADC_CONVERT_DATA_RESOLUTION((__VREFINT_ADC_DATA__), \
 (__ADC_RESOLUTION__), \
 LL_ADC_RESOLUTION_12B) \
 )

[waclawek.jan]

As TDK said, the formula is correct, but the devil may be in details, e.g. rounding at division or overflow at multiplication.

[shameless self-advertisement] my take on the topic [/shameless self-advertisement]

JW