STM32f103c8T6 inaccurate ADC readings

[DSimp.1]

Hello, I'm using stm32f103c8t6 board and trying to get simple voltage divider readings. I'm also using multimeter to measure voltage and well I'm getting 0.2 sometimes 0.4 V error. Is this normal? I really need precision, 0.2 or 0,4 volts are really to much for me. I'm using regular conversion on 1 channel every 10 seconds with maximum 239.5 cycles sampling time. This is the code I use (yes I try to calibrate it and that doesn't work)

HAL_ADCEx_Calibration_Start(&hadc1);

HAL_Delay(10);

HAL_ADC_Start(&hadc1);

HAL_ADC_PollForConversion(&hadc1,100);

ADC_VAL=HAL_ADC_GetValue(&hadc1);

HAL_ADC_Stop(&hadc1);

voltage=ADC_VAL*3.3/4096;

#[STM32 MCUs]​ #ADC​ #STM32F1​ 

[TDK]

Multimeters take their reading by long term averaging. Perhaps the signal you're reading is oscillating.

[DSimp.1]

nope, I'm trying to measure Li-ion cell's voltage through voltage divider and it should be around 2.7 volts I get 2.4-2.8 volts and sometimes 3.2 volts for no reason. Are stm32f1's ADC really so inaccurate ? Maybe the problem is that I bought this mcu on alliexpress, so maybe it's faulty or something like that?

[LP4PTA]

Have the same problem.

Caps have no effect to stabilize voltage reading so only conclusion is that AD is inaccurate.

To make it worse, even after I average the reading by software and use a calibration factor the average float over time.

Back to PIC?

#Actual Voltage = 3.1V

Vout(0) = 329
Vout(1) = 330
Vout(2) = 329
Vout(3) = 329
Vout(4) = 329
Vout(5) = 332
Vout(6) = 328
Vout(7) = 328
Vout(8) = 329
Vout(Max) = 332
Vout(Min) = 328
Delta = 4
Vout(Average) = 329.00
Vpin = 0.266V