Legal Removal Request

Post: hi all, as you well know with i2c it is very tricky sometimes. if i meassure external power source, let's say car battery or simply one 18650 which are not used to power rpi, i get a nice accurate readings 1x 18650 multimeter: 4.127V adc_pi readings: 4.1167 but since i meassure the battery,which is powering rpi_zero [they have common GND], then i am not able to get accurate reading. this is the setup: input 12v car battery -> dcdc stepdown --> output 5v --> usb power rpi_zero then i have a voltage divider with R1 = 51k and R2 = 6k8. this is factor 7.5 so for 12V U1 = 10.588 V U2 = 1.41 V simple schema: +12 V --> R1 --> [ +A1 ] --> R2 --> -12 V / [ -A1 GND] all verified with multimeter that math was done correctly. so for 12V betwen +A1 and -A1 i get nice 1.411V with multimeter when rpi is off [i am using same setup to measure 12v and also 5v] i have to have order like 51k first and 6k8 second so i get +A1 in between both resistors [otherwise the voltage will be not 1.4V but +11V due current travel through i2c bus, so it will kill ADC pi] running a simple code: ### from ADCPi import ADCPi adc = ADCPi(0x68, 0x69, 12) #default i2c settings adc.set_conversion_mode(1) samples = 1000 def test(pin): value = [] for x in range(0, samples): value.append(adc.read_voltage(pin)) averagevalue = float(sum(value))/max(len(value), 1) return averagevalue ### >>>a_result = test(1) >>> a_result 1.0171278460000053 # this should be close to 1,41V >>> 12 / a_result 11.797926924517547 #this should be close to 7.5 factor the same error goes for 5V readings. when i verify the ADC accuracy and meassure 3.3V from RPI output, i get nice 3.300895234000026V please, do you have any suggestion on how to improve my circuit for voltage divider to get accurate readings? thank you, pavel