NUCLEOL476RG read battery level using ADC.

[stm32f4]

Posted on May 28, 2018 at 17:21

Hello ST All.

I have a NUCLEOL476RG and i have some troubles trying to read a battery level using ADC.

I made a voltage divider partitor to get 3.3V starting from 14.1V.

1)  using 162K resistor + 50K resistor the adc read 2.9v

2) using 68.8k + 20.78k the adc read 2.6v.

Using multimeter on partitor in both read 3.20v .

If i read 3.3 from Nucleo (CN7  pin 2 or 4) adc read correctly.

please, Anyone may helm me with hints?

Regards,

Carmelo

[S.Ma · Accepted solution]

Posted on May 28, 2018 at 18:04

ADC input is high impedence. Then ADC sample the voltage by connecting a RC to the ADC input pin for a while, then disconnecting the input pin, and convert. When connecting the RC, the cap maybe discharged...

The input impedence should be small (ideally a supply voltage), and here it is very resistive, the RC curve slows down and is cut before reaching the asymptotic ideal voltage.

==> If this situation is correct, increasing the sample time to maximum will yield a value closer to the expected one.

==> If the required sample time shall be higher than possible, you could stop the ADC clock during sampling phase to 'freeze the time'.

==> Alternatively, put a decoupling cap to ground (10uF?) at the ADC sampling point. Battery level changes slowly so the response time of the divider maynot be critical...

[T J]

Posted on May 29, 2018 at 02:36

decoupling cap should be closer to 1uF or even 0.1uF,

your problem is the high impedance.

if you require a fast response, then you should use an external OpAmp to drive the pin with low impedance.

this will give you a much faster changing voltage on the pin from the source and a rapid read capability.

[stm32f4]

Posted on May 29, 2018 at 08:54

Thanks a lot.

i, as you suggested , increase sample time