suggest me any voltage divider circuit with resistors and capacitors to measure a dc voltage of 48v using 12 bit 3.3v ADC...

 

suggest me any voltage divider circuit with resistors and capacitors to measure a dc voltage of 48v using 12 bit 3.3v ADC...

Start by looking up the maximum source impedance for the ADC input as suggested by the manufacturer.
Use this value as the lower leg of your divider, calculate the value of the upper leg using the division ratio 48/3.3

 

I use 62k/2.7k. That gives 2.0V output for 48V input, and allows enough headroom for a 48V battery on equalisation charge. The output of 2.0V is handy as it is the volts per cell of the battery.
The output impedance of the divider (2.6k) is low enough for most A/Ds and high enough that it doesn't need resistors of more than 1/8W.

 

the volts per cell of the battery.

I use 62k/2.7k. That gives 2.0V output for 48V input, and allows enough headroom for a 48V battery on equalisation charge. The output of 2.0V is handy as it is the volts per cell of the battery.
The output impedance of the divider (2.6k) is low enough for most A/Ds and high enough that it doesn't need resistors of more than 1/8W.

do we need to put any capacitors across the tapping(3.3v side ) to compensate noise and any suggestions for its value

 

do we need to put any capacitors across the tapping(3.3v side ) to compensate noise and any suggestions for its value

That depends on quite a few things:
how noisy is the supply
how often do you sample the voltage

If there is any significant signals Above half your sampling frequency, then it needs filtering.

 

A capacitor to ground across the bottom resistor will reduce noise.
It will also slow the response time to a change in battery voltage, but that may not be a particular concern here.

 

Remember, if you want to maintain that 12-bit resolution (or more with the proper sampling), accuracy and precision with higher voltage divider ratios you need to use better, more stable resistors and voltage references for the ADC.
I normally use binary valued external (the internal controller vref is usually marginal) reference voltages like 2.048V for 3.3v system and double that, 4.096V refs for 5v systems with precision metal film resistors where then ADC scaled readings are of high system importance. Be careful using the general logic supply voltage for a ADC vref.
https://www.analog.com/en/resources...ns-for-a-lowcost-sensor-and-ad-interface.html

https://cdn-shop.adafruit.com/datasheets/lm4040-n.pdf
https://www.adafruit.com/product/2200

https://www.mouser.com/c/semiconductors/power-management-ics/voltage-references/?output voltage=2.048 V

 

But do you really need 12 bits for a battery voltage monitor? And if you do, are you willing to spend for a reference voltage worthy of 12 bits? If not, then just find a good stable voltage regulator and some stable 1% resistors plus a capacitor and then do a factory calibration at your MPU so your system reads correctly - and be happy. Tip: using resistors from the same manufacturer and series may further reduce the tempco of your system.

 

The TS didn't actually say it was a battery. I suggested it might be.
I use a LP2951 to supply the microcontroller and use that as the reference. The LP2951 tends to be better than the 1% accuracy the datasheet claims.

 

But do you really need 12 bits for a battery voltage monitor? And if you do, are you willing to spend for a reference voltage worthy of 12 bits? If not, then just find a good stable voltage regulator and some stable 1% resistors plus a capacitor and then do a factory calibration at your MPU so your system reads correctly - and be happy. Tip: using resistors from the same manufacturer and series may further reduce the tempco of your system.

It depends on the battery chemistry. Some have really flat discharge profiles that make voltage to SoC approximations (static and dynamic) require more resolution (I like 10X of the needed measurement unit accuracy requirement) as the battery stack voltage increases.

 

When the voltage discharge curve gets that flat a coulometer is likely a more valuable instrument.

 

When the voltage discharge curve gets that flat a coulometer is likely a more valuable instrument.

For my latest project with a LiFePO4 Battery bank I just track energy in Wh (from energy sensors). I don't really care about or directly track coulombs or Amps as the LiFePO4 battery energy storage is very linear from full to empty and back.