Go Down

Topic: Minimum input current for a reliable reading? (Read 744 times) previous topic - next topic


I'm going to be monitoring battery voltage and other voltages which are higher than the MCU's Vcc for my next project. Therefore I'm using a simple voltage divider to get the voltage down to an acceptable level. Of course, I'd like to waste as little power as possible in the divider so I need to know what the minimum input current to the ADC is so that I can adjust my resistors accordingly. The conversions will take place once a second if it makes a difference. I've already tested and found that less than 30uA is sufficient for digital inputs but I'm not sure about analog inputs.


The ADC input is a high impedance input so in theory the current is not important at all.It will drain a very very low current.
Be careful about the input voltage levels.You must insure the input level does not exceed the max input which is about the mcu power supply
Arduino Mega 2560
Arduino Nano
Arduino Duemilanove
MAC OS Montain Lion
Raspberry PI Model B


That's what I figured. But how low is past the point of "not important at all"? Would you say 50uA is good enough?


You need to put a 10 to 100nF capacitor to ground on the relevant analog pin - then the divider can be almost any resistance.

If you are only ever using one analog input this probably doesn't matter, but if you are reading more than one its important the impedance is 10k or less for full accuracy - a capacitor will drop the (dynamic) impedance to a very low value without compromising on resistor values.

Since you are measuring battery voltage the bandwidth can be very low without issues. (the cap and the resistor network form a low-pass filter).

[ When you read different analog input pins an internal multiplexer circuit will switch just before the ADC conversion - the voltage source to the pin then has to charge the ADC's sample/hold capacitor in a few microseconds, hence the 10k or less requirement.  Intrinsically the ADC's input impedance is extremely high at DC, you can treat it as infinite if not switching between pins ]
[ I won't respond to messages, use the forum please ]


Alright then, I'll have a capacitor to ground as well. Seeing as the voltage will change less than 1 volt per hour there shouldn't be any problem with a 1uF cap or even higher, correct? I seem to be running out .1uF caps!

Go Up