# Current drawn by microcontroller

How to estimate current drawn by microcontroller(on the ADC port) when interfaced to a module/IC/sensor theoretically?

• If you can give us a link to the datasheet for your microcontroller, we can probably show you where this is specified. Dec 9 '18 at 16:01
• – Megh
Dec 9 '18 at 16:05

The parameters of the pin you are interested in are the "LOW-level input current" and "HIGH-level input current". In case of that MCU it is 3 µA.

For the ADC you also have to factor in the sample and hold capacitance which is the "analog input capacitance" (1 pf). Also look at figure 12 "Suggested ADC interface", it shows the ADC input in more detail.

• will it be right to say that this the load current if we are measuring the power consumed by the sensor?
– Megh
Dec 9 '18 at 16:29
• Sensor power consumption and sensor output current are (almost) totally independent. You can have a sensor that consumes 5W to run a heater and still outputs just 1 mA signal.
– filo
Dec 9 '18 at 16:33
• So now if I need to calculate the power consumed by a sensor(specifically analog type temp sensor) interfaced to my controller, I need to find P=VI, where V is the supply voltage and I is the sum of load current and quiescent current at that very temperature. So what should I take as the load current?
– Megh
Dec 9 '18 at 16:39
• Which sensor? Thermocouple? Pt100? LM35? Provide a datasheet.
– filo
Dec 9 '18 at 16:40
• @Megh, I showed how in my answer. Given the specs filo pointed out, though, I am not sure the datasheet is 100% self-consistent, and to be conservative you should probalby use the 3 uA figure. Dec 9 '18 at 16:59

This isn't specified directly in your datasheet, but there are a couple of clues in there.

First, there's a spec for the maximum resistance of the sensor or signal conditioning circuit that drives the ADC:

where Note 8 refers us to Figure 11:

The limit on the source resistance tells us that the maximum input leakage current, when run through a 40 kΩ resistor, won't cause a voltage drop big enough to disturb the ADC reading (by more than the specified absolute error of counts). Taking the minimum VDDA value of 2.6 V, divided by the resolution of the ADC of 1024 levels, this implies

$$I_l < 4 \frac{2.5\ {\rm mV}}{40\ {\rm k\Omega}}$$