Main Site Documentation

Current leakage on analog input pins


#1

What is the maximum current leakage on the analog pins for the G-400S?

Max current leakage is used to calculate worst case tolerances on inputs. For instance, if you are monitoring a 10V signal, you would typically create a resistor divider to provide a 3V3 max signal to the input. Current leakage at the micro would cause additional voltage drop across the pull-up resistor in the divider, throwing the reading off by an amount relative to the leakage.

The datasheet for the SAM9X doesn’t list a leakage current, it lists an “Input Peak Current” of 2.5 mA, which would be ridiculously high for a leakage current. This would provide up to 2.5V of tolerance even on a resistor as small as 1kohm. I’m hoping I am reading the datasheet wrong somehow…


#2

@ bigtwisty - There is not a specific leakage current value. That would depend upon VREF and the input parameters. Input impedance is the specification you are interested in…

This article is a start.

http://www.planetanalog.com/author.asp?section_id=510&doc_id=559252

Simplest way is measure the input voltage, from your divider, without the ADC connected. Then connect it and see if the divider voltage has dropped to an unacceptable level.


#3

Thank you for the response, Mike.

The input impedance listed for the SAM9X datasheet is 50 ohms. Do I understand correctly that this represents a 50 ohm impedance from input pin to ground? Unless my understanding of Ohm’s law is way off, any resistance in series with the input pin would see significant voltage drop.

Is this buffered at all on the G-400S? How do people typically monitor a signal without significantly impacting it with this type of input impedance? Unity gain op-amps?


#4

I believe the 50 ohms represents a resistor between the measured voltage and a capacitor. To understand what is happening, you would have to read up on analog to digital converters (ADC) and RC circuits.

I would not worry about the impact of the ADC on the measurement. These devices are not made for high accuracy. Take a voltmeter and measure what is happening. If you are not happy with the results, then you would need to use a better external ADC.

Of course, the design of your voltage divider plays an important part in the results. Someone who has worked more with ADCs and voltage dividers can help you if necessary.


#5

Understood.

The device I am designing is an automated exerciser controller for an ECU I am putting through design validation testing in a couple of months. During environmental testing my device will be sending scripted signals and LIN messages to the ECU under test, logging faults and monitoring several signals involved. It isn’t something I can just plug a multimeter in for, as it must be completely automated and log results.

If I can get this working the way I want to, I will be using these SoM’s in more of my future load simulator and exerciser designs.