I have what I hope is a simple question. Can I adjust the output impedance of an IC by simply placing a resistor in series/parallel to the analog In pin on the Duem?

I'm playing around with the ADC clock and have set the prescalar to 4 giving me a sampling time of 4Mhz. Using a function generator with a 600 ohm output impedance with a sine wave (DC offset ofcourse) set to 5kHz I don't see any loss in accuracy of the reproduced sine wave.

Ultimately what I'm doing is interfacing a MAX4372 to sample current and I'm trying to maximize the sampling time to get accurate readings of current spikes. I wanted to do some testing to see if accuracy was a function of output impedance. I know the datasheet for the Atmel says it's optimized for circuits with an output impedance of 10K or less but doesn't specifically say whether or not accuracy is correlated to impedance. I wanted to play around with and try and collect some data for future projects.

Oh btw, setting the ADC prescaler to 2 doesn't work, it constantly outputs 1023.

Can I adjust the output impedance of an IC by simply placing a resistor in series/parallel to the analog In pin on the Duem?

Wouldn't that be input impedance then?

Simply no you don't change the impedance of a pin by doing this. You do however change the impedance seen by the pin, which is something quite different.

I don't see any loss in accuracy of the reproduced sine wave.

So how much accuracy do you expect to be able to see, if you work it out it will be not that great.

but doesn't specifically say whether or not accuracy is correlated to impedance

It's not, or at least not directly. If you have a high impedance source feeding into the analogue input then it won't be able to charge the sample and hold capacitor fully before the digitization phase starts. An in that respect what you measure will not be a full reflection of the voltage on the input. Note this is not what people normally mean by accuracy but it is what you are meaning by it.

Quote: but doesn't specifically say whether or not accuracy is correlated to impedance

It's not, or at least not directly. If you have a high impedance source feeding into the analogue input then it won't be able to charge the sample and hold capacitor fully before the digitization phase starts. An in that respect what you measure will not be a full reflection of the voltage on the input. Note this is not what people normally mean by accuracy but it is what you are meaning by it.

Just for future reference what would the correct term be? (as I type this the word distortion just popped into my head).

Quote: Can I adjust the output impedance of an IC by simply placing a resistor in series/parallel to the analog In pin on the Duem?

Wouldn't that be input impedance then?

Simply no you don't change the impedance of a pin by doing this. You do however change the impedance seen by the pin, which is something quite different.

That's exactly what I want to do, change the impedance seen by the Arduino ANAI so I can where if any distortion starts to occur in the signal at given frequencies with respect to changes in impedance.

For my current project I'm not too worried since the output impedance of the MAX4372 is only 1.5 ohms. I'm just one of these people that comes across a question while doing something and likes to get as much info as possible and do some testing (probably why it takes me forever to get things done). Thanks again for the help.

Just for future reference what would the correct term be?

Actually I am not sure because you are talking about minimizing any input error. Accuracy refers to the voltage step than can be discriminate. This is a good site for terms:-
http://www.maxim-ic.com/appnotes.cfm/appnote_number/641/CMP/WP-36

If you put a 10K resistor in parallel with the analogue input it will see 10K impedance. However the thing driving the input has to have a low impedance to be able to drive that 10K. If not you loose voltage because your source canâ€™t drive the impedance so you are no better off.