SparkFun Forums

[roach]

I'm desiging a project that involves using all 8 ADC pins on the ATMega32. I've read in various places that EMF interference is to be avoided at all costs when using the ADC. Since this will be on a motor-controller board, near some pretty "loud" motors, I want to isolate the traces from the input pin headers to the uC somehow. Apart from keeping these traces as short as possible, and using twisted-pair or shielded cable for the inputs, is there any other strategy I can employ that will still give me the full 0-AREF voltage range on these pins?

thanks in advance,

[leon_heller]

I'm desiging a project that involves using all 8 ADC pins on the ATMega32. I've read in various places that EMF interference is to be avoided at all costs when using the ADC. Since this will be on a motor-controller board, near some pretty "loud" motors, I want to isolate the traces from the input pin headers to the uC somehow. Apart from keeping these traces as short as possible, and using twisted-pair or shielded cable for the inputs, is there any other strategy I can employ that will still give me the full 0-AREF voltage range on these pins?

thanks in advance,

Keep the analogue and digital grounds separated, and filter the analogue supply. Details are in the data sheet.

Leon

[roach]

Thanks Leon! I checked the data sheet, and they give recommendations, but I'm not sure I'm completely grasping the concept of a seperate Ground plane for AGND.

Electrically, how does AGND connect to GND? I mean, obviously, they need to be seeperated, but at some point they have to connect. How is this typically done?

thanks again.

[leon_heller]

It's usually done with a small piece of track, joining them at one point. This is a good document on the technique:

http://www.microchipc.com/PIC_tips_3.asp

However, ADI has a different view:

http://www.analog.com/library/analogDia ... ry/12.html

I've never bothered doing it with the AVR, just filtering the analogue supply works well enough.

Leon