DECIMATING DIGITAL DATA

[CESSNA1]

HELLO EVERYONE: What are the advantages and disadvantages of decimating digital data. Say we take data at 100 hz and decimate it by ten to 10 hertz. What problems can be expected of solved? One that comes to mind is aliasing. If decimation effectively moves the sample rate below the filter frequency then we couold have aliasing problems.

What is your understanding of this?

Regards

Dave

 

[GregLocock]

The advantages are enormous, but you have oversimplified the processing.

You need digital AA filters as well. Look at the architecture of the old HP Paragon A/D boxes, they sample at 256 kHz (?) after fixed frequency (100 kHz?) input AA filters and after that everything is digital, even if you only want 10 Hz sampling frequency.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[tomirvine]

I agree with Greg that you need digital AA filters as well.

Decimation with filtering can be performed in you are only interested in the signal's frequency content well below the sample rate.

Decimation may also be performed to reduce the size of the data files for storage or post-processing purposes.

Tom Irvine

http://www.vibrationdata.com

 

[CESSNA1]

GREG AND TOM: Thank you for your responses. I would like to clarify one thing. I think you are saying to use a different AA filter when than the data is decimated than when it is orignially taken. For example, If we use, say, a 100 Hz filter to acquire data; then I beleive you are saying we shoudl use, say, a 10 hz filter when it is decimated. Is that true?

Thank you and Regards
Dave