Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Decimation causing information loss over spectrum edge

Status
Not open for further replies.

bluespec9

Newbie level 5
Joined
May 17, 2008
Messages
8
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,339
Hi techies,

I am facing information loss in my spectrum after decimation.

My 40MHz (with center frequency 40MHz) spectrum contains 8 sub-channel of QPSK having 5MHz bandwidth (respective center frequency of 22.5, 27.5, 32.5 ...52.5 and 57.5). Sampling freq of this spectrum is 160M.

I am shifting this spectrum to baseband, so after shifting it lies between -20MHz to +20MHz. I am also doing decimation by 4, so effective sample becomes 40M. If I abstract each 5MHz channel from spectrum, I am getting correct information for 4 channel near center but I am not getting correct information for 4 channel near edge.

Any idea, why decimation is disturbing my spectrum over edge?

I know decimation in MATLAB do low pass filtering (passband 16MHz and stopband 20MHz) and then do down sampling. But why lowpass filtering is disturbing my corner spectrum.
 

Hi bluespec9,

You say that the signal in the baseband occupies [-20MHz, +20MHz] spectrum, and then you say that low pass filtering has passband up to 16MHZ and stopband from 20MHz. This means that the transition part of the filter attenuation is from 16MHz to 20MHz, and as I understand all components of your signal from 16MHz to 20MHz will be attenuated (from the smallest attenuation on 16MHz to the biggest attenuation at 20MHz). Right?
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top