I have sampled my transmitted signal, it's looks like sinusoidal , frequency sampling is =2048Khz, bitrate of transmittion=100Kbaud , and my problem is that I have designed in matlab my LPF filter which it's :


and I got a problem that my filter is moving my DC signal away from zero, how could I do my LPF filter to not move the DC of my signal after filtering ? In other words, I want to do LPF filter to my signal but not moving the DC of my filtered signal and it should be as before filtering (in my case DC of my signal is approximately zero, assume it zero) but after filtering it my DC signal moves aways from zero. input signal: enter image description here attaching here my filtered signal after it passed my designed LPF:

enter image description here

what should I do to my LPF designed in order to not moving my DC of filtered signal comparing to my Original signal before filtering?

About LPF filter, what should it length be in order to get good results? how many bits it should be?

atttaching my LPF plot: enter image description here

  • $\begingroup$ -1, there isn't enough here to help. Include input and output signals as well as frequency responses. That will likely answer your own question. $\endgroup$ – Engineer Jul 16 '20 at 16:27
  • $\begingroup$ @Engineer I attached $\endgroup$ – LiamLony Jul 16 '20 at 16:34
  • $\begingroup$ Hi ! I don't know really why it's for you shwing HPF, but here's another matlab syntax using designfit simulink, LP_FIR = dsp.LowpassFilter('SampleRate',Fs,... 'DesignForMinimumOrder',false,'FilterOrder',N,... 'PassbandFrequency',Fp,'PassbandRipple',Ap,'StopbandAttenuation',Ast); $\endgroup$ – LiamLony Jul 17 '20 at 19:37
  • $\begingroup$ but whatever Im not asking why it's this ! , asking how can I build my LPF for not letting my DC of my signal moving after filtering .. $\endgroup$ – LiamLony Jul 17 '20 at 19:37

As provided pics show, there is no difference between the DC value of input and output signal of the filter i.e. they are both zero.

but there is a problem in your filter design and that is, you set the 'FilterOrder' key to 1. I think its better to remove 'FilterOrder' key and its value and let the algorithm choose the smallest filter order for you.

As for number of bits it's about implementation and it's not an issue in high level programming languages like Matlab.

Note: when a LTI filter has zero DC value signal in its input then the signal at its output has zero DC value no matter what is the frequency response of your filter.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.