Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic 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.

Register Log in

who can explain this DSM matlab code to me - I am new for this

Status
Not open for further replies.

6716914

Newbie level 5
Joined
Nov 16, 2012
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,353
%% Simulate the DSM in time
Nfft = 2^13;
tone_bin = 31; what is tone_bin?
t = 0:Nfft-1;

u = 1*(nLev-1)*sin(2*pi*tone_bin/Nfft*t); why the input signal has to be this, what this means
v = simulateDSM(u,Ha,nLev);

figure()
n=1:1000;
plot(t(n), u(n), 'g');
hold on; grid on;
stairs(t(n), v(n), 'b');
title('DSM time-domain Simulation');

%% Plot the Spectrum
spec = fft(v.*ds_hann(Nfft))/(Nfft*(nLev-1)/4); what this means
snr = calculateSNR(spec(1:ceil(Nfft/(2*OSR))+1), tone_bin);
Neff = (snr-1.76)/6.02;

NBW = 1.5/Nfft; why 1.5
f = linspace(0, 0.5, Nfft/2+1);
Sqq = 4*(evalTF(Ha, exp(2i*pi*f))/(nLev-1)).^2/3;

figure()
plot(f, dbv(spec(1:Nfft/2+1)), 'b')
hold on; grid on;
plot(f, dbp(Sqq*NBW), 'm', 'Linewidth', 1);
title('DSM Output Spectrum');
 

guguwuwu

Newbie level 6
Joined
Apr 10, 2012
Messages
12
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,283
Activity points
1,342
I don't know your real means and think of for:
"tone_bin" is a variable to determine the length of the signal.
"u = 1*(nLev-1)*sin(2*pi*tone_bin/Nfft*t)" is an sine signal. The frequency is tone_bin/Nfft.The amplitude is (nLev-1).
"spec = fft(v.*ds_hann(Nfft))/(Nfft*(nLev-1)/4);" is an FFT operation to the signal in bracket.
 

6716914

Newbie level 5
Joined
Nov 16, 2012
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,353
thanks, i know your mean, but i asked why it has to be like that, for example, why the sin signal frequency has to be the tone_bin/Nfft. and for NBW, why NBW=1.5/Nfft
 

monsoon

Member level 3
Joined
Oct 19, 2012
Messages
57
Helped
38
Reputation
76
Reaction score
35
Trophy points
1,298
Activity points
1,598
If you have a signal with periodicity N, and you take N point DFT, then the DFT resolution is 2pi/N. 2pi/N and its harmonics are called bins. If the signal is of the form sin(2pi*31*n/N), it is said to be on 31st bin.

Say you choose sin(wo n) as your input signal. If you intend to take N point DFT, then this signal must be periodic with period N. This implies that wo=2*pi*k/N, where k is any integer. In other words signal must be on a bin. If signal is not on a bin, it leads to what is called spectral leakage and the noise floor rises well above its actual level.

In delta sigma even when the signal is taken on a bin , the output is not periodic because of strong presence of shaped quantization noise. To overcome spectral leakage, output sequence is multiplied by a window ( in this case hann window). Windowing removes spectral leakage.

As for your last question, I am not sure whats going on
 

6716914

Newbie level 5
Joined
Nov 16, 2012
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,353
If you have a signal with periodicity N, and you take N point DFT, then the DFT resolution is 2pi/N. 2pi/N and its harmonics are called bins. If the signal is of the form sin(2pi*31*n/N), it is said to be on 31st bin.

Say you choose sin(wo n) as your input signal. If you intend to take N point DFT, then this signal must be periodic with period N. This implies that wo=2*pi*k/N, where k is any integer. In other words signal must be on a bin. If signal is not on a bin, it leads to what is called spectral leakage and the noise floor rises well above its actual level.

In delta sigma even when the signal is taken on a bin , the output is not periodic because of strong presence of shaped quantization noise. To overcome spectral leakage, output sequence is multiplied by a window ( in this case hann window). Windowing removes spectral leakage.

As for your last question, I am not sure whats going on
Thanks for your reply, my last question was, why NBW equals 1.5/Nfft, is there any explantion for the 1.5/Nfft.
 

6716914

Newbie level 5
Joined
Nov 16, 2012
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,353
For a rectangular window NBW is 1/Nfft, but for a hann window it is 1.5/Nfft. You may refer to page 15 and 16.
ThKs~~~ thanks

- - - Updated - - -

Do you have the other lecture slides about the DSM, I hope I can refer something...

Thnaks
 

Status
Not open for further replies.
Toggle Sidebar

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top