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.

what's the basic difference between DPCM and DM ?

Status
Not open for further replies.

desperado1

Full Member level 3
Full Member level 3
Joined
Sep 4, 2006
Messages
189
Helped
8
Reputation
16
Reaction score
2
Trophy points
1,298
Activity points
2,503
what's the basic difference between DPCM and Delta Modulation (DM) ?
 

Hi,
DPCM does oversampling i.e., signal are sampled at hte slightly higher than the nyquist rate.
Note:
In oversampling, sampling rate is much higher there the sampling interval is very much lower ( sampling interval is the reciprocal of sampling rate)
i.e., we take samples very closely ( so that the signal amplitude will not vary much in amplitude levels)so the correlation (similarity)between adjacent samples are
very high.
So we deploy predictor ( in the feedback path) which when subtracted from the next adjacent sample gives error signal which is then fed to quantizer .

Actually the error signal will give an idea about how exactly the predictor fails to predict the next adjacent i/p signal.

Hope this will give you clue to proceed further.

Delta Modualtion(DM) makes use of correlation (similar propety of signal with the adjacent samples having done oversampling). It actually increases the correlation between the adjecent samples, so that we can deploy simple quantizer for constructing the encoded signal.

In short DM is a one bit (two level) version of DPCM.


Happy learning
 

    desperado1

    Points: 2
    Helpful Answer Positive Rating
Thanks for the awesome reply. You are damn gud in digital communications. I have one question further ............How can we realize a Prediction filter physically ?
 

Linear predictor is basically a transversal filter ( a tapped delay line), where tap gains are set equal to prediction coefficients.There may be N-order predictors. Larger the N, better prediction we uasually get. but it causes time delay.

we can also realize computationally simple predictor adaptation algorithms for implenting predicting operations for speech coding

Happy learning

Added after 13 minutes:

previously i talk about DPCM
Now reg, DM :
FOr Delta modualtor, we need one-bit quantizer (physically we can employ comparator (two level)followed by sampler i.e., switch) and one accumulator (where we add one-bit quantizer o/p & one bit delayed acimulator o/p).acumulator is implemed phaysically by integrator amplifier.

In delta demodulator we deploy one accumulator followed by low pass filter. ( as i said befor, one can implement accumulator by using integrator amplifier)

Hope it might give some idea.


Happy learning
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top