Browse Prior Art Database

Apparatus for Decreasing Delta Modulation Distortion

IP.com Disclosure Number: IPCOM000081915D
Original Publication Date: 1974-Aug-01
Included in the Prior Art Database: 2005-Feb-28
Document File: 3 page(s) / 59K

Publishing Venue

IBM

Related People

Machol, GK: AUTHOR [+2]

Abstract

In delta modulation, it is desired to minimize the tracking error between a time varying analog input applied to the modulator and the representative stream of "1's" and "0's" generated therefrom. Delta modulation is a coding process. It requires a periodic sampling of a waveform and comparing the sampled package magnitude with a reference. If a sample value exceeds the reference then a positive signal is generated. If the reference exceeds the sampled magnitude a negative signal is generated. The reference value is, in turn, incremented by a value opposite in phase to that of the modulator output.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 3

Apparatus for Decreasing Delta Modulation Distortion

In delta modulation, it is desired to minimize the tracking error between a time varying analog input applied to the modulator and the representative stream of "1's" and "0's" generated therefrom. Delta modulation is a coding process. It requires a periodic sampling of a waveform and comparing the sampled package magnitude with a reference. If a sample value exceeds the reference then a positive signal is generated. If the reference exceeds the sampled magnitude a negative signal is generated. The reference value is, in turn, incremented by a value opposite in phase to that of the modulator output. It is well appreciated, for example, that if a stream of successive 1's or 0's were generated, then it could be concluded that the sampled waveform either always exceeded the reference value or the reference value always exceeded the sampled value. In either case, the coder would not be closely tracking the signal.

It is recognized that better tracking can be obtained if it is possible to either anticipate the trend of the input signal or find the same type of information in the output of such a system and make adjustments accordingly. As applied to delta modulators, the prior art has taught that tracking accuracy is improved by changing the sampling frequency or the step size (the amount of unit change algebraically added to the reference value each time an output signal is generated).

Hereinafter, in the described apparatus the delta tracking error can be minimized by complementing the modulator output bit stream during the ith + 1 time interval, if the error occurring during the ith interval was greater than a unit step. Because the apparatus monitors the modulator error and inverts the appropriate pulses in the modulator output bit stream, then it is not necessary to either modify the sampling frequency or the step size.

Reference is made to Fig. 1 where the feed forward detection and correction logic for the delta coder tracking error is set forth.

Delta coder 2 is responsive to analog input 1 over path 3 by generating a binary 1 or 0, i.e., Delta or Delta out from flip-flop
7. Difference amplifier 5 causes flip-flop 7 to generate a Delta out at each sampled interval when the analog input on line 3 exceeds the algebraic sum of the Delta values obtained from integrator 9 on line
11. Similarly, flip-flop 7 generates Delta when the magnitude on line 3 is less than the algebraic sum on line 11.

Tracking error "e" is representatively shown in Fig. 2 as the instantaneous magnitude difference between the input signal and the integrated output signal. So that there will be no ambiguity, the signals labeled Delta or Delta represent sequences of 1's and 0's in the digital sense, and also refer to magnitude values generated by flip-flop 7 when applied to integrator 9.

Referring to Fig. 2, there is shown the tracking by the delta coder of the time varying analog input over a succession of sa...