Proposal

Delta-Sigma Modulation
for
Analog to Digital Converters
on
Integrated Circuits

The Engineering Honors Committee
119 Hitchcock Hall
College of Engineering
The Ohio State University
Columbus, OH 43210

by
John Sheridan Fisher
137 West Norwich Ave.
Columbus, OH 43201-4348

October 14, 1996


Abstract

This honors research project will explore Delta-Sigma Modulation (DSM) of analog to digital (A/D) converters. DSM is an extremely fast, accurate methodology for converting between analog and digital, as well as a highly innovative and new idea in A/D technology. This project will encompass the complete setup and operation of a Motorola 56000 evaluation system including the EVB board with on-board DSM A/D conversion. This system will be used to explore DSM in a stand alone manner and to evaluate other systems. Once the Motorola 56000 system is working properly, various discrete DSM A/D converters will be built and evaluated. Hopefully, this is where DSM response improvements will be made. Eventually, all of this work will migrate to an integrated circuit. This integrated circuit will become part of a much larger project, a programmable analog integrated circuit.

Introduction

The purpose of this honors research project is to explore Delta-Sigma Modulation of A/D converters. This is highly innovative and new idea in A/D technology. Much remains to be learned in this area, especially in the application of Delta-Sigma Modulation (DSM) on an integrated circuit (IC). In order to further develop the IC design of DSM, some precise test systems will be required. Consequently, this project will encompass the complete setup and operation of a Motorola 56000 evaluation system including the A/D board with on-board DSM A/D conversion. This system will be used to explore DSM in a stand alone manner and to evaluate other systems. The possibility exists that this system will also be used as an emulation station for a DSM system depending on the course of the project. Once the Motorola 56000 system is working properly, various discrete DSM A/D converters will be built and evaluated. ideally, this is where DSM response improvements will be made. Eventually, all of this work will migrate to an IC.
This integrated circuit will become part of a much larger project: a programmable analog IC. Programmable digital ICs are commonplace, but only in the past few years has the idea of a programmable analog IC surfaced, let alone the whole field of mixed signal design. There is a one programmable analog IC on the market made by IMP. Their IC does not have an on-board A/D converter; this research project will work toward placing the A/D on IC using DSM. Hence, the overall project will design and implement a complete analog IC that will be competitive with the IMP chip.
A design of this sort has seemingly never been attempted. There are inherent problems of speed, frequency response, and timing in trying to integrate an A/D on an IC. However, the possibility of creating a complete analog system on an inexpensive single IC is compelling. The applications are endless, and given the recently achieved level of technology, this project can now be attempted.
In working on this project, the researcher hopes first-and-foremost to learn. There is also a vested interest in gaining experience in research. This experience can highly valuable to anyone either in graduate school or as a part of the work force. The methodology used to solve problems that have never been investigated is a highly valuable tool in any business. In addition, the researcher would like to explore a personal interest in research to determine if this is a field of true interest.

Objectives

The objectives of this research project are threefold and integral to the completion of the overall project. Firstly, the Motorola 56000 evaluation system must be prepared for testing and evaluation of other systems. Secondly, various discrete DSM A/D converters must be built. These may then be tested using the Motorola 56000 system. Finally, the IC characteristics and limits modeled by the discrete A/D systems must be improved such that the once on the IC, the A/D is not creating a lag in computation while still providing robust frequency response. Obviously on a single analog IC, there is a absolute need for fast, accurate, and robust A/D converters. Normally, it is the connection between IC chips slows the system. That connection has been integrated to increase speed. However, A/D converters are historically slower devices, thus the use of DSM. DSM will provide the speed and IC characteristics required.

Methods of Procedure

As previously stated, this research project has three parts. The first is, of course, being to prepare the Motorola 56000 evaluation system. This undertaking includes not only hardware but also software. There is a Motorola 56000 evaluation board hardware manual, and there is also a Motorola 56000 EVB board hardware manual. Both of these will be required reading to prepare the Motorola 56000 hardware. This hardware should be ready near the beginning of Autumn of 1996. Additionally, there is a Motorola evaluation board software manual which discusses the actual code used by the Motorola 56000 evaluation board, and there is a Motorola 56000 evaluation board C compiler manual. Both of these are necessary reading in order to run the tests with the Motorola 56000 evaluation system. Of course, various C programs will have to be written for the Motorola 56000 system. These should be finished near the beginning of Winter of 1997. Once the evaluation system has been brought on-line, the various test programs will be implemented to prepare to evaluate the discrete DSM A/D converters.
These discrete systems will not only be built and evaluated, but also simulated. This process should begin in the middle of Autumn of 1996. Matlab, Spice, and Mentor Graphics will be used for simulation. Matlab will be used to evaluate the frequency response of the theoretical transfer equations derived from the designed circuits. Spice will be used for its robust, realistic frequency response evaluation techniques of discrete components from the Spice library. Mentor Graphics will be used to simulate the A/D system because of its powerful ability to feed complicated wave forms into a designed system. It could also be used to design the complete analog IC and integrate the various portions together including the DSM A/D.
Once plausible systems have designed on computer, they will be built and evaluated using the Motorola 56000 evaluation system. The researcher would like to reach this point by the beginning to middle of Winter of 1997. This will begin the long, circular process of design, simulate, implement, evaluate, and redesign. It is in this area that this project may make the most significant contributions to the area of DSM. Various existing resources will be used including, but not exclusive to, Digital Signal Processing in Telecommunications by Kishon Shenol, Analysis and Design of Analog Integrated Circuits by Gray and Meyer, Discrete-Time Signal Processing by Oppenheim and Schager, Modern Digital and Analog Communication Systems by B.P. Lathi, and Microelectronic Circuits by Sedra and Smith. Once a final DSM A/D has been completed, the design may be laid out in Magic. This would should happen near the middle of Spring of 1997. Creating this layout will enable a direct implementation and integration of the DSM A/D into the final analog IC design.

Capability

The researcher has successfully engaged the following free elective classes specifically for the purpose of doing research in this area of Electrical Engineering: EE600, Digital Signal Processing; EE601, Communication Theory; EE650, Linear Systems with Random Inputs; EE721, Advanced Electronic Circuits; EE722, Active Network and Logic Circuit Design by Digital Computers. In addition, the researcher has completed all of the required courses for the major which gives a strong skill base from which to work. Further, the researcher has had four co-op positions, one of which provided exposure to DSM A/D in implementation. The researcher feels that the only component of the undergraduate educational experience that has not been completed is research. This research project will facilitate that interest and involvement in research.

Please any questions, comments, or corrections to John Sheridan Fisher

Last Modified February 14, 1997