Single-chip microcomputers can be easy to program

ABSTRACT

Most single-chip microcomputers (MCUs) use the split-memory Harvard architecture. A few single-chips trace their architectural heritage to large computers due to the common-memory Von Neumann organization. The major differences are that a Harvard-based MCU costs less in its undistorted form, and a Von Neumann-based MCU is more expandable and easier to program.

Since the traits of Harvard-based single-chips are quite well known, though perhaps not by that name, the focus is placed on the programming benefits of a Von Neumann MCU. Programming costs can be lowered while increasing program reliability. Data organizations can be more flexible in both RAM and ROM. Program changes can be incorporated more quickly. The generalized instruction set is easier to understand. The M6805 family of MCUs is used to illustrate these benefits.