COP8SA-DM National Semiconductor, COP8SA-DM Datasheet - Page 19

• Manufacturer Part Number: COP8SA-DM
• Description: MODULE DEBUGGING FOR COP8SA
• Manufacturer: National Semiconductor

Specifications of COP8SA-DM

• Module/board Type: Debugger Module
• For Use With/related Products: Cop 8
• Lead Free Status / RoHS Status: Lead free / RoHS Compliant
• Other names: *COP8SA-DM

The single address bus of the Von Neumann architecture is used sequentially to access
instructions from program memory and then execute the instructions by retrieving data
from and/or storing data to the data memory. This means that an instruction fetch cannot
overlap a data access from memory.
The obvious advantage of a Von Neumann architecture is the single address bus and
single data bus linking memory with the CPU. A drawback is that code can be
inadvertently executed from data memory, opening up the possibility for undesired
operation due to corruption of the program counter or other registers.
1.5.2
Harvard Architecture
The Harvard architecture was named after the Harvard Mark 1 and the early
electromechanical computers developed at Harvard by Howard Aiken, another computer
pioneer. This architecture has separate program memory and data memory with a
separate address bus and data bus for each memory. One of the benefits of the Harvard
architecture is that the operation of the microcontroller can be controlled more easily in
the event of corrupted program counter. A modified (enhanced) Harvard architecture
allows accessing data tables from program memory. This is very important with modern
day microcomputers, since the program memory is usually ROM or OTP/EPROM) while
the data memory is RAM. Consequently, data tables usually need to be in program
memory so that they are not lost when the microcontroller is powered down.
The advantage of a modified Harvard architecture is that instruction fetch and memory
data transfers can be overlapped with a two stage pipeline, which means the next
instruction can be fetched from program memory while the current instruction is being
executed using data memory. A drawback is that special instructions are required to
access RAM and ROM (or OTP/EPROM) data values, making programming more
difficult.
1.6
HOW DOES A MICROCONTROLLER OPERATE?
The CPU can request information from memory (or read an input port) by calling it by
its memory address. The address with all its bits is stored in the CPU as binary number
in a temporary data latch type memory called a register. The outputs of the register are
sent over multiple wires (or single wire) to the microcontroller memory and peripherals.
The group of wires (parallel) or the single wire (serial) that carries the address is called
the address bus. The word “bus” refers to one or more wires that share a common path
to/from multiple places. The address register holds address bits. The number of address
bits depends on the microcontroller type.
Data is sent to the CPU over a data bus. The data bus is different from the address bus
in that the CPU uses it to read information from memory or peripherals and to write
information to memory or peripherals. Signals on the address bus originate only at the
CPU and are sent to the other blocks attached to the bus. Signals on the data bus can
either be inputs to the CPU or outputs from the CPU. The information on the data bus is
sent or received at the CPU by the data register. In other words, the data bus is bi-
directional and the address bus is uni-directional. The width of the address and the data
bus may also be different, depending on the microcontroller type and memory size.
MICROCONTROLLER BASICS
1-7