Browse Prior Art Database

Bus Sparing for Fault Tolerant System Design

IP.com Disclosure Number: IPCOM000122398D
Original Publication Date: 1991-Dec-01
Included in the Prior Art Database: 2005-Apr-04
Document File: 2 page(s) / 52K

Publishing Venue

IBM

Related People

Singh, S: AUTHOR [+2]

Abstract

A failure in a driver or receiver circuit or a faulty wire in a common bus are difficult to isolate and therefore difficult to correct in the field. However, by designing spare wires and driver/receivers which can electronically replace defective components, a quick solution is possible. Since such fault occurrences in the systems are few and far between, a small number of wires or wire sets with associated driver/ receiver circuits are more than adequate. If the bus is supported by Error Correction Code, then the sparing can also be used in preventive maintenance.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 97% of the total text.

Bus Sparing for Fault Tolerant System Design

      A failure in a driver or receiver circuit or a faulty
wire in a common bus are difficult to isolate and therefore difficult
to correct in the field.  However, by designing spare wires and
driver/receivers which can electronically replace defective
components, a quick solution is possible. Since such fault
occurrences in the systems are few and far between, a small number of
wires or wire sets with associated driver/ receiver circuits are more
than adequate. If the bus is supported by Error Correction Code, then
the sparing can also be used in preventive maintenance.

      The figure describes a method of automatic bus sparing and is
illustrated by an example of a 4-bit-wide bus. Suppose a driver
labeled (A) develops a fault, which causes malfunction on bus wire
(1).  By writing control vector "0001" in control registers, this
will electronically disconnect the driver (A) and associated bus wire
(1) and substitute spare driver and bus wire (5) in its place.  A
faulty receiver will also be handled in the same fashion. The circuit
structure which allows such a replacement is a simple gate/degate
function controlled by the appropriate bit in the control registers.
There can be many variations on implementations, for example, a set
of wires or set of drivers can be a replacement unit, even though the
failure may impact only one wire.  Such an implementation could be
advantageous in situations where a certain kind of...