March 2016 13
I hope this gets things moving for you. For the record,
Figure 4 shows what is on the shelf behind me at this very
That’s a completely refurbished Commodore SX- 64
with an SDIDE and an original Commodore 1702 monitor
connected with a Chroma/Luma cable for best picture
quality! I hope I have helped answer your question. Please
let me know if you get your old Commodore to put out
T & L Publications, Inc.
QI have built and used many electronic devices over the years, which have had varying useful ifetimes. I have always wondered how to determine the lifetime of such devices. Could
you explain how to predict the reliability of an electronic
AReliability is a quality which is highly sought after by most designers, builders, and users of electronic and mechanical devices. Reliability is essentially the ability of a device to perform
a specified function for a certain amount of time with a
particular degree of accuracy. Reliability is usually rated in
the ability to resist failure. Failure is anything that prevents
the device from performing its function over the design life
with the desired accuracy. Any discussion of the reliability
of a device must first address what is meant by failure and
any possible failure modes.
For example, a transistor radio fails anytime you
cannot hear a particular station which you have been able
to receive in the past. Failure modes could be due to the
failure of discrete components such as resistors, capacitors,
inductors, diodes, transistors, wiring, speakers, circuit
boards, etc. Each component can then have several failure
modes, such as a resistor can fail open (burned out; also, I
have found a carbon composition resistor invisibly cracked
but it was open and the device would not operate),
shorted (wire wound turns short together), or there is
tolerance drift with time.
Reliability requirements can be different for similar
circuits depending on the end use of the circuit. For
example, if a device is designed as a one-time use device
for entertainment, the cost of improving the circuit’s
reliability may be too much to allow the device to be sold
competitively for a reasonable profit. At the other end of
the spectrum, if the device is embedded into a spacecraft
system on a deep space probe where there is no chance
of repair and the costs of the probe are in the millions of
dollars, the reliability will be maximized to the highest level
The cost of improving can increase exponentially
after we reach a specific level using available technology.
Devices that we build on our workbenches are most likely
not candidates for high reliability. We can improve the
reliability of our circuits by using high quality components
(e.g., mil spec) and careful assembly techniques, but these
will increase the cost of our projects.
Reliability of components is not related to their specific
tolerances. Tolerance is a measure of the accuracy to
which the component is manufactured, where reliability
is a measure of how well the component performs its
function under a given set of conditions for a specified
period of time.
Reliability of components can be visualized as a failure
rate curve as shown in Figure 5, which is often called the
“bath tub curve.” When a large group of components is
put into service, there are many components that fail early
in the life of the device as seen in the “burn-in period” of
the failure rate curve. For critical systems, some electronics
QUESTIONS and ANSWERS
Post comments on this article at
n FIGURE 4.
Typical Failure Rate Curve
l = failure rate during the useful life
Image Credit - www.ece.cmu.edu
n FIGURE 5.