One of the most widely used fi gures of merit when comparing the reliability of industrial DC power supplies is
MTBF (Mean Time Between Failures). A power supply with a higher MTBF number stated on the specifi cation
sheet may be chosen over a competitor due to a perceived improvement in reliability. Th is notion can be far from
the actual truth. Demonstrating the MTBF through actual fi eld or experimental test data makes the calculated
MTBF more plausible. Th is paper will examine the methods that can be used to correlate MTBF test data with the
calculated values.
