There is a very large database which clearly indicates that
subjecting electronic hardware to thermal cycling and random
vibration is a very effective method of eliminating workmanship and
component defects and providing the ultimate user with very reliable
equipment. See Reference 30. This process has been defined as
Environmental Stress Screening (ESS). A case history of the thermal
cycle screen and resultant measured reliability of a high-
reliability, low-voltage power supply is presented in "Thermal Cycle
Screening of Militarized Power Supplies." See Reference 25.
RECOMMENDED ESS PROCEDURE
The minimal requirements for performing ESS on power supplies
should be implemented in the following sequence:
(1) Parametric testing
(2) Random vibration with continuous monitoring
(3) Thermal cycling with continuous monitoring
(4) 100% parametric testing per the customer Acceptance Test
Procedure (ATP).
The above ESS process applies to all low-voltage power supplies
and high-voltage power supplies. High-voltage power supplies should
be subjected to the additional ESS detailed in Section 5.5 of this
publication.
RANDOM VIBRATION
There are two general types of random vibration recognized for
the screening process; i.e., pure random and pseudo or quasi-random.
Either type is acceptable. Random vibration should be done prior to
performing the thermal cycle portion of ESS. The power supply should
be operational, fully loaded (resistive) and continuously monitored.
The profile and levels of the random vibration are approximately 6
grams (guidelines are provided in NAVMAT P-9492). The power spectral
density versus frequency is illustrated in that publication. The
vibration should be applied to a minimum of two axes for 10 minutes
per axis. At least one of the axes should be perpendicular to the
plane of any PWB/MIBs in the power supply.
The data gathered during random vibration screening of
militarized digital avionics, and used in the development of these
requirements, was examined in "Random Vibration Screening of Six
Militarized Avionic Computers." See Reference 26. A comprehensive
guide to ESS is provided by "Environmental Stress Screening for
Assemblies." See Reference 27.
THERMAL CYCLING
Twelve thermal cycles are required in accordance with the profile
in Figure 4-1. The last two cycles must be failure-free. The
temperature limits are:
(1) High Temperature
The chamber ambient conditions should be such that the average
temperature measured at the surface of the components of the power
supply is a maximum temperature of +55ºC during fully operational
(power-on) conditions.
(2) Low Temperature
The chamber ambient conditions should be such that the average
temperature measured at the surface of the components of the power
supply is a minimum temperature of -55ºC with the power off.
(3) Thermal Rate of Change
The average rate of change of temperature at the surface of the
components should be at least 5ºC per minute, with 15ºC per minute
providing optimum results.
(4) Thermal Stability
The power supply should be considered to have reached thermal
stability at the high- or low-temperature portions of the thermal
cycle when the component with the greatest thermal inertia has a
rate of change of less than 5ºC per hour, or is within 5ºC of the
ambient temperature of the chamber. To achieve this stability in as
short a time as possible, the cooling fans should operate
continuously during the cool-down part of the cycle.
In order to validate that the above thermal conditions are
achieved, a thermal survey should be performed under the required
temperature and duty cycle to identify the component of greatest
thermal inertia and establish the time/temperature relationship
between it and the chamber air.
The 3-hour thermal cycle shown in Figure 4-1 is a typical example
for power supplies. The power supply should be turned on (energized)
under full load (characterized) conditions at the start of the
temperature transition from low to high temperature; this is defined
as the cold-start condition. Except for planned on/off cycling
lasting no more than a few minutes, the power supply should then be
operated continuously until thermal stability is reached. During the
transition from high to low temperature the power supply should be
turned off and remain de-energized until the start of the transition
from low to high temperature.
Provision for ESS temperature ranges should be incorporated in
the design specification. For the purpose of ESS, it is required
that the power supplies be turned on at the end of the cold soak
period of the thermal cycle. This provides the maximum thermal shock
and stimulates the failure of weak components and elements. It has
been observed that most failure mechanisms occur during the
transition, with power on, from low to high temperature. The power
supply would not be expected to operate within its performance
criteria until its temperature rises above that specified for
power-on operating conditions by the customer specification.
The power supply should be turned off for no less than one
minute, at least four times during the thermal cycle: twice when the
chamber ambient is between low temperature and 0ºC, and twice at
high temperature after thermal stability has been reached. The input
power should be applied and removed in an abrupt manner, as by a
relay or mechanical switch.
POWER SUPPLY LOADING
Each output of a power supply should have a static
(characterized) load equal to the maximum rated load. This load
should be applied to the power supply during all ESS.
SPARES
Power supplies and/or their assemblies that are delivered as
spares should be subjected to the same ESS as the original
assemblies.
FAILURE DATA AND CORRECTIVE ACTION
A reliability program should be implemented to collect and
analyze the data from ESS, determine the failure mechanism and
implement corrective action. |