Basics of HALT, HASS and HAST
- highly accelerated life test, stress screening and stress
testing
The principles of environmental stress screening and life testing are
well established and provide an essential element of many production and
development test programmes. One of the problems with these programmes is
that it can take a considerable length of time for them to achieve their
goals if no steps are taken to speed the results. As a result techniques
including Highly accelerated life test (HALT), highly accelerated stress
screening (HASS) or highly accelerated stress test (HAST) have been
introduced to good effect.
Techniques such as HALT and HASS are not required in every application,
but as many electronic items are used in critical applications and failure
rates must be reduced to the very lowest levels possible. It is in these
applications where stress screening and stress testing are essential.
The need to accelerate
The problem with environmental life testing and stress screening is that it
can take many years for testing programmes to run, and stress screening
times can be well in excess of what is acceptable in a production
environment.
Life testing is an essential part of many development programmes. Once an
item has been designed it is necessary to look at the reliability. It is too
late to wait until the product enters service and then monitor the failures
in the field. However to undertake life testing is not only costly, but it
can add significantly to the development time, and as a result of the fat
development times required by today's electronics industry this is clearly
unacceptable.
As a result techniques such as running the unit under test on duty cycles
much greater than their normal operating duty cycle can help, but this on
its own is rarely sufficient. As the life of a product is often several
years, these timescales are clearly unacceptable in any development
programme today.
A similar situation occurs in a production environment where stress
screening can take several days. Much of the screening is undertaken
automatically and does not need manual intervention, so labour costs may not
be a major issue. Nevertheless the time taken for the units to pass through
the stress screening programme can cause a backlog in production. With many
units passing through the process a significant level of investment is
required to accommodate all the units. Accordingly there are major gains to
be made by accelerating the process, while still maintaining the
effectiveness of the programme.
Developing HALT and HASS programmes
In order to accelerate the life testing and screening programmes, careful
analysis of what is required must be undertaken and the schemes must be
evaluated to ensure they fulfil their requirements properly. It is also
necessary to device programmes for HALT and HASS to be complementary. In
this way no holes are left in the programme, and equally elements of the
development testing are not undertaken in the production cycle, as often
happens with specification testing.
Both HALT and HASS rely on shortening the timescales of their programmes
by applying much higher stresses than exist under normal use conditions.
This forces the failures to occur in much less time than would be the case
under normal operation conditions.
HALT programmes need to find the weaknesses in the design, finding
elements of the unit that will fail as a result of the design. These can
then be re-designed to ensure that they do not fail during normal use. To
achieve this temperature and vibration are used as the chief elements. Other
conditions such as humidity, thermal cycling, burn in, application
over-voltage and over-current conditions as well as anything else that might
conceivable occur during use. By using the units continually, with a much
higher duty cycle than would normally be achieved, this will assist in
identifying the weak points of the design. Any problems that occur must be
re-worked in order to ensure that they are able to withstand the rigours of
use.
HASS programmes have a different aim. Its purpose should not be to
highlight design problems, but instead to highlight any latent defects in
the unit that have arisen during the production process before the unit
leaves the production environment. In this way the infant mortality failures
should have all occurred. Obviously if a particular failure occurs then this
can be investigated by the design team. These may occur as a result of a
design failing, or a production process that needs changing.
The HASS schemes employ high stressing of the unit, often beyond the
qualification level, but not to the extremes that are applied during the
HALT programme. Typically scenarios such as temperature (with fast
temperature cycling) and sometimes some vibration are used. It is also
normal to run the units during HASS, and to monitor their operation so that
failures can be identified as soon after they occur as possible. Once the
unit has completed its HASS programme, it should then undergo a full
production test to ensure that it fully meets its specification. The
monitoring of the unit undertaken during the HASS test itself is normally
limited as a result of access, and the equipment required to fully test all
the units undergoing HASS.
Determining limits for HASS
It is extremely important to determine the correct limits for the stress
screening process. If too little is undertaken then some latent failures may
remain in the unit under test. If the conditions that are applied to the
unit are too great then additional failures may be introduced and there is a
risk that the overall reliability may be impaired.,/p>
To determine the limits the minimum and maximum operating temperatures
must be determined. Burn-in may be accelerated by running the unit at high
temperature, and also at a high duty cycle. Temperature cycling is also
effective in revealing latent mechanically related problems such as poor
solder joints etc.
Additionally the vibration conditions must be investigated. Typically
some vibration is applied, but this is limited and kept to within the
operating limits of the unit. As part of this process it is necessary to
simulate the transport conditions, as transporting units often brings out
defects.
In some programmes, humidity testing may be applicable. This can be very
useful in finding latent shorts. It is also helpful in accelerating the
ageing process, and as a result it is a powerful technique used in HALT.
Equipment
When developing a HASS programme, it is necessary to take account of the
expected equipment throughput. The very nature of HASS means that numbers of
units will be run through the programme together. To enable the throughput
to be optimised it is necessary to have a good idea of the planned
throughput. In this way it is possible to select he correct size for any
environmental chambers and also construct the equipment required to operate
and monitor the equipment under test.
It is obviously ideal to monitor the equipment as it is undergoing HASS.
However the fact that a large number of units are operating together and
often in an environmental chamber can restrict the testing that can be done.
Techniques such as the extensive use of built in self test (BIST) can help.
It may also be necessary to develop and construct test heads that can be
placed close to the equipment under test, as it is not advisable to subject
the test instrumentation to continual HASS as well.
A number of problems will need to be addressed at the early stages of the
development of a HASS programme. These include the dissipation of power
generated by the units being tested. This can be a particular problem for
power supply units. Running these can generate large amounts of power. If
this is dissipated within the chamber it may not be possible for the chamber
to maintain the temperature correctly. Even the power dissipated by the
units under test themselves may cause a problem if there are large numbers
of them.
Other problems may be encountered by running large numbers of units in
close proximity. Electrical noise, combined with the long lead lengths often
needed to make measurements remotely is a particular problem as they will be
more susceptible to pickup.
Monitoring schemes will also need to be developed to identify any
failures during the process. There should also be adequate protection to
ensure that if any units do fail in a catastrophic manner, then damage
further damage is not caused to the unit, other units near by or the HASS
equipment.
As it is seldom possible to undertake a full test of a unit during HASS,
it is often necessary to complete a final test of the unit to ensure that
after the stress screening it still meets its specification before being
shipped.
Summary
A integrated HALT / HASS programme can bring significant benefits in many
circumstances. However to achieve the full benefits, careful planning and
continued monitoring of the system are vital. Only in this way will it be
possible to feedback information into the development and production
programmes to ensure that the most reliable product its shipped.
|
