Charles,

I believe that the question you are asking depends on your application.
My criteria for determining these types of tests go as follows:

1)  Is this a stress screening on all boards or is this just a
qualification test.

a)  If just a qualification test you might look into a HALT test.  They
seem to be highly effective in breaking things and giving appropriate
data.
b)  Screening tests mean that you need to look carefully at sucking all
the reliability out of the card.

2)  Are you testing the solder quality or functionality of the board and
it's components.

3)  Is the board Unpowered or Powered.

a)  Unpowered tests can allow the board to go to higher extremes in
temperature and this type of test seems to be harder on solder joints.

If I were doing a stress screening test, I would cycle the boards ten
times with the board being powered up.  The ramp rate would be as fast
as the machine could go (typically in the 50 C/min category) from the
minimum temperature the board would ever see to the maximum temperature
the board would ever see.  I would dwell 15 minutes at each extreme.
Depending on what you are trying to find, I would monitor the inputs and
outputs.  If a power supply were on board and the minimum operating
temperature was -40 C or less, I would cycle power (turn off for 10
minutes) at each extreme and make sure that it started up.

Really this is a huge rule of thumb and to do this test correctly you
would have to verify that it was breaking the right things.  This may
cause you to have to cycle a couple of boards 100 times and verify that
your test is effective.

Now if it's a HALT test, the rules are different and I would lead you to
some people that are experts in that field.  However, for those that
want a brief description of what you do in a HALT test here it goes.
Set up your equipment with the power on and verify that everything is
working correctly.  Go cold temperature until module breaks.  Decide if
you need to fix it or if it is good enough.  Go hot until the module
breaks.  Decide if it needs fixed or if it is good enough.  Do
increasing temperature ramp rates until module breaks, decide.  Shake at
room temperature until module breaks, decide.  Shake at cold
temperatures until it breaks, decide.  Shake at hot temperatures until
breaks, decide.  Shake at temperature ramp rates until breaks, decide.
After this is all done, you should know the weak parts of the system and
whether your module design is good or not.

Opinions expressed herein may line up with the planets, are not
necessarily fictional, and may cause managers to groan as they pay for
quality products.

Max Bernhardt
 ----------
From:         Charles Carey
To: [log in to unmask]
Subject: [TN] Industry Conditions for Card Assembly Thermal Stress