Reading your email I get the impression that this method has caused
numerous problems in the past.  Perhaps you could provide some anecdotal
accounts and really educate us.  Considering that the SIR tests I am aware
of take 1000 hours to complete it would be nice to be able to weed out the
most reactive compounds with the least time and therefore expense.  Then I
can tie up my test equipment with the materials that I think have
demonstrated the highest probability of passing the test.  As you state the
test is probably suited for comparing different materials and this appeared
to be the goal of the original post.  I apologize if you took offense to my
suggestion.

Denis

At 06:05 PM 8/20/97 -0700, you wrote:
>Denis Meloche wrote:
>>
>> An interesting test I have seen performed no clean flux as an indicator of
>> reactability (new word) is to print the no clean product on the comb
>> pattern and then relfow.  Apply power across the comb pattern and the
>> current should be zero with a good no clean (lets hope!).  Place a drop of
>> DI water on the comb pattern.  If the power does not change it is probably
>> a good sign that the product does not absorb moisture and cause EM.  If the
>> circuit breakers blow and the emergency lights come on then you want to be
>> more critical about the product.  This is not intended as a replacement for
>> the longer moisture and temp tests but it is reported to provide a short
>> term indicator of the long term results.  The theory being that DI water is
>> so starved for ions that if any are present they will readily go into
>> solution and cause a circuit path with the water.
>>
>
>The infamous water drop electromigration test rears its ugly head again.
>If you do this test in ordinary air, you can grow dendrites on
>ultra-clean comb patterns.  The DI water will absorb CO2 and other ionogenic
>material from the air and increase it's conductivity anyway.
>In fact, it's easier to grow dendrites on very, very clean, bare metal combs
>than uncleaned combs wave soldered with 15% solids RMA flux.  This tesst is
>only useful in comparing current vs. time curves for combs process with
>different fluxes/materials and or conditions.  Uncleaned, high residues, RMA-
>type fluxes and solder pastes almost always perform better than low residue
>type no-clean materials, even those that perform well in long-term
temperature/
>humidity/bias exposures.  The RMA residue, even though it's full of ions,
>slows the water from contacting the metal surfaces of the comb and allowing
>conduction and electromigration.
>
>
>> At 03:31 PM 8/18/97 MDT, you wrote:
>> >     Hello,
>> >
>> >     Based on my literature research, my preliminary plans would be:
>> >
>> >     1. Obtain SIR comb pattern coupons, run them through the assembly
>> >     process w/o components. Use interested flux and solder wire, reflow.
>> >
>> >     2. SIR test  OR
>> >
>> >     1. Run the product board through the assembly process w/o components.
>> >     Use interested flux and solder wire, reflow.
>> >
>> >     2. Ionic test like Omagemeter
>> >
>> >     Definitely, the second plan is simpler for us. But I am not sure it
>> >     can work. Could someone provide some advices on a simple workable
>> >     qualification procedure for no-clean rework?
>> >
>> >     Another related question:
>> >
>> >     I am aware of 2 options to make test coupons: one is to get coupons;
>> >     the other is to get Gerber file of the test pattern and generate
it on
>> >     the product board. Which one is better?
>
>The second.
>> >
>
>