In a message dated 5/27/99 1:55:58 PM Central Daylight Time,
[log in to unmask] writes:
<< Terry, I sent this message out awhile back but didn't get any
feed back.
Based on your message I would think you could help answer..
Hello Jesse
I believe the electrical potential has to be DC (?). Does
the speed of
growth increase with increased voltage? Is there a point
where increased
voltage does not increase the speed?
We have been able to show dendrite growth with DC primarily, but also with
AC. This is an area we currently are working at better understanding
(spacing, voltage, current and contaminant level).
From a practical stand point, given two processes, one using
water soluble
flux requiring a wash cycle after soldering and one with a
no-clean flux
requiring no wash cycle, would you say the chance of seeing
dendritic growth
with the no-clean process would be almost nil compared to
the water soluble
process or could the environment seen by the end product
cause the no- clean flux to absorb moisture and therefore
cause dendritic growth?
Well I will answer your question from my project files of actual failures.
We have seen product recalls with WSF processes (high chloride or bromide
activated fluxes) with low standoff parts leaving trapping large amounts of
flux residue over dirty bare boards which the water having a high surface
tension will never clean well. This process lends itself to holding moisture
and contaminants and has metal migration failures even below conformal coated
hardware. But then I have companies with clean bare boards and proper
cleaning protocol with no field performance problems.
But the No Clean assembly with dirty bare boards will also have catastrophic
failures, but clean bare boards (by Ion Chromatography and CSL guidelines)
and excess unreacted or uncomplexed flux will also cause excessive electrical
leakage even typical humidity environments of 50 to 60% RH. But I also have
clients that have No Clean processes for 8 years with no problem either by
having clean boards or by applying some much flux and complexing all of it
that it entraps the residue similar to that of rosin based fluxes
traditionally did. We received so much protection from that clear coat of
varnish (even after any Solvent cleaning) that we never knew we had dirty
bare boards.
So my answer to your question is great processes are understood and
controlled and bad processes are everything else that is unknown.
In a real life case we have seen dendritic growth across 805
size capacitors
under a potential of about 2V-DC. We attributed the problem
to water
entrapment in voids in the surface mount adhesive. We
changed to a new
adhesive with lower humidity absorption, reduced the adhesive
dot size, and
adjusted the cure profile to prevent the voids.
Are you sure that the moisture was trapped moisture from a cleaning process,
or could it have been the monomer of the adhesive driven to the dot center by
a fast hot cure profile creating the voids. Because a skinning effect will
not allow the monomer to escape through the material and drive it into the
uncured material thus creating the voids. Now if the voids created a
channel connecting both ends of the 0805 package and a liquid monomer filled
this channel along with the contaminants from the bare board (incoming
cleanliness) that comes from the fabricators HASL flux and cleaning process
(15-20 ft/min with tap water or worse (Chicago River Water)) then wouldn't it
be the ideal condition to grow dendrites with only 1.5 to 2 volts. We have
already published a case study in Circuits Assembly Process Rx column Dec 98
" Microcanyons"
The follow up question from my boss was
"would there be less risk of dendritic growth with a no-clean
process?"
Electromigration shorts / Electrical Leakage failures on sensitive circuits
are happening at an alarming rate that few people / companies want to admit,
but these critical parameters that ensure good clean processes, materials,
handling and rework procedures can be evaluated and identified as to there
impact on reliability. These steps of the process, and materials can be
qualified and monitored weather it is cleaned or not cleaned.
The process that is at the greatest risk is the process which is not
understood. How clean are the incoming materials, how well are the boards
handled, what effects do the process residues have on the reliability of the
hardware and how is this proven. Many times the increase in No Trouble Found
(NTF) field returns is an indication that there may be a residue problem that
hasn't seen enough humidity yet. This failure will occur because a sensitive
circuit fails due to moisture, contamination, and stray leakage, but drys out
before coming back to the bench for testing and is fine. Hence the issue No
Trouble Found means there still was a problem.
Thanks for the soap box everyone
Terry Munson
CSL Inc.
765-457-8095
################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe: SUBSCRIBE TechNet <your full name>
To unsubscribe: SIGNOFF TechNet
################################################################
Please visit IPC's web site (http://www.ipc.org) "On-Line Services" section for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################
################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe: SUBSCRIBE TechNet <your full name>
To unsubscribe: SIGNOFF TechNet
################################################################
Please visit IPC's web site (http://www.ipc.org) "On-Line Services" section for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################
|
