You can not use Tin or Tin/Lead as a contact material. In humidity and air
an oxide film grows which will continually raise the contact resistance over
time. ATT years ago did
a study of metals for physical contacts and found those two platings
unsuitable.
WE used some contact connectors on PWAs that had solder coated fingers. The
vendor claimed that a hermetic seal would form in the contact from the
connector pushing into the solder. We had all kinds of electrical problems.
The solder contact had to be perfectly flat and flat within a specified area
in order for the contact to maintain good conductivity and this was in an
air conditioned office environment. We had to do Ammonium Sulfide and Nitric
Acid vapor exposure tests to guarantee that we had a hermetic seal.
-----Original Message-----
From: Ed Holton [mailto:[log in to unmask]]
Sent: Thursday, March 16, 2000 3:06 PM
To: [log in to unmask]
Subject: Re: [TN] Electromigration demonstration...
My first experience with dendretic growth involved a capacitor on a board
in a highly humid environment (automotive application). We had dendretic
growth underneath the capacitor. There were a few variables that helped
aggrevate the situation. I did a simple experiment to show everyone what
was happening, but it also showed that even if your board was perfectly
clean, as long as there was water and voltage, you will get a dendretic
growth.
I hooked wires onto each end of a 1206 capacitor. These wires were then
hooked to a 12v power supply (to simulate a car battery). A drop of DI
water was placed ontop of the capacitor and the power supply turned on.
After a few moments, a dendrite jumped from one end of the capacitor to the
other. It was pretty cool to watch. This little experiment earned me the
nickname Mr. Wizard for the next few months. We had the video camera from
our microscope system focused on the capacitor so everyone could watch it
on the monitor. We did find in later repeat experiments that it was
sometimes necessary to add water, as it sometimes evaporated quickly due to
the conditions. (I need to thank John Maxwell, he help me set the
experiment up originally)
Ed Holton
Manufacturing Engineer and Group Leader
Hella Electronics
Telephone (734) 414-0944
Fax (734) 414-0941
KK Chin
<KK.Chin@ARTE To: [log in to unmask]
SYN.COM> cc:
Sent by: Subject: Re: [TN]
Electromigration
TechNet demonstration...
<[log in to unmask]
ORG>
03/16/00
02:13 PM
Please
respond to
"TechNet
E-Mail
Forum.";
Please
respond to KK
Chin
Steve,
Dr. William Kenyon responded to my technet posting last year with the
demonstration method which was of great help to me. The passage is pasted
below.
The article was also posted in his column in the SMT magazine, April 99.
K.K. Chin
Artesyn Technologies
Fremont, CA
----------------------------------------------------------------------------
-------------------
There is a simple EmR (electromigration resistance) test you can
do -- I put it in a recent (April) SMT column that I write each
month on cleaning issues.
Here it is, just copied and pasted from the last edit.
Fun with Dendrites-
Readers have expressed interest in the reliability concerns
caused by dendritic
growth, or cathode-anode filament formation. Whatever the name,
it usually
results in catastrophic in-service failure of the PCB in service.
Since many PCB
fabricators are reluctant to make any process changes based on
the possibility
of an obscure failure mechanism, as long the bare and assembled
PCBs pass the
ionic cleanliness test, it seemed appropriate to provide a simple
demonstration
of the phenomenon that assemblers can use to get the attention of
their PCB
suppliers.
There Here is a simple, inexpensive test that anyone using
hot-air solder
leveled (HASL) or infrared fused PCBs can do. First, obtain a
scrap board that
has been through PCB supplierUs HASL process. Find two parallel
traces on the
PCB that are at least 5 cm long. Cut the traces at both ends to
isolate them
from the rest of the circuitry. Solder some 5 cm. wires to the
positive and
negative terminals of a standard 9 V battery, then attach
mini-alligator clips
to the ends of the wires. Using a sharp knife straight-edge
razor, peel up the
opposite ends of the cut traces, so you can attach the
mini-alligator clips to
the bent-up ends of the traces. (One could solder the battery
wires to the
traces with type-R rosin, but this could introduce doubt in the
results. The
clips or a connector if the design has wires going to a set of
Fax To: Amanda OUBrien @ SMT
Magazine Page 4/5
connector fingers is better.) Obtain some DI water. and a decent
10-30x
microscope. If a microscope is not available, a powerful
magnifying glass can
be used. Put the isolated traces under the microscope and bring
them into focus.
Arrange the PCB to see both traces and the laminate gap between
them. Next, hook
the 9 V battery terminal wires to the opposite ends of the
isolated traces; so
one trace is positive and the other is negative. Now, very
carefully place a
drop of DI water on the isolated traces so it bridges the gap
between the traces
in the field of view of the microscope. See how long it takes for
the dendrites
to grow from cathode to anode, timing the progress with a stop
watch if desired.
Usually one sees the drop of water start to bubble a little, and
then a metallic
tree-like growth will cross the gap from cathode to anode. The
resulting
metallic dendrites are quite brittle, which probably accounts for
the well-known
fix of banging a defective PCB on the table and having it spring
to life again.
What has happened is that all the dendrites (current leakage
paths) have been
shattered. Will it happen again over time, especially with
exposure to elevated
humidity conditions? Absolutely. This test will make believers
out of those who
think all one has to do is pass the ionic cleanliness test and
that pure water
never hurts anything. The problem with HASL or infrared fused
tin/lead plated
PCBs is was typically the organic fluid that forms the basis of
the traditional
HASL or reflow fluids. It only took takes a little on the
surface to give the
current leakage, resulting in the dendritic growth problem. The
only pass/fail
value available that is based on test work is about 1
microgram/sq. cm per
0.001" of trace or pad separation. So a 0.050S space can
tolerate 50 mg., while
a 0.010S space can tolerate 10 mg. According to this source,
most PCBs will
have hundreds to thousands of mg./sq. cm of such fluid residues
on them.
Literature data are is available on the effects of various flux
residues and
cleaning methods on test PCBs that had such fluids on them or had
any traces of
these fluids removed prior to the soldering and cleaning tests.
Newer fluids
either eliminate or minimize this problem, but the test is so
easy to do on
scrap boards that it is good insurance prior to accepting a lot
of PCBs and
releasing them to production.
##############################################################
TechNet Mail List 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 web site (http://www.ipc.org/html/forum.htm) for
additional
information.
If you need assistance - contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
##############################################################
##############################################################
TechNet Mail List 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 web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
##############################################################
##############################################################
TechNet Mail List 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 web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
##############################################################
|
