On 29 August 1995, Ed Tidwell wrote:
>> I measured some tin/lead thickness on some 20 mil pitch pads (12 mil
>> width, 8mil spacing) on some raw fabs and found the largest majority
>> of the measurements to fall in the 450-600 microinch range (These
>> measurements were made on a SFT 7455 X-ray fluorescence system.)
>> However, there were a number of isolated pads with thicknesses
>> exceeding 850 microinches which were were considerably more 'dome'
>> shaped compared to those pads in the nominal 450-600 microinch range.
>> SMT is decidedly having more problems (i.e., shorts) with the boards
>> with the isolated 'thick' pads, but incoming QA says the boards are
>> within spec. Can anyone share their experiences or data on an upper
>> threshold thickness for tin/lead on 20 mil pitch boards? I'm not sure
>> whether we have a part problem, process problem or both.
>> Ed Tidwell
>> DSC Analytical Lab
>> [log in to unmask]
Ed,
What you mention is something that I've always observed with HASL
PCB's. However, I'm not so sure that's the cause of the shorts.
When printing solderpaste, the stencil thickness and apeture geometry
determines the amount of paste deposited, at least that's the way it's
supposed to work on FLAT pads. With "Dome-shaped" pads, the amount will
decrease according to the extent of the "dome"...('scuse me whilst I fire
up my ascii-cad)....
___ ___ ___ ___
|___| |___| |/#\| |___|<----Printed Paste
___|###|____|###|____|###|____|###|<----Surface Mount Pad w\plating
Normal Normal Dome'd Normal
Pad-1 Pad-2 Pad-3 Pad-4
So what I'm trying to say is, that no matter if there is a dome, that
shouldn't cause you a bridge....unless of course ALL the pads are
severely dome'd and they wind-up blocking the stencil openings so that
you're printing paste on top of an already large volume of solidified
solder. The domes I've seen fit within the apetures and I may see (on
large domes veiwing them under a microscope) the top of the dome running
down the center through the printed paste...this normally hasn't caused
me a problem...it doesn't look good, but it's never been a big deal.
I did have problems with bridging on 20-mil but it was due to soldermask
issues and some other problems. I'll paste my post and the responses at
the end of the email.
Good Luck!!!
////
| @@ / Steve Gregory
C ) Process Training Engineer
////
\_o Pragmatech Incorporated
| @@ / 101 Nicholson Lane
C ) San Jose, California 95134
\_o ////
[log in to unmask]
| @@ / Phone:(408) 943-1151
C ) Fax:(408)943-1461
\_o
-------------------------------------------------------------
....the views expressed above are my own, not my employers...
*************************************************************************
I'm looking for some opinions on two issues that we seem to find very
difficult to convince a customer of ours that they should implement a
change to immediately. It has to do with TSOP Pad Geometries and solder
mask "Dams" between pads on fine pitch components.
The TSOP is a 40-pin rectangular device at 20-mil pitch...(or it
could be 19.7-mils, my eyes don't QUITE have the resolution they used to)
The foot of the gullwing lead measures 10-mils by 20-mils.....and of
course the PCB doesn't have mask in between the pads. The pads themselves
measured out to 10-mils by 70-mils. We've not had much luck with these
parts at reflow.....BRIDGE CITY!!
On the surface when you look at the foot geometry vs. what the pads
are, it would seem that there should be plenty of pad area to disperse
the solder and not bridge. But the leads are so small and don't have much
of a shoulder. Being that we use convection for reflow, we get wicking to
the leads because they heat up so fast and then bridge...BTW, >the PCB is
about 16" X 18" and 16 layers.
The customer is talking about increasing the pad length to
85-mils...we're trying to get them to shorten it. I was wondering, could
there possibly be a reason that we don't know about to increase the pad
length? We were thinking maybe the reason the pads are so much longer
than the current part foot is to accomodate the size variations in
packages from other vendors.
Soldermask dams....same customer.... can't convince them to put
soldermask dams between the pads on their PCB's.....I heard they got
burned a while back from a FAB vendor that didn't do a very good job on
their bare fabs....had mis-registered mask all over everything, pads,
test points, PTH barrels, etc. I guess now they've become a little
"gun-shy" about trying it again....they were burnt pretty bad.
Right now they're using a dry film mask....ick!
What I am asking for is just your opinions...we're trying to do
everything we can to keep from having to put a soldering-iron to the
board during the builds, and feel that the customer is going in the wrong
direction to fix this problem.....I sure would appreciate any feedback at
all!!
__\/__
. / ^ _ \ .
|\| (o)(o) |/|
#-.OOOo----oo----oOOO.-#
# Steve Gregory #
# [log in to unmask] #
# #
#____________Oooo._____#
.oooO ( )
( ) ) /
\ ( (_/
\_)
*************************************************************************
Steve... Keep after them on the DAMS! Too many LPI masks on the market
can do the job on .5mm devices today. Perhaps a "show & tell" of their
product with another similar along with yield data, again their product
vs. another. Probimer from Ciba Giegy is the mask of choice here and we
also offer other varieties including Coates, Enthone and Taiyo! It can be
done!
Regards,
[log in to unmask]
Tom Bresnan
Sr. Product Engineer
*************************************************************************
SMOBC should help with your problem. Also try reducing the amount of
solder paste applied to reduce bridging (try using a 6 mil stencil if you
are currently using an 8 mil). Round your land ends to help flatten tin
lead plating if used. Don't use dry film!!!!!!
Good day!
Edward K. Current
CAD Supervisor
Engineering Services
Power Convertibles Corp.
Ph#(520)628-8292ext.167
[log in to unmask]
|