Steve,
Thanks for the sound advice. I agree that a bullet-proof process and
set-up procedure will produce the results necessary for high yield and
that metal blades will help reduce variation. The situation I have seen
here is that metal blades can, and will, get damaged in our facility (I've
even seen poly blades get nicked!). In addition, I have seen several
stencils get damaged, through careless handling or otherwise. The result
is inconsistent print volume, leading to solder defects. The one area
where I can't completely reduce variation is with operator skill, or
diligence in following common sense practices. For example, soon after I
got here I was troubled by the fact that boards were allowed to slide down
a ramp at the end of reflow, sliding into each other and piling up.
Unbelievably, I was questioned by some as to "why" I requested an
operator at the end of reflow at all times. Some glued on components were
actually being sheared off.
The main problem we face here in arid western Colorado is the huge swing
in RH and temperatures (swamp cooling in summer & gas heat in winter), and
the very narrow process window of our current paste (Multicore NC-40). In
the winter, RH can be around 8% at 70 - 76 deg F, and in the summer RH is
up to 70% with temps up to 83 F. We have a significant problem with paste
dry out on the boards and in the stencil apertures - sometimes a delay of
less than 20 minutes can cause a problem. Many of the current solder
pastes may eliminate this problem, and I am working on changing the paste
- once I can get through the customer approvals and all that non-technical
stuff. Until we can get the process in control, we must monitor the
prints and reduce every variation that we can. Any feedback out there on
enclosed printheads, ie ProFlow to help with our indoor climate problem?
Howard Watson
Manufacturing Engineer
AMETEK/Dixson
"Stephen R. Gregory" <[log in to unmask]>
Sent by: TechNet <[log in to unmask]>
07/31/01 09:19 PM
Please respond to "TechNet E-Mail Forum."; Please respond to SteveZeva
To: [log in to unmask]
cc:
Subject: Re: [TN] Solder Paste Measurements
Hi All!
Just want to "muddy the waters" a bit...
How many of you that use metal squeegees, and pretty much have your
stencil
thickness down (as far as specifying thickness for a given pitch) have
found
that it is really "crucial" to measure your paste thickness?
I for one, have found that when I spend the time and the money on
equipment
to measure paste thickness, when using metal squeegee blades, and having a
solid set-up procedure, that I'm spending a bunch of time and money
measuring
things that are always good...meaning that time could be spent better
elsewhere as long as you use metal squeegee blades, and have a good
procedure
in place ensures that the printer is set-up correctly..which is not rocket
science...
Am I over simplifying things? It's always worked for me since the advent
of
metal squeegees...I don't know how you can go wrong with a metal blade.
Look
at the gerbers you're given, see if they match the pad geometries, and
then
get the stencil made. Things are pretty straight forward after that...
-Steve Gregory-
Howard, in the past, I was forced to use to use and invest in smaller,
less
expensive systems (I've since been lucky to have nice EXPENSIVE
toys-teehee). One of my favorite tools has always been the benchtop
Cyberoptics unit. I don't remember the model name, I'm sure it's on their
website, or a price figure. But I do know it was cheaper than most
systems.
The beauty of it is it shoots a laser at an angle, laser is viewable on
monitor and is deflected by paste height. You line up the reference lines
on the screen and get pretty true paste height. If your using 1:1
apertures
or if you're doing aperaure reductions, you can calculate your volume.
Hope I've helped.
Jason Gregory
SCI Systems, Inc.
Software Specialist - NPI Group
(256)882-4107 x3728
[log in to unmask]
>>> [log in to unmask] 07/31/01 11:49 AM >>>
Hi Howard,
I've always found measuring solder paste volume to be very difficult. Yes,
numbers can be generated, but even taking large amounts of data doesn't
guarantee the correct impression. Generally, regardless of equipment used,
the best results one can hope for are for trends, which at best could be
described to be of the order of -1 or +1 from true. It's not a bad thing
to
try but don't expect absolute results.
Best regards,
Andrew Hoggan
BBA Associates Ltd
www.bba-associates.ltd
-----Original Message-----
From: TechNet [mailto:[log in to unmask]]On Behalf Of Howard Watson
Sent: 31 July 2001 14:47
To: [log in to unmask]
Subject: [TN] Solder Paste Measurements
Dear Technetters,
My question is: how do you measure solder paste height and volume
accurately? My situation is with using the VisionMaster Model 150A, I am
having trouble obtaining reasonably accurate measurements. The
VisionMaster
is a small bench topsystem that uses a template consisting of a "region of
interest" (the solder paste), and reference regions (the areas on each
side
of the pad). Because there are inconsistencies in the PCB (HASL finish),
like raised areas from traces, valleys surrounding the pad, and
irregularities in the solder mask, the readings I get cannot possibly be
true. Many times the measured readings for weighted average height are
over
7.5 mils using a 6 mil screen and 9.5 mils using an 8 mil screen. The
process specifics are Multicore NC-40 paste, shore 94-97 polyurethane
blades, DEK 265 Infinity, and correct squeegee pressures, print gap, etc..
I believe the bricks are good, I just ca! n't use the measurements for SPC
as it shows the process to be out of control.
In theory, polyurethane blades should "scoop" if anything, leaving a
shorter brick than the stencil thickness. It seems logical to me that the
best way to obtain accurate measurements would be to use the pad as a
reference region and measure the height from the pad, but the VisionMaster
system does not allow me to do this. Does anyone have knowledge or
recommendations on what I can do here? How do the more expensive systems
measure solder paste? Oh yea, spending $$ on new equipment is not a real
good option at this point!
Thanks in advance for the assistance,
Howard Watson
Manufacturing Engineer
AMETEK/Dixson