Good afternoon - is there ever an easy answer that will satisfy everyone? I
believe all test coupon conditions need to reflect all real board
situations. Example - conductor spacing would have to change depending on
base copper used ( say 4 oz. ) for good soldermask flow and unique thickness
if required by customer. Soldermask should be continuos flow across pattern.

regards - Ken

-----Original Message-----
From: Rene Martinez [mailto:[log in to unmask]]
Sent: Tuesday, August 01, 2000 10:32 AM
To: [log in to unmask]
Subject: Re: [IPC-600-6012] Solder Mask coupon coverage


You have a point there.  The failures often occur where the trace somes out
of the soldermask and that condition is not covered on the coupons.  The
coupon should support, large copper areas and also traces with partial and
full coverage.  Are we looking at a new coupon design?

-----Original Message-----
From: C. Don Dupriest [mailto:[log in to unmask]]
Sent: Monday, July 31, 2000 12:24 PM
To: [log in to unmask]
Subject: Re: [IPC-600-6012] Solder Mask coupon coverage


OK here is more detail on the solder mask failure situation.

We experienced Solder Mask adhesion failure on a board, but not the coupon.
The current coupon (flooded continuous with mask) has no true material
transition area.  It only transitions under the mask from copper to laminate
and back and forth within the hatch geometry.  The actual mask material does
not transition as you would typically see in a real design (say the web area
around certain component features).  The failure in this case is at the
transition from "solder mask" to "bare opening" to "over copper" and back to
"clear opening".  Follow that?    A tape test over these areas on the actual
board reveals mask failure (mask removing onto the tape) locally at that
point.  A tape test over the coupon and over continuous circuits revealed no
failure.  Therefore I think we may have a Test Method that is not optimized
for today's demanding features.  I think we may want to explorer a "G
coupon" design that includes a masking transition from "clear" (void of
mask) to "covered laminate" to "covered copper" and so forth.  Need a
picture to explain better.

When speaking with someone of our Materials Test Lab background they agreed
that is why they gang scribe samples in paint adhesion test methods to
include a separation in the paint into smaller features to check true
adhesion.   When we test over a large continuous area even though it is
going over copper and base laminate it gains adhesion strength due to the
larger surface area.

Does that make sense.


        C. D. (Don) Dupriest
        Lockheed Martin Missiles and Fire Control - Dallas
        Electronics Manufacturing Engineering
        Mgr. - PWB Process Development
        Ph. 972/603-7724 fax: 972/603-3548
        Email: [log in to unmask] <mailto:[log in to unmask]>


        ----------
        From:  Rene Martinez [SMTP:[log in to unmask]]
        Sent:  Monday, July 31, 2000 12:45 PM
        To:  [log in to unmask]; [log in to unmask]
        Subject:  RE: [IPC-600-6012] Solder Mask coupon coverage

        We ask for a continuous flood of material.  I think it makes more
sense because you ned to test the transition areas from laminate to traces.
We also use these coupons to measure thickness--- instead of the Y coupon.
I am not sure if it is better or not.  What is your opinion?

        Rene
        -----Original Message-----
        From: C. Don Dupriest [mailto:[log in to unmask]]
        Sent: Monday, July 31, 2000 9:52 AM
        To: [log in to unmask]
        Subject: [IPC-600-6012] Solder Mask coupon coverage


        Greetings all,

        A question has come up concerning the solder mask coverage over the
"G"
        coupon (cross hatch).  After reviewing in 3 locations TM (2.4.28.1)
; 2221
        (12.4.6) ; and SM-840 (6.4) no mention is made to the distinct type
of
        coverage over the coupon.  The question is should the coverage be a
        continuous flood over the hatch or should the coverage be a matching
hatch
        pattern over the copper only.  The thinking being that an adhesion
problem
        would be more readily detectable if the tape test were pulling over
solder
        mask squares over copper vs. pulling over an entire pattern of
laminate and
        copper.  Any thoughts on this.  Our spec wording just states
coverage not
        the type of coverage.  Anyone have a thoughts on this as too what
would be
        the most accurate test method.  Do we need a hatch over laminate and
a
        separate hatch over cooper to be most accurate?

        Thanks for any input,

        PS. The G coupon is up for review in the 2221 amendment 2 (fig
12-15) also.

                C. D. (Don) Dupriest
                Lockheed Martin Missiles and Fire Control - Dallas
                Electronics Manufacturing Engineering
                Mgr. - PWB Process Development
                Ph. 972/603-7724 fax: 972/603-3548
                Email: [log in to unmask] <mailto:[log in to unmask]>