After building 4000 PCA's for a customer, which has 14 Case Size "A"
Tant. Caps out of approx 400 parts on the wave side, it was found that
PCA's averaged 14 "No Solders"/PCA after wave. (50% joints of 28 in
random positions) Almost all the defects were on these Case Size "A"
component types. Case size "D" Tant Caps and 805 ceramic parts were
defect free. What is unique about Tant SM caps is that the leads only
travel half-way up the body , unlike ceramic chips which terminal
surround the ends. The Case size "D" had larger lands. Occasionally
defects for excess solder were found as a result of trying to increase
the  solder flow to these difficult parts. (eg slower feed rate, higher
chip wave etc.  The SOLTEC wave soldering machine uses a foam fluxer (No
Clean flux) with good foam height and bubble size. An experiment to hand
spray flux reduced the qty of defects to 4/PCA, but did not make the
process defect-free. (Yes we know foamers are not the best kind)
To make the process Defect-Free and specifically free of "No Solder"
joints, we measured a sample of exposed solder lands and checked a
variety  OK and "bad (NS)" joints. The result was there were 0 joints
with 0.060 (1.5mm) exposed lands that failed.  Since it was too late to
change the pad design for this production run, we considered 1) do
nothing except continued 100% inspection, 2) hand-solder both ends, 3)
offset the part, leaving only one joint requiring hand solder.  This was
considered better than leaving inspection to chance. To offset the parts
0.5mm or 20 mil on the Fuji placement offset, would provide 100%
probability of lead soldering . The pad size allows contact with the
0.8mm (32mil) terminal length with the offset and then manual soldering
of only 1 joint is required. This is expected to reduce the process
variable of Case Size lead solderability to zero and make this part of
the process defect-free. 

Other Opportunities
====================
 Interesting enough we theorize this solution will fix SOT-23 "No
solder" problems.  I've read the history of people who have had problems
on this part. If the land is big enough for the solder to adhereit will
flow to the part. If you can,  increase the land size to accomodate your
wave soldering process or in the Size "A" Tant Case here .060" or 1.5mm.
This doesn't apply to ceramic chips where the termination surrounds the
end of the part and wets easily. 

I don't put too much faith into IPC-782 design stds. They
are only guidelines.  Determine your own optimum LAND Designs and
provide input to your customer's design. I'd like to get more DFM
details from the Solder Land committee. Comments?

EXPERIMENTAL DATA for Case Size "A" Tantalum caps.
<land length>               
Exposed Lands   Inspected Joints to IPC class2
[thou]  [mm]    % bad   % OK
======  =====   =====   =====
0.000   0.00    100%    0%
0.010   0.40    100%    0%      
0.020   0.80    50%     50%
0.030   1.20    50%     50%     worst case 
0.040   1.60    33%     67%     this is IPC782 guideline size
0.050   2.00    43%     57%
0.060   2.40    0%      100%   defect free on several batches inspected
0.070   2.80    0%      100%
                                
0.018   0.73            Kemet MIN. Land Length  Case size "A" Tant Caps
0.031   1.23            Kemet NOM. Land Length  
		        Note that KEMET does not have a MAX!! 
0.040"  1               IPC782A MAX Land Length (calculated) 
                        Haven't checked Rev B yet
0.060   1.5 mm          C-MAC MIN Land Length for No clean flux with 		
foam fluxer on SOLTEC wave

RSVP by e-mail if you wish.

-- 
regards,
Tony

Eng Manager     		Web:      http://cmac.ca
C-MAC Electronic Systems Inc.   Phone:    (204) 633-7200
1455 Mountain Avenue            Direct:   (204) 631-7030
Winnipeg, MB, Canada R2X 2Y9    FAX:      (204) 631-7294

[log in to unmask]

***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To subscribe/unsubscribe send a message <to: [log in to unmask]>   *
* with <subject: subscribe/unsubscribe> and no text in the body.          *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact   *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask]      *
***************************************************************************