I ran into this problem in a previous job with discrete chips being placed on HASL boards. We traced the problem back to paste being squeezed out during part placement. During reflow the paste was trapped under the component body and would not wick back into the solder fillets. 
The homeplate itself didn't prevent paste from still being squeezed out but it did reduce the amount of paste trapped. Found that we had to go to a modified homeplate design - we had to move the inside edges of the homeplate actually back from the edges of the component pad.

Regards
Tom

-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Steven Creswick
Sent: Friday, October 05, 2012 10:11 AM
To: [log in to unmask]
Subject: EXTERNAL: Re: [TN] solder splash under SMT caps

Phil,

You will likely get more expert opinions than mine, however, I would also add old/expired/improperly stored solder paste to the list as well as excessive placement force during P&P.

Steve C

-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Phil Nutting
Sent: Friday, October 05, 2012 9:56 AM
To: [log in to unmask]
Subject: [TN] solder splash under SMT caps

Here is my question for today.

We have an SMT board that measures 16 inches by 17 inches be 0.062 thick (it is well supported in the end application) with 32 voltage multiplier circuits so the board develops 12.5kV.  On the 2225 capacitors we had failures (arcing) under the parts and our customer removed some caps only to find solder splashes.  We used the standard IPC pad configuration for a 2225 part (3.7mm gap) with no solder mask under the capacitor, ENIG finish and I suspect tin/lead paste.  I have a photo I will post if Steve Gregory is still offering his site.

Our customer wants us to use a specific capacitor by AVX and the AVX pad layout of 4.6mm gap. Our contract manufacturer wants us to open up the gap between pads by 1mm (essentially the 4.6mm recommended by AVX).

My theory on this failure is the following; (could be any one or a combination of any items listed below) Too much solder paste Incorrect reflow pre-heat Incorrect liquidus temp

I have read John Maxwell's white paper and see that he recommends a larger space, but he didn't see significant performance enhancement above the 1808 part.  Refer to
http://johansondielectrics.com/technical-notes/application-notes/new-impact-
of-pad-design-and-spacing-on-ac-breakdown-performance.html.

I agree that this greater distance will help with arcing, but will it help with the solder splash?

Will separating the pads have that much affect that the solder will be drawn out to the outside visible pads?

Inquiring minds want to know....

Phil Nutting
Design for Manufacturing Engineer
Kaiser Systems, Inc.
126 Sohier Road
Beverly, MA 01915
Phone: 978-922-9300 x1310
Fax: 978-922-8374
e-mail: [log in to unmask]<mailto:[log in to unmask]>
www.kaisersystems.com<http://www.kaisersystems.com>
www.linkedin.com/in/philnutting<http://www.linkedin.com/in/philnutting>


________________________________
Note: All the information contained in this e-mail and its attachments is proprietary to Kaiser Systems, Inc. and it may not be reproduced without the prior written permission of sender. If you have received this email in error, please immediately return it to sender and delete the copy you received.


______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] ______________________________________________________________________


______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] ______________________________________________________________________

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________