We have been stumped by occasional opens at the corner balls of some BGA components after reflow.
We have experimented with many minor adjustments, yet this occasional problem continues. Do you have any suggestions?
Based upon the warp and twist characteristics of the printed board laminate materials and the BGA this could be the major starting point for investigating this condition.
It has been seen over the years that the BGA component does warp and many times creates shorts in the center of the device itself. Based upon the size of the component, when this happens the edges of the components are rising up off the surface of the PCB.
The same condition can occur if the PCB itself warps and twist during the soldering process.
So I would recommend reviewing the thermal profile, the warp and twist characteristics of both the board and the components. Secondly as a minor issue is the paste deposition process to make sure the paste is sufficient on the outer pads of the component.
Leo Lambert Vice President, Technical Director EPTAC Corporation At EPTAC Corporation, Mr. Lambert oversees content of course offerings, IPC Certification programs and provides customers with expert consultation in electronics manufacturing, including RoHS/WEEE and lead free issues. Leo is also the IPC General Chairman for the Assembly/Joining Process Committee.
Have you tried making the stencil aperture larger for the 4 corners?
Bill Coleman Vice President Technology Photo Stencil For over 18 years, Dr. Coleman has been the vice president of technology for Photo Stencil, working closely with customers to understand their printing requirements. His efforts have resulted in several new stencil products.
This is a classic example of Solderability-related Head-in-Pillow (SR-HiP). There is usually not a 100% fix, however, the best way to approach this problem is to locate the cause of warping and find the point in reflow where it occurs.
In this type of HiP, component warpage is a possibility. In component warpage, the BGA warps during thermal profile causing the lifting of the BGA from out of the solder paste. As the solder paste and the solder ball oxidize, the solder paste does not have enough flux activation to wet the ball.
Often, a longer cool down side of the profile will help. This gives a longer time above liquidus, allowing the solder paste and solder ball to form a proper solder joint. If the problem still exists, using a solder paste with uniform volume transfer during printing, as well as using an extended wetting time paste, could be a possible solution.
If the BGA is warped prior to reflow, the defect is CP-HiP (Co-planarity Head-in-Pillow). CP-HiP occurs when some of the BGA balls do not contact the paste. In this situation, some BGA will arrive with a co-planarity of .008 mils while the printer might only be depositing .005 mils of solder paste.
This combination leaves some of the BGA balls hanging in the air, therefore they will not wet. Printing a slightly thicker paste deposit may eliminate this defect. It is also a good idea to check the co-planarity tolerance on the parts you are receiving. If it is within the limit of your print height, this is probably not the issue. If all modifications fail, the problem ultimately lies within reflow. The best way to approach this is to minimize warping during reflow or change to a more robust paste formula.
Deck Street Consultants In his 32 years of industry experience, Mr. Seelig has authored over 30 published articles on topics including lead-free assembly, no-clean technology, and process optimization. Karl holds numerous patents, including four for lead-free solder alloys, and was a key developer of no-clean technology.
The most likely cause is probably warpage of the BGA Package, this seems to be the reason for the opens as it appears always at the corner balls. On inspection does it look as though the corner paste has melted and that the reflow is in that case OK?
If it all looks good, then the likely explanation is warpage where the device is curling up or smiling at the corners. You could look at reducing the ramp rate of the reflow profile to try to minimize the cte differences that might cause the warpage. As this effect happens only occasionally it could need just a minor change.
The other explanation could be Head in pillow, this is where the paste at the corner of the device is exposed to most heat and therefore uses up the most flux during preheat. This leaves insufficient flux left to achieve good reflow and the solder does not wet to the ball correctly.
On inspection you will see a line between the ball of the device and the solder or the solder will look dull and lumpy. If this is the case, then a more active or robust flux might be the answer. Also check after printing to make sure you are getting good consistent print deposits as a small reduction in volume can cause this effect.
If you cannot change the paste or improve the print, can you reflow in nitrogen?
Richard Boyle Global Product Champion Henkel Electronics Richard Boyle is a Global Product Champion at Henkel Electronics. He has over 25 years experience in the electronics assembly industry and is responsible for the global technical service of all of Henkel's solder materials.
There couple good feedback, apart I recommend
check is Board finish HASL, ENIG. HASL finish has more prone of such defect due
coplanirity issue, switch to ENIG finish PCB.second- PCB warpage: check PCB
comes form supplier preferable to be in sealed condition, if opened PCB keep in
20%RH till use to avoid board warpage.
Subrat Prajapati Supplier Quality Leader Ge Healthcare Subrat has 10 year of extensive experience in PCB assembly process optimizing for quality, process includes screen printing, wave, reflow. He has a copyright in stencil design published in Apex Expo2010 at Las Vegas US.
In order to mitigate the effects of warp and twist of the
board and BGA causing Opens or Head In Pillow defects ample past needs to be
For the more difficult air atmosphere reflow additional
paste volume is needed than in a nitrogen atmosphere, no less than 1:1 paste to
pad or over print may solve the problem. Nitrogen atmosphere reflow is more
forgiving and if employed it may also solve the problem providing ample paste
volume is present, generally 1:1 will provide the required volume in this atmosphere.
RL Brown, Consultant, USA
A head-in-pillow open solder joint is often caused by a "perfect storm" of three contributing factors. The primary cause is, of course, warping of either board or component near the high end of the profile. However in many cases, as the part cools, the tension relaxes, bringing the sphere and paste into contact while still molten. So long as there is sufficient flux activity remaining, and sufficient volume of solder to make contact, the solder will still wet and a good joint will form.
Solutions to head-in-pillow opens, then, usually take three approaches: 1) attempting to minimize warping of component (or board, in some cases); 2) increasing the volume of solder in the problem area(s); or 3) increasing flux activity at the end of the profile. Each of these process tweaks comes with some cost, risk, or trade-off.
Component warpage may be mitigated by slowing the ramp rate of the thermal profile, but keep in mind that adding extra ramp time or soak zones increases the overall profile time, which might reduce the activity of your flux when you need it most. Deflection (warping) of the assembly itself can sometimes be corrected by special supports or custom fixtures.
Solder volume variation is one possible issue, and indicates a review of the printing process is needed. It's also possible that your stencil is thin, such that the paste deposit is too shallow to overcome the coplanarity deltas of the warped part, resulting in no contact during reflow. Besides increasing the stencil thickness, you can try increasing the aperture size over the problem pads, but be careful not to exceed the pad diameter too much, or you risk creating "stray" solder that can result in shorts.
You might have success with a more aggressive flux. You can work with your paste vendor to see what they have available. You'll have to decide whether it's worth all the re-qualifying, profile development, validation, etc. that might accompany such a change. Another possibility is doing a flux dip of the component, prior to placement. A good, no-clean gel flux would provide plenty flux activity during a long profile. However, this is often a manual process, one which might not scale well enough to cost-effective at higher volumes.