Electronics Assembly Knowledge, Vision & Wisdom
Round or Square Stencil Apertures?
Round or Square Stencil Apertures?
What is your opinion about using round versus square apertures for BGA patterns in solder paste stencils? Which wins, square or round? The Assembly Brothers, Jim Hall and Phil Zarrow, answer these questions.
Board Talk

,{url:'http://www.circuitinsight.com/videos/board_talk_round_square_stencil_apertures.mp4'}], clip:{autoBuffering:true, autoPlay:true, scaling:'scale' } }).ipad();
Board Talk is presented by ITM Consulting
Phil Zarrow
Phil Zarrow, ITM Consulting
With over 35 years experience in PCB assembly, Phil is one of the leading experts in SMT process failure analysis. He has vast experience in SMT equipment, materials and processes.


Jim Hall
Jim Hall, ITM Consulting
A Lean Six-Sigma Master Blackbelt, Jim has a wealth of knowledge in soldering, thermal technology, equipment and process basics. He is a pioneer in the science of reflow.
Transcript

Phil
Welcome to Board Talk. This is Jim Hall and Phil Zarrow, with ITM Consulting - Pick and Place - the Assembly Brothers. We're here to attempt to solve your problems in assembly process. And what question do we have today, Jim?

Jim
We have a question about stencil printing. "What is your opinion using round versus square apertures for BGA patterns in our solder paste stencils?"

Of course in printing the big trick is to get consistent fill and, more importantly, consistent release of solder paste from every aperture in a stencil, all those thousands of little holes in there. If any one of them doesn't release completely every time, and you have the potential for inadequate soldering defects.

So as apertures get smaller, the question has been around for a long time. Is it more efficient to use a round aperture or a square aperture, particularly when you're dealing with BGAs where the pads tend to be round anyway?

Many years ago, I think more people favored round apertures. Based upon the published studies and hands-on experience that I've seen with our customers, I feel that for smaller apertures, .5 millimeter, 20 mil pitch and smaller chip components, 0402s, 0201s, and so forth. I think that Square wins!

Although the area ratio would tend to favor a round aperture having less perimeter per unit area, more of the data tend to support square or rectangular apertures for more consistent transfer efficiency, which means more consistent release of paste from the small apertures.

Phil
Let's face it in printing, that's what it's all about. This is a kind of a related subject, but with regard to non-BGA pads, some of the other pads people work with what - so called Roundy pads. It's a different matter, a different situation, but on passives and things along those lines...

Jim
Most of them are based upon square or rectangular pads. Then modified with a cut-out substitute radius corner to get the home plate or reverse home plate pattern. We're talking about changing the shape of the aperture other than a pure circle or a pure square or rectangle. That's a more complex subject than we can get into at this point now.

Phil
Remember, in printing, the bottom line is fill and release and the idea of getting the effective, proper solder volume. So thank you for the question. This is Phil Zarrow and Jim Hall.

Jim
And no matter what process you're using, reflow, wave, hand or selective, don't solder like my brother!

Phil
And please don't solder like my brother!

Reader Comment

It's nice to se so many agree to disagree but the bottom line is none of this will give consistent results unless everything from the program, blade selection, paste, fab, stencil materials and thickness, etc. are taken into account. I prefer square with radius corners but again there are no instant fixes. As devices get smaller, processes have to get tighter. We will reach a point where stencils just won't do anymore.

Mark Maheux Sr., Honeywell
Reader Comment

It is preferred to have a square with rounded corner which will help the solder paste to transfer and maintain its brick shape. This also indirectly increase the volume with out increasing the size.

Sakthivel Padmanapan, ASAHITEC STENCILS PVT LTD, India
Reader Comment

Just to support the subject. In my little time almost, 10 years with Panasonic, since we switched our BGA pads from round to square, we right away noticed that our insufficient solder defects on these pads, had a positive impact due to a paste deposition improvement, so, definitely, squared is better.

Abraham Noriega Lizarraga, Custom Sensors and Technologies (CST), Mexico
Reader Comment

In my experience we've tried both round and square apertures on our BGA's with mixed results. We switched to a 1 to 1 aperture using a rounded rectangle with perfect results (corners rounded to .003")

Doug Christen, Ascentron Inc., USA
Reader Comment

I have my stencil house convert the round BGA to square by keeping the same area, this results in a smaller square but the same volume of solder paste is deposited and that also reduces the amount of overprint. I had to do this to get better release of the paste from the stencil when there is not very much solder paste to work with on the micro BGAs.

Scott McKee, Goldenwest Technologies
Reader Comment

I have my stencil house convert the round BGA to square by keeping the same area, this results in a smaller square but the same volume of solder paste is deposited and that also reduces the amount of overprint. I had to do this to get better release of the paste from the stencil when there is not very much solder paste to work with on the micro BGAs.

Scott McKee, Goldenwest Technologies
Reader Comment

Using square stencil apertures and printing square paste bricks for round BGA or CSP (including flip-chip) pads may lead to disaster. Here's why:

1. Whenever printing paste, it is seldom a good idea to print paste off the pad. Solder fines have a tougher time agglomerating into a single solder joint (during reflow) when overprinted off of the pad and onto the solder mask, even more so if the pad is not solder-mask-defined because then they need to overcome the gap between the solder mask and the edge of the pads.

As a result, you can end up with many renegade fines trapped under the BGA that are not part of any solder joint. These may well go undetected, even when using an automatic X-ray scan because quite often they are too small or are outside of the X-ray's "target zone."

Typically, for automatic X-ray equipment, the software is quite similar to an AOI, and of course some X-ray inspection machines are actually a combination of both, called an AOX, AXI, or OAX, depending on which configuration and inspection methodology is used. The machine first finds a localized fiducial, and from there it goes into the programmed inspection routine. This routine is typically telling the machine to perform a series of inspections based (first) on the offset distance from the fiducial, then into the step-and-repeat sub-routine where it measures every ball based on the pre-programmed pitch, target ball size, roundness algorithm, and voiding gray-scale measurements (remember I said typically, not all operate the same way).

What this means is that NO inspection is performed in the areas between each target location. If there are individual fines or, worse yet, semi-agglomerated renegade solder balls in these "nether regions", they will go undetected by the automatic X-ray. Yet they can be large enough to violate the minimum electrical clearance required between the vias and the BGA solder joints, or between the vias and adjacent conductors or the finished solder joints and adjacent conductors.

The result is can be poor high-RF performance (crosstalk), and either permanent or intermittent solder shorts caused by all of these renegade conductors floating around the finished CCA. If water soluble chemistry is used, the wash process usually will get rid of some of these renegade solder balls, but never all of them.

If no-clean or RMA flux chemistry is used, there will be hundreds of these entrapped renegades, and while they may be held in the no-clean or RMA flux residues they will eventually work themselves loose and cause all kinds of havoc.

2. Some of those nice folks reading your column may interpret what you are saying as "square pads and square stencil apertures may provide better paste printing results for BGAs." Again, this can also lead to disaster. In this scenario, the round BGA, CSP, or flip-chip solder balls are placed onto square pads with square paste deposits. Instead of a finished solder joint similar to a round vase on a round table, you now have a solder joint that is round on the upper half connected to the round pad on the bottom of the BGA, but where it connects to the PWB the solder joint has four corners.

During the cool-down stage of the reflow profile, and during its service life where the temperatures are cycling, differing coefficients of thermal expansion between the BGA and the circuit board causes severe stresses on the solder joints between them.

When four stress risers are present on every square pad under the BGA, and if the stresses are sudden and severe enough, the corners of each pad will be pulled up. This upward leverage causes the bond between the PWB pad and the PWB itself to fracture, often severing the traces connected to these pads.

This is called "pad cratering", and with the higher temperatures required to solder lead-free solder plus the more brittle nature of lead-free solder joints this phenomena is being seen more and more every day. It is a bad enough issue with standard round pads, adding the four stress risers has been shown to make it orders of magnitude worse.

In both of these examples, the issue can go undetected until after all CCAs have completed assembly, inspection, and test and are deployed into the field. More than one company has found out the hard way they have just generated a whole bunch of scrap, with the full cost of materials and manufacturing lost with no possibility of recovery.

As always, you must warn your audience not to make these changes without going through some type of qualification steps to ensure these issues are not present before jumping in and implementing them on production CCAs. With all due respect to your audience, some may not understand the ramifications of not performing the due diligence aforehand. I do not want that to happen to anyone, hence this note.


Richard Stadem, Advanced Scientist/Engineer, General Dynamics Nuclear and Space Products Division
Reader Comment

If converting to square pads from round on the stencil, would the correct conversion be to make the square sides the size the diameter of old round pad?

James Puritch, VP of Research and Development, Reliable Controls Corporation
Reader Comment

For a moment I thought that you were going to say round ... but as you said, in my experience we found that squares for BGA worked better than circles.

It is good to be in-sync. Keep it up! Regards from Mexico amigos.


Guillermo Velazquez, Manufacturing Engineering Manager, Rainbird
Submit A Comment

Comments are reviewed prior to posting. Please avoid discussion of pricing or recommendations for specific products. You must include your full name to have your comments posted. We will not post your email address.

Your Name


Company


E-mail


Country


Comments





Related Programs
bullet Most Important Factors for Screen Printing
bullet Reliability of Stacked Microvia
bullet Stencil Manufacturing and Impact on Precision
bullet Digitally Printed Battery Design
bullet Printing of Solder Paste - A Quality Assurance Methodology
bullet Are There Standards Governing Polarity Marks?
bullet Round or Square Stencil Apertures?
bullet Stencil Design Improves Drop Test Performance
bullet Mixed Technology - Which First
bullet Sources of ESD in a Production Line
More Related Programs
About | Advertising | Contact | Directory | Directory Search | Directory Submit | Privacy | Programs | Program Search | Sponsorship | Subscribe | Terms

Circuit Insight
6 Liberty Square #2040, Boston MA 02109 USA

Jeff Ferry, Publisher | Ken Cavallaro, Editor/Business Manager

Copyright © Circuitnet LLC. All rights reserved.
A Circuitnet Media Publication