I totally agree with the idea that "avoiding" voids can result in stronger joints. I do not see any advantage in having voids at all. I understand that in reality it is something we have to live with due to the constraints that we are faced with. Nevertheless, each void is a weak spot. P.S. Please do not shoot the guy, he might have some children. Ha, ha, ha. Ramon -----Original Message----- From: TechNet [mailto:[log in to unmask]] On Behalf Of Stadem, Richard Sent: Tuesday, July 19, 2005 8:55 AM To: [log in to unmask] Subject: Re: [TN] BGA voids Voids in solder paste are normal, I do not disagree with Werner on that point. However, what is "normal"? I see literally hundreds of X-rays of solder joints per week, and based on experience I know that producing void-free solder joints is entirely possible, but let me qualify that with that now-famous line: "It depends.....". There are a number of ways voids can form. You can have them in your BGA solder spheres when the components come in the door, they can be created from outgassing around the solder joint during reflow (especially in soldermask-defined boards), they can come from chemical contaminants on the board prior to print and reflow, they can be induced by the solder paste, they can be created by contaminants caused by improper handling of the board and/or components, they can be caused by improper reflow profiles, and many other ways. But most of them are typically caused by too much moisture in the solder paste. If you get rid of that problem, your chances of producing solder joints void-free, or nearly void-free, improve dramatically. Are voids bad? IMHO they do not cause a problem if they are smaller than the 25% limit, but again, it depends. I have worked at companies where the voiding caused a component to break off during a gunshot. The fracture went through the void and caused the component to lift off of the board on one end. Now get this, every capacitor on the assembly fractured on one end and lifted up during the gunshot, but only on one end. This was the end where the split terminal induced a void, but the solder wetting and IMF were otherwise good. There have been several published papers that state the void in a BGA solder joint can prevent total failure if a crack is present, as the void alleviates the pressure once the crack propogates into it. But two of the papers ended up stating that it is less likely for a solder joint to develop a crack if the grain structure is tightly defined (not coarse), and no voiding is present in the first place, and the pads are not soldermask-defined. If voids can be avoided by the use of a solder paste that was formulated to prevent it, why not take advantage of that? If you don't have any voids in your solder joints, then you do not have to depend on an operator to make judgement calls to determine if it is over 25%. You do not need to worry about whether your X-ray system is good enough to detect them or not. My feeling is that if all of the solder joints on a given assembly are extensively riddled with voids, there is something wrong somewhere in the process. Regarding which paste to use, I switched to a particular paste for a previous company. It almost totally eliminated all voiding seen in both SMT and BGA solder joints. Furthermore, it has been an evaluation/qualification champion in three different companies that I have worked with. There are other solder paste formulations out there that were created to prevent voiding as well, but this one seems to be the best overall performer in several different categories. There are new formulations coming out from different vendors quite often. If you are seeing a lot of voiding (or spattering, or wetting issues, or solder strength issues, etc.) then I strongly urge you to undergo a paste evaluation so you can go to sleep at night knowing that you have the best solder paste on the market being used on your assemblies. A new evaluation should be performed every so often to take advantage of the improvements being made. I do not want to recommend any paste over the other, as I think all solder process engineers should perform a paste evaluation themselves to see what works best for their particular process and product. The things you learn when you perform an extensive and detailed paste evaluation are invaluable. It is well worth the time, effort, and cost. -----Original Message----- From: TechNet [mailto:[log in to unmask]] On Behalf Of Steve Gregory Sent: Monday, July 18, 2005 7:12 PM To: [log in to unmask] Subject: Re: [TN] BGA voids Hi Richard! I know one goal is to have zero voids. But is seeing them really a problem? This all came about when I've been looking at upgrading our current x-ray system. To be honest, I haven't seen voids here using our current equipment. I've seen them now with some images from a sample that I've gotten with a system that has more capability than our current system has. I asked about the voids to the applications engineer, and was told that the voids that he observed were normal, and not to worry about them, with the caveat that if they should become larger, or more concentrated in any one area. He told me that he has about 25-years experience working with x-ray inspection...so I'm not that quick to go out and shoot him. He didn't say not to worry about voids, but the voids that he observed from the sample I sent him were not a concern...he said they were normal. It seems that Werner agrees... From the differing opinions, it seems the jury is still out about what is still acceptable, and what really is rejectable when it comes to voids. Right now, the J-STD-001 and -610 call out that more than 25% of the area of the solder sphere is rejectable...it does not specify where they are located. Anymore thoughts? Anyone? -Steve Gregory- *************************************************************** >Hi Steve, >Seeing voids is indeed normal. >What you are referring to is in IPC-7095, Section 7.5.1.6 Process >Control Criteria for Voids in Solder Balls. >This was a suggested process control example, and is NOT a requirement, >nor was it meant to be one--otherwise, I would have voted negative on >the document. >Putting it like your customer did is absurd. >Werner Voids can be caused by a number of things, but the most common cause is the moisture content of the solder paste, and its condition. That being said, a certain paste that has been known to be the mainstay for many years currently is the worst one today as far as voiding. A different paste manufacturer made an excellent paste that was formulated to prevent voiding. I recently performed a paste evaluation of standard 63/37, a leadfree water soluble, and a leadfree no clean paste. There were 27 different tests, with 5 vendors. On the standard 63/37, the paste formulated to prevent voiding had ZERO voids in 1330 BGA joints, multiplied by 15 samples. The previous paste vendor had voids in every single BGA joint. All variables were strictly controlled, and I did not know which sample was being tested (they were numbered by someone else to prevent any bias). While I do not want to give out vendor names on the forum, I would be glad to share this info with anyone offline. The application engineer who told you not to worry about voids should be taken outside and shot. While some data exists that says voids are OK if they are small enough, you should certainly strive for an alloy that provides you with an excellent grain structure free of voiding, with evidence of good IMF at the pad/ball interface. Sorry, pal, but you cannot determine this with an X-ray machine. It can only be done with a good microsectional analysis. -----Original Message----- From: TechNet [mailto:[log in to unmask]] On Behalf Of Stephen Gregory Sent: Monday, July 18, 2005 1:07 PM To: [log in to unmask] Subject: [TN] BGA voids Good Day Technetters! I have a question about BGA voids. We are bidding on some work and it looks like they have taken statements that are in the IPC-7095, and placed them as notes on their drawing. One of the notes has to do with a limit on the percentage of balls on a BGA device that are allowed to have voids. For example; "Less than 5% of the balls can have voids, and the percentage of the void area must meet class 3 requirements." I can understand the void area being critical, but the percentage of the number of balls that are allowed to exhibit them? The reason that I'm asking, is that I was told last week by a application engineer from a x-ray machine company (who shall remain nameless), that seeing voids is normal. Just as long as they aren't too big, or being concentrated at either the pad/ball interface, or the ball/device interface. He said you should worry if you don't see any voids on BGA's with eutectic balls. Just curious as to what you all think about this? Kind regards, -Steve Gregory- Senior Process Engineer LaBarge Incorporated Tulsa, Oklahoma (918) 459-2285 (918) 459-2350 FAX --------------------------------------------------- Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815 ----------------------------------------------------- --------------------------------------------------- Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815 ----------------------------------------------------- --------------------------------------------------- Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815 -----------------------------------------------------