| | | Yesterday as I was creating a profile on a plastic BGA using Conceptronics freedom 2000 rework station, I happend to discover that I was able to remove solder bridging from a plastic BGA. | | | | | | This is what I did. | | | | | | Created a removal profile, I had a board that had 3 bridges on the BGA. Run the profile and as soon as saw the BGA that was ready for removal, I brought down the nossle Picked up the BGA and came up almost 20mils than dropped back the BGA. I X-Ray the BGA, no bridging. The solder balls all maintain it's shape and all looked to be consistant. | | | | | | Has Anyone done something similar.. | | | | | | | | | Thanks in advance. | | | | | Very cool, Tony | | | I agree, very cool. Next time I run into this well-known bridging trouble with PBGAs, I will try it out. | However, I see a potential problem with this practice. | When reworking PBGAs, bridging usually happens because of non-uniform heating of the package. During heat-up the package corners, who get the heat first, are down while the center of package is up. So the solder of the corner balls is being spread out, and two or more balls can touch. That's the most common reason for shorts. The best benchtop rework station will never give you the uniformity of a reflow oven. You see shorts mostly in the corners, don't you? When the package cools down, it's the opposite: if you cross-section a reworked package, you see something like the roof of a chinese restaurant. | Now, when your bridging is caused by this normal warpage problem, your solution is great. | However, there can be other reasons for shorts. For example, after removal of the defective part you may damage the solder mask when cleaning the site, so you may have small areas of exposed copper between BGA pads, very close to the pads themselves. These can cause shorts too. You may be able to remove such a short with your cool idea, but the part may then prematurely fail in the field because of reduced space between a trace and a joint. | In such cases, it is better to remove the part again and find out what's wrong. | There is something I did left out on the explained process.
I soaked the site with liquid flux..I guess this helped the solder to string durring the 20mil separation.
I am going to continue doing more experiments and will keep U posted..
Anyone doing something similar please share with us your your techniques...
Thanks
reply »