Reliability of Reworked BGAs/CSPs

I'd like to find some technical papers on the reliability of reworked BGAs. If you know of any, please let me know.

I have a concern about this because our current rework process does not apply solder paste to the pads prior to placing the part; this results in a lower standoff and reduced fatigue life...but reduced how much?.

reply »

| I'd like to find some technical papers on the reliability of reworked BGAs. If you know of any, please let me know. | | I have a concern about this because our current rework process does not apply solder paste to the pads prior to placing the part; this results in a lower standoff and reduced fatigue life...but reduced how much?.

Hi Mike,

Well, I've not had to rework any (yet, anyway...) but if you point yer' browser to: www.Tessera.com (or click on the link below)

and when you get there, click on; Technical Info, then click on; Assembly Process Development for Chip Scale BGA Devices. It's an article that was written by Vern Solberg whose been working for Tessera for quite a number of years as the Director of Advanced Manufacturing ( or something along those lines) There's a bunch of other articles there as well.

Vern says just to use no-clean flux...but I imagine that'll depend what kind of alloy the balls are made of as you can get them in a variety of alloys. He says that you need to use a new device as well, to not try to remove and replace the same chip...(but they WOULD say that, coming from a business point of view, huh?)

Hope this helps ya'!

-Steve Gregory-

reply »

| I'd like to find some technical papers on the reliability of reworked BGAs. If you know of any, please let me know. | | I have a concern about this because our current rework process does not apply solder paste to the pads prior to placing the part; this results in a lower standoff and reduced fatigue life...but reduced how much?. | The humungous company for whom I am now working has had the wisdom and foresite to allow me to support manufacturing efforts in various areas. Wonderful as this sounds to me, the one area I am working, and am most interested, is repair/rework - especially BGA and beyond. We employ the SRT 1000 to do this.

As Steve said, we and many like us, do not apply solder paste as this adds about less than 7% to the solder joint volume and provides a mechanism for other problems not associated with normal reflow operations. Eutectic balls are used. You know, the standard stuff my balls are made of (63/37). Sometimes during reflow, high temp balls are positioned at the corners to prevent collapse and on it goes.

Oh yes, about reliability, all I can say is initial quality is very well established during X-Ray, but these guys won't let me near their long term reliability stuff. I can only guess, but I guess pretty well after running so many rework cycles.

Earl Moon

reply »

| I'd like to find some technical papers on the reliability of reworked BGAs. If you know of any, please let me know. | | I have a concern about this because our current rework process does not apply solder paste to the pads prior to placing the part; this results in a lower standoff and reduced fatigue life...but reduced how much?. | Michael: A quote ...

"Motorola does, in fact, rebump packages which are returned from the field to allow them to be tested. Motorola does not, however, recommend this process be used to allow re-assembly of the BGA into actual product. BGAs that are shipped by Motorola are typically qualified to withstand from two to four reflows, depending on the intended application and the Motorola product group involved. That is, the packaged devices are subjected to from two to four reflows of preconditioning prior to any reliability stressing, as part of qualification testing. When a BGA is removed from a board, rebumped, and reflowed back onto a board, it has potentially seen from four to six reflows, and may have exceeded the number for which it was qualified. Additionally, there is no traceability to distinguish a rebumped BGA. Any questions in this area should be directed to the Reliability and Quality Assurance group dealing with the particular BGA."

From this link.

Have a nice day ;-)

Dave F

reply »

| I'd like to find some technical papers on the reliability of reworked BGAs. If you know of any, please let me know. | | I have a concern about this because our current rework process does not apply solder paste to the pads prior to placing the part; this results in a lower standoff and reduced fatigue life...but reduced how much?. | Assuming you are putting a new PBGA on your board, when you rework, and not a PBGA which has been reballed, the difference in solder joint reliability would be about a 16% shorter life with the flux only process as compared to using solder paste, in 0�C to 100�C temperature cycling. There are two reasons for the shorter life, one is the decreased standoff height of the BGA, to the PCB, and the other reason is the BGA you solder with just flux, will not have solder joint voids, as compared to BGA's which have been soldered with paste. The voids create a time delay in the total time it takes a crack to propagate through the bulk solder. I'm sending you a paper we've published, that will give you the number of cycles to failure, for flux vs. solder paste attached BGA's, as well as voids, vs. no voids.

reply »

Actually, solder paste adds about 12% to the joint volume for PBGAs, or about 31% for MicroBGAs -- assuming standard (recommended) pad sizes and stencil thicknesses, and paste with 90% metal content (50% metal by volume).

| | I'd like to find some technical papers on the reliability of reworked BGAs. If you know of any, please let me know. | | | | I have a concern about this because our current rework process does not apply solder paste to the pads prior to placing the part; this results in a lower standoff and reduced fatigue life...but reduced how much?. | | | The humungous company for whom I am now working has had the wisdom and foresite to allow me to support manufacturing efforts in various areas. Wonderful as this sounds to me, the one area I am working, and am most interested, is repair/rework - especially BGA and beyond. We employ the SRT 1000 to do this. | | As Steve said, we and many like us, do not apply solder paste as this adds about less than 7% to the solder joint volume and provides a mechanism for other problems not associated with normal reflow operations. Eutectic balls are used. You know, the standard stuff my balls are made of (63/37). Sometimes during reflow, high temp balls are positioned at the corners to prevent collapse and on it goes. | | Oh yes, about reliability, all I can say is initial quality is very well established during X-Ray, but these guys won't let me near their long term reliability stuff. I can only guess, but I guess pretty well after running so many rework cycles. | | Earl Moon | |

reply »

| Actually, solder paste adds about 12% to the joint volume for PBGAs, or about 31% for MicroBGAs -- assuming standard (recommended) pad sizes and stencil thicknesses, and paste with 90% metal content (50% metal by volume).

Michael,

In the "real world" I agree. In my current world it is as I said and the solder joints are acceptable with less than "outside world" solder volume.

I still don't know reliability here. I do know some of it outside and it's getting better.

Thanks,

Earl Moon

|

| | | I'd like to find some technical papers on the reliability of reworked BGAs. If you know of any, please let me know. | | | | | | I have a concern about this because our current rework process does not apply solder paste to the pads prior to placing the part; this results in a lower standoff and reduced fatigue life...but reduced how much?. | | | | | The humungous company for whom I am now working has had the wisdom and foresite to allow me to support manufacturing efforts in various areas. Wonderful as this sounds to me, the one area I am working, and am most interested, is repair/rework - especially BGA and beyond. We employ the SRT 1000 to do this. | | | | As Steve said, we and many like us, do not apply solder paste as this adds about less than 7% to the solder joint volume and provides a mechanism for other problems not associated with normal reflow operations. Eutectic balls are used. You know, the standard stuff my balls are made of (63/37). Sometimes during reflow, high temp balls are positioned at the corners to prevent collapse and on it goes. | | | | Oh yes, about reliability, all I can say is initial quality is very well established during X-Ray, but these guys won't let me near their long term reliability stuff. I can only guess, but I guess pretty well after running so many rework cycles. | | | | Earl Moon | | | | | |

reply »