Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

BGA failure at Functional Test

Views: 4301


BGA failure at Functional Test | 7 January, 2009

Recently on one of my assemblies(sn/pb) I started seeing leaded BGA failure at functinal test for opens, but when I inspected them through 5DX and ERSA Scope it doesnt seem like there is any opens or insufficient and half no. of boards passed functional test and rest of the boards failed. half no. of failure boards passed after re-reflow and rest of them had to change or replace BGA in order to pass at FCT. Same assembly we ran couple of times and never had this issue only last time this happend. we havent change the profile, paste(Kester R562), stencil, SMT line. profile for this board is very good and I created profile on fully populated PCB and ramp rate is 1to 1.5 dec C/sec, 65 to 80 sec soak time (time bet 150 to 170),TAL is 70 to 80 sec with peak temp 210 to 225. when we did non-destructive analyis by removing BGA substrate or body without removing BGA balls from the pad we observed very good solder joint formation, and doesnt look like as if its contaminated and I dont even think that this is bad part since half of the failure board passed the test after re-reflow, if it was not known good die it would have failed even after re-reflow. could you guys please help me finding the root cause for this issue? I really appreciated and thanks a lot in advance for your input.

Milan Shah

reply »



BGA failure at Functional Test | 8 January, 2009


Is there any PTH Assembly after reflow, we found cracked joints on some BGAs which was caused by operators ing Oupin connectors which were a tight fit. This caused the PCB to flex hence cracking joints.

One way of checking for this during test is to apply pressure on the BGA and see if it passes.Try each corner individually first.


reply »


BGA failure at Functional Test | 12 January, 2009

Hi Milan, One cause for such BGA failures could either be a laminate warp (bow or twist) formed during your Reflow or Wave soldering process, or perhaps a warp seen on the BGA component package itself prior to processing on SMT line. IPC guidelines for warp acceptance may not suit your application, and hence needs to be separately evaluated. It is essential to consider BGA storage conditions at your end and that MSL labels are correctly referred to for any Pre-baking (if required) prior to processing on SMT line. You could also derive and analyse any common trend for the faliures such as Component Batch code, Supplier Lot, Lot date, testing conditions, etc., which will help you filter out the problem. Just in case you are using this BGA for the first time in your assembly, you should verify its datasheet to confirm its process temperature handling capabilities.

reply »


BGA failure at Functional Test | 16 January, 2009

Kester R562 is a 63/37 paste. Not ideal for no-lead parts. I bet you are seeing no-lead BGAs and they will become very weak using your old proile and paste (been there - done that). You can try a hotter profile but will probably just burn off your flux way too soon. Try an paste designed for no-lead parts that has lead. Re tweak your profile for the no lead parts and you'll get a strong intermatallic. This will eliminate your problem.

reply »

Nitrogen Capable SMD & BGA Rework Station

Fully automatic selective soldering stations