BGA Shorting problem

We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. We are printing with a .020" diameter stencil aperature. Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components.

reply »

| We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane.

| One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | We are printing with a .020" diameter stencil aperature. | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components. Is your BGA moisture sensitive? Sometime popcorn issue would cause short at BGA. I would advise you to bake the BGA.

reply »

| We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. | One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | We are printing with a .020" diameter stencil aperature. | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components. Hi Kelly, A couple of questions first. what type of BGA are these components? CBGA, PBGA, TBGA, etc. What is your aperture diameter on your stencil and what stencil thickness are you using? I think I can be of more help once I know the answers to those questions. Regards, Justin

reply »

| | We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. | | One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | | We are printing with a .020" diameter stencil aperature. | | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components. | Hi Kelly, | A couple of questions first. what type of BGA are these components? CBGA, PBGA, TBGA, etc. What is your aperture diameter on your stencil and what stencil thickness are you using? I think I can be of more help once I know the answers to those questions. | Regards, | Justin Since the problem seems to come and go, Consider the operators as the source of the variability. (aside from baking to prevent moisture induced delimitation) A few things to check 1) Placement, Some operators will tweak a part if not perfectly placed on pad. This is a bad thing. If yes slap the operator, if no continue to step 2 2) Handling, paste and/or part becoming disturbed after the printing process. If yes slap the operator, if no continue to step 3 3) Printing, Repeatable? Are you measuring the Paste volume? Automatic screen printer? Are all shifts and/or operators setting up the screen printer the same? If yes train the operators to follow the same procedures, if no continue to step 4 4) Are the boards warping in the reflow oven? If yes, slap yourself and fix the oven, if no Continue to step 5 5) If you are still having problems you should do some imperical modeling to take advantage of the inherent continuity of quantitative variables. Good luck Mike

reply »

| | | We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | | | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. | | | One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | | | We are printing with a .020" diameter stencil aperature. | | | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components. | | Hi Kelly, | | A couple of questions first. what type of BGA are these components? CBGA, PBGA, TBGA, etc. What is your aperture diameter on your stencil and what stencil thickness are you using? I think I can be of more help once I know the answers to those questions. | | Regards, | | Justin | Since the problem seems to come and go, Consider the operators as the source of the variability. (aside from baking to prevent moisture induced delimitation) | A few things to check | 1) Placement, Some operators will tweak a part if not perfectly placed on pad. This is a bad thing. | If yes slap the operator, if no continue to step 2 | 2) Handling, paste and/or part becoming disturbed after the printing process. | If yes slap the operator, if no continue to step 3 | 3) Printing, Repeatable? Are you measuring the Paste volume? Automatic screen printer? Are all shifts and/or operators setting up the screen printer the same? | If yes train the operators to follow the same procedures, if no continue to step 4 | 4) Are the boards warping in the reflow oven? | If yes, slap yourself and fix the oven, if no Continue to step 5 | 5) If you are still having problems you should do some imperical modeling to take advantage of the inherent continuity of quantitative variables. | Good luck | Mike Post Script Oh yea one more thing to check 1) If that Imperical modeling BS does not work, check the placement machine to make sure the part is not being smashed into the paste, Bad placement of board supports could cause the board to rest at a higher level. I agree with Justin more info is needed. The frequency of the problem would be nice. Is it 10% or 1% of the parts are shorting? How often does it show up, what equipment are you using? Questions questions questions Mike

reply »

| We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. | One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | We are printing with a .020" diameter stencil aperature. | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components.

reply »

| We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. | One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | We are printing with a .020" diameter stencil aperature. | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components. Hi Kelly, I encountered a very similar kind of phantom shorts at ICT. Root cause was copper shorts in rawcards (PWB) between bga pads. Usually it is very hard to see as it is usually under the solder mask. It is also very hard to detect without removing the comp.(BGA) itself. X-ray (high- focus), thermal / IR imaging would not detect this kind either. It happens during external layer exposing (Rawcard)- most probably lint or piece of hair on phototool (diazo) would usually cause that. This might affect the whole batch or some of them.... Date code isolation method would work. Hope it helps, good luck. Regards Kc.

reply »

| We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. | One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | We are printing with a .020" diameter stencil aperature. | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components. Is it an internal short or the short that related to the process out of control? If the short occur during the ICT testing and X-Ray it,there is no problem then it is the PCB internal short.But when you X-RAY and found short then this is you processes related problem. Processes related could link to Component sensitivity,Aperture opening on the stencil(make sure some apertures are not over etch),screen printing process...support pin, dented stencil. Hope this help!!!! Rin

reply »

| Hi Kelly, | A couple of questions first. what type of BGA are these components? CBGA, PBGA, TBGA, etc. What is your aperture diameter on your stencil and what stencil thickness are you using? I think I can be of more help once I know the answers to those questions. Justin: It is a 6mil stencil with 20mil aperatures. The part is a Globtop BGA. | Regards, | Justin

reply »

| | We are experiencing shorting on a BGA. The part has a 1mm pitch with a 196 I/O count. The problem seems to pop up without warning & disappear in the same manner. | | Our pad design is .020" diameter pads with .025" solder resist diameter. The center pads on the PCB are all tied together as a ground plane, therefore, the pads are solder mask defined and are .025" diameter. The shorts mostly occur on the center pads on this ground plane. | | One thing that is strange, is that there is another 144 I/O BGA very close to this one on the PCB, with the same pitch and pad with center ground plane designs, but we do not experience any shorting problem. | | We are printing with a .020" diameter stencil aperature. | | Any advice or input would be appreciated. I would be interested to hear what other people are doing for pad and aperature designs for similar components. | Is it an internal short or the short that related to the process out of control? If the short occur during the ICT testing and X-Ray it,there is no problem then it is the PCB internal short.But when you X-RAY and found short then this is you processes related problem. Processes related could link to Component sensitivity,Aperture opening on the stencil(make sure some apertures are not over etch),screen printing process...support pin, dented stencil. | Hope this help!!!! | Rin Kelly, The guys pretty much covered all of the possibilities. I don't see anything out of control on what you've told me so far. Do you do paste height measurements? Definitely check the board supports on your placement equipment. Are the components dimensionally stable. Glob top BGAs have a tendancy to curl up like potato chips. Those components have an FR5 substrate. It's either a 1, 2, 3, or 4 layer construction depending on the IO count. The more layers the more dimensional stability. Likewise, the larger the die, the more stability. FR5 has a Tg of about 165 degrees C. If you're hitting 220 to 240 for reflow temps, you're way over the Tg and that can't be helped. The fiber bundles in the substrate can migrate all over the place and turn your once flat package into something remeniscent of a topographical map. You are at the mercy of your package. Check your profile and get it as cool as you can. Check with your paste vendor first though and see when the volatiles burn off. That is going to determine your peak temperature. Hopefully you can get away with 205 to 210. Other than that, you might want to try baking a batch of the components. 125C for 24 hours and see where the yields go then. Hope some of that long winded crap was helpful. Regards, Justin Medernach

reply »