problem in micro-BGA assembly

Dear guys,

I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!!

I increase the printing pressure and printing speed, but no help!

Some people suggest that I should (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or (ii) use size 4 powder.

Is it really help? DO you have any experience on assembling micro-BGA with solder paste?

Thanks in advance.

reply »

| Dear guys, | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | I increase the printing pressure and printing speed, but no help! | | Some people suggest that I should | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | (ii) use size 4 powder. | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | Thanks in advance. | Vic: Think volume, you are height as a proxy for volume. Good luck. Dave F

reply »

| Dear guys, | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | I increase the printing pressure and printing speed, but no help! | | Some people suggest that I should | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | (ii) use size 4 powder. | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | Thanks in advance. | Vic Take a hard look at the bare FAB. It sounds like the Mask around the pad is keeping the stencil from contacting the pad (thick silk screen will do this also). I have seen where the board house will plug the VIA�s and in doing so they leave a mound of mask on the via. This will cause the stencil to be raised higher than the pad causing the extra 2-3 mills of paste.

MD Cox

reply »

| | Dear guys, | | | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | | | I increase the printing pressure and printing speed, but no help! | | | | Some people suggest that I should | | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | | (ii) use size 4 powder. | | | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | | | Thanks in advance. | | | Vic | Take a hard look at the bare FAB. It sounds like the Mask around the pad is keeping the stencil from contacting the pad (thick silk screen will do this also). I have seen where the board house will plug the VIA�s and in doing so they leave a mound of mask on the via. This will cause the stencil to be raised higher than the pad causing the extra 2-3 mills of paste. | | MD Cox | | | As MD says, many causes effect excessive paste print height/volume. We've had people rely on "clean" stencils and come to find "crud" as dried on solder paste contacting board surfaces making off contact a reality. Check it all as "clean" stencil and board surfaces as well as anything else causing the off contact condition.

Earl Moon

reply »

| | Dear guys, | | | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | | | I increase the printing pressure and printing speed, but no help! | | | | Some people suggest that I should | | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | | (ii) use size 4 powder. | | | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | | | Thanks in advance. | | | Vic: Think volume, you are height as a proxy for volume. Good luck. Dave F | Vic, That is a pretty significant increase. Check the actual stencil thickness. No doubt it is actually 8 mil. If you are using rubber squeegees don't even bother checking paste height. You will never get consistent results anyway. If your stencil thickness meaures true I would question the method or paste measuring tool accuracy.

reply »

| | Dear guys, | | | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | | | I increase the printing pressure and printing speed, but no help! | | | | Some people suggest that I should | | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | | (ii) use size 4 powder. | | | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | | | Thanks in advance. | | | Vic: Think volume, you are height as a proxy for volume. Good luck. Dave F |

reply »

| Dear guys, | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | Vic, Sometimes its the obvious that we forget to check. Since you increased your squeegee pressure; are you getting good board support. Without the support it will also cause an over-deposit of paste. | I increase the printing pressure and printing speed, but no help! | | Some people suggest that I should | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | (ii) use size 4 powder. | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | Thanks in advance. |

reply »

| Dear guys, | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | I increase the printing pressure and printing speed, but no help! | | Some people suggest that I should | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | (ii) use size 4 powder. | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | Thanks in advance. |

The 400 mesh paste should solve your problem. We use Alpha Metals NC, the tact time is much longer than other pastes that we have experimented with. For some real good advice on stencil and paste, contact Tom Luckenbach at 919 460 0111 *120. They have been placing the TV46 uBGA from tessera for a few years now.

reply »

| | | Dear guys, | | | | | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | | | | | I increase the printing pressure and printing speed, but no help! | | | | | | Some people suggest that I should | | | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | | | (ii) use size 4 powder. | | | | | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | | | | | Thanks in advance. | | | | | Vic: Think volume, you are height as a proxy for volume. Good luck. Dave F | | | Vic, | That is a pretty significant increase. Check the actual stencil thickness. No doubt it is actually 8 mil. If you are using rubber squeegees don't even bother checking paste height. You will never get consistent results anyway. If your stencil thickness meaures true I would question the method or paste measuring tool accuracy. | | I agree. Don't forget, your foil thickness is NOMINAL. Also, don't forget the solder bricks can be deformed and elongated as the stencil breads-away from the PCB This will cause a shift in the AVERAGE solder brick height. I have witnessed this again and again using laser mesurement techniques. Essentially you get a "scoop" shape rather than a square solder brick. This occures because the stencil to paste contact has drag as they seperate. Also, verify your baseline measurement error. Especially when using lasers. Example: My (brand name witheld) laser height machine has a baseline error of 0.0004 inch error. This can easilly be determined by statistical techniques and a simple experiment. Bottom line all these nominal and average errors accumulate. So, think volume. PS using statistical sampling techniques would account for this phenomenon and bring you back into the green! Think volume dude! | | Dean.

reply »

| | | | Dear guys, | | | | | | | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | | | | | | | I increase the printing pressure and printing speed, but no help! | | | | | | | | Some people suggest that I should | | | | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | | | | (ii) use size 4 powder. | | | | | | | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | | | | | | | Thanks in advance. | | | | | | | Vic: Think volume, you are height as a proxy for volume. Good luck. Dave F | | | | | Vic, | | That is a pretty significant increase. Check the actual stencil thickness. No doubt it is actually 8 mil. If you are using rubber squeegees don't even bother checking paste height. You will never get consistent results anyway. If your stencil thickness meaures true I would question the method or paste measuring tool accuracy. | | | | I agree. Don't forget, your foil thickness is NOMINAL. | Also, don't forget the solder bricks can be deformed and elongated as the stencil breads-away from the PCB This will cause a shift in the AVERAGE solder brick height. I have witnessed this again and again using laser mesurement techniques. Essentially you get a "scoop" shape rather than a square solder brick. This occures because the stencil to paste contact has drag as they seperate. | Also, verify your baseline measurement error. Especially when using lasers. Example: My (brand name witheld) laser height machine has a baseline error of 0.0004 inch error. This can easilly be determined by statistical techniques and a simple experiment. | Bottom line all these nominal and average errors accumulate. So, think volume. PS using statistical sampling techniques would account for this phenomenon and bring you back into the green! Think volume dude! | | | | Dean. | | Volume is the critical measurement when printing for BGA. Increasing the speed and pressure is not always the best way to go. There are other factors which come into play. Without knowing what screen printer you are using, I would suggest doing an experiment with the following as factors: Squeegee pressure, Squeegee speed, screen gap, Snapoff, Snapoff speed. And as someone says above, use metal squeegees. I have had more consistent paste deposition when I slow the squeegee down, with less pressure, and adjust the snapoff, snapoff speed, and screen gap. There are some special supplements to SMT Magazine that are excellent: Guide to Soldering and Printing, and SMT in 10 Easy Steps.

Good Luck Brian

reply »

| Dear guys, | | I have some problem on assembling micro-BGA with solder paste (63/37). The micro-BGA we assmeble is 0.75mm pitch and with 45 pin (solder bump is 96.5/3.5Ag). The stencil is a laser-cut stencil with 5 mil thickness. We are now using a no-clean solder paste, with the metal mesh size 325. We find that after stenicl printing, the paste height goes to 8 mil! Where our internal spec is 5-6 mil!!!! | | I increase the printing pressure and printing speed, but no help! | | Some people suggest that I should | (i) use a nickel-plated stencil, thus the paste can leave the stencil easily; or | (ii) use size 4 powder. | | Is it really help? DO you have any experience on assembling micro-BGA with solder paste? | | Thanks in advance. | Vic

We had release problems when we tried using circular apertures on the stencil. While releasing from the stencil, the brick would pull up around the edges making the brick much taller. I altered the stencil apertures for the BGA to make them square and cut down on the size of the square to match the volume of the circular pad. We got better separation from the stencil and a much more repeatable volume. The corners of the square actually hung over the pad edges but reflowed into the joint very nicely. Hope it helps!

Kris

reply »