Micro Vias in Pads

Anyone have an idea how small a via must be to inhibit significant solder volume escaping into the via?

reply »

| Anyone have an idea how small a via must be to inhibit significant solder volume escaping into the via? Typically, micro vias are drilled and plated to about .003" diameter. To avoid significant solder "drain off," they should be drilled as shallow as possible consistent with specified dieletric thickness. This usually means to a depth reaching through a dielectric thickness of not more than 5-7 mils. The solder still will drain off but the via can be filled as part of the printing process. Be careful of outgassing back up through the via causing solder "blow outs" effecting unacceptable solder joints. Characterize your solder process to address and prevent this problem. Another, less costly way to avoid micro vias all together is to drill "normal" holes in SMT pads, when possible, fill them as part of the relamination process, then plate over the top providing a smooth, flat surface that is hole free.

reply »

We have successfully put .016" vias in .030" pads. There is no drain off with .030" solder balls, since the actual volume in the .002" depression is low. We did notice random bubbles where, apparently, the solder reflowed and trapped the air in the pad. Extensive reliability testing showed no adverse affects. Steve | | Anyone have an idea how small a via must be to inhibit significant solder volume escaping into the via? | Typically, micro vias are drilled and plated to about | .003" diameter. To avoid significant solder "drain | off," they should be drilled as shallow as possible | consistent with specified dieletric thickness. This | usually means to a depth reaching through a dielectric | thickness of not more than 5-7 mils. | The solder still will drain off but the via can | be filled as part of the printing process. Be careful | of outgassing back up through the via causing solder | "blow outs" effecting unacceptable solder joints. | Characterize your solder process to address and prevent | this problem. | Another, less costly way to avoid micro vias all together | is to drill "normal" holes in SMT pads, when possible, | fill them as part of the relamination process, then | plate over the top providing a smooth, flat surface | that is hole free.

reply »

We have successfully put .016" vias in .030" pads. There is no issue with .030" solder balls, since the actual volume of the .002" via in pad is small compared to the ball+solder paste volume. We did notice random bubbles where, apparently, the solder reflowed and trapped the air in the via. Extensive reliability testing showed no adverse affects. Steve | | Anyone have an idea how small a via must be to inhibit significant solder volume escaping into the via? | Typically, micro vias are drilled and plated to about | .003" diameter. To avoid significant solder "drain | off," they should be drilled as shallow as possible | consistent with specified dieletric thickness. This | usually means to a depth reaching through a dielectric | thickness of not more than 5-7 mils. | The solder still will drain off but the via can | be filled as part of the printing process. Be careful | of outgassing back up through the via causing solder | "blow outs" effecting unacceptable solder joints. | Characterize your solder process to address and prevent | this problem. | Another, less costly way to avoid micro vias all together | is to drill "normal" holes in SMT pads, when possible, | fill them as part of the relamination process, then | plate over the top providing a smooth, flat surface | that is hole free.

reply »

| We have successfully put .016" vias in .030" pads. | There is no drain off with .030" solder balls, since | the actual volume in the .002" depression is low. | We did notice random bubbles where, apparently, the | solder reflowed and trapped the air in the pad. | Extensive reliability testing showed no adverse affects. | Steve | | | Anyone have an idea how small a via must be to inhibit significant solder volume escaping into the via? | | Typically, micro vias are drilled and plated to about | | .003" diameter. To avoid significant solder "drain | | off," they should be drilled as shallow as possible | | consistent with specified dieletric thickness. This | | usually means to a depth reaching through a dielectric | | thickness of not more than 5-7 mils. | | The solder still will drain off but the via can | | be filled as part of the printing process. Be careful | | of outgassing back up through the via causing solder | | "blow outs" effecting unacceptable solder joints. | | Characterize your solder process to address and prevent | | this problem. | | Another, less costly way to avoid micro vias all together | | is to drill "normal" holes in SMT pads, when possible, | | fill them as part of the relamination process, then | | plate over the top providing a smooth, flat surface | | that is hole free. Steve, I concur concerning such a depression in such applications. However, I have seen significant problems with HASL, micro vias with some fab suppliers. I worked with vias in BGA pads early in the 90's and have much data the solderability problems I addressed. Again, I agree the problems typically do not occurr. However, one should be aware they could arise and bite you when and where it hurts.

reply »

| We have successfully put .016" vias in .030" pads. | There is no issue with .030" solder balls, since | the actual volume of the .002" via in pad is small | compared to the ball+solder paste volume. | We did notice random bubbles where, apparently, the | solder reflowed and trapped the air in the via. | Extensive reliability testing showed no adverse affects. | Steve Steve, Is the 0.016" dimension a depth or a diameter? Also: was this a CSP / micro-BGA appliction? My company is considering several different CSP packages for new designs (e.g., Tessera's uBGA, Amkor's Chip Array @ 0.8mm and 0.5mm). One potential problem for us might be cleaning our water-soluble flux from beneath the CSP, since the standoff is so small. Do you have any experience or information on this topic? P.s. I tried to email you directly, but the message bounced back (?).

reply »