sjm1027
11-26-2007, 03:53 AM
>*Currently our PCB designs utilize "encroached" solder mask via pads on
>pin escapes for all of our products.
>This was an adequate treatment for products with 1.27mm but as BGA lead
>pitch starts shrinking, the issue of adequate solder dams between ball
>pads and their pin escape vias becomes more problematic.
>Our standard pin escape via pad is typically .024" with .012" Finished
>Hole Size (FHS). Our requirement is solder mask encroached 1:1 with FHS.
>*First dilemma is that fabricators require .006" solder mask clearance
>over primary drill size.
>*This yields .001" solder mask encroachment on a .024" via pad with a
>.012" FHS under ideal conditions with no mask misregistration.
>*Problem becomes more pronounced with .8mm BGA pin escape vias. Our
>typical pad geometry for .8mm BGAs is .016". The pin escape via pad
>diameter for these devices is typically .018". That will yield only
>.005" clearance between the via pad and adjacent BGA pad. If
>fabricator requests .006" over primary drill, that leaves 0 solder mask
>on pin escape via pad with perfect solder mask registration. Problem is
>further exacerbated when fabricators increase solder mask apertures to
>accommodate misregistration and prevent solder mask from getting onto
>SMT pads. This results in as little as .003" solder dams.
>*Predict that there will be an abundance of issues whenever parts need
>to be reworked.
> Our preferred PCB surface finish is ENIG with SMOBC construction.
>*Typical practice in industry is to "tent" over vias with LPI. However,
>this practice is problematic and can lead to long-term PCB reliability
>issues.
>*Tenting vias with LPI solder mask will cause ENIG plating chemistry to
>become entrapped in vias.
>I would like to know how others are handling vias under BGA's with a
>pitch of .8mm or less.
>
>If we plug our vias we will have the following issues:
>
>*
> Via fill process will increase cost of PCB and will
>potentially increase lead time as well.
>*
> *Need to have vias covered with solder mask on BGA placement
>side to prevent potential for solder wicking onto via land. This
>becomes very critical during rework operations.
>
>*
> *Need to have vias exposed on non-placement side for test
>access.
>
>*
> *Vias filled with non-conductive media will cause probe
>contact issues during flying probe and In-circuit test.
>
>*
>Any help would be appreciated
>
>pin escapes for all of our products.
>This was an adequate treatment for products with 1.27mm but as BGA lead
>pitch starts shrinking, the issue of adequate solder dams between ball
>pads and their pin escape vias becomes more problematic.
>Our standard pin escape via pad is typically .024" with .012" Finished
>Hole Size (FHS). Our requirement is solder mask encroached 1:1 with FHS.
>*First dilemma is that fabricators require .006" solder mask clearance
>over primary drill size.
>*This yields .001" solder mask encroachment on a .024" via pad with a
>.012" FHS under ideal conditions with no mask misregistration.
>*Problem becomes more pronounced with .8mm BGA pin escape vias. Our
>typical pad geometry for .8mm BGAs is .016". The pin escape via pad
>diameter for these devices is typically .018". That will yield only
>.005" clearance between the via pad and adjacent BGA pad. If
>fabricator requests .006" over primary drill, that leaves 0 solder mask
>on pin escape via pad with perfect solder mask registration. Problem is
>further exacerbated when fabricators increase solder mask apertures to
>accommodate misregistration and prevent solder mask from getting onto
>SMT pads. This results in as little as .003" solder dams.
>*Predict that there will be an abundance of issues whenever parts need
>to be reworked.
> Our preferred PCB surface finish is ENIG with SMOBC construction.
>*Typical practice in industry is to "tent" over vias with LPI. However,
>this practice is problematic and can lead to long-term PCB reliability
>issues.
>*Tenting vias with LPI solder mask will cause ENIG plating chemistry to
>become entrapped in vias.
>I would like to know how others are handling vias under BGA's with a
>pitch of .8mm or less.
>
>If we plug our vias we will have the following issues:
>
>*
> Via fill process will increase cost of PCB and will
>potentially increase lead time as well.
>*
> *Need to have vias covered with solder mask on BGA placement
>side to prevent potential for solder wicking onto via land. This
>becomes very critical during rework operations.
>
>*
> *Need to have vias exposed on non-placement side for test
>access.
>
>*
> *Vias filled with non-conductive media will cause probe
>contact issues during flying probe and In-circuit test.
>
>*
>Any help would be appreciated
>