RE: BGA Assembly

["John Haddy" <john.haddy@radiata.com>]

Holes in pads = bad idea unless: they're microvias (< 0.1mm dia)
or they're blind vias.

Consider (assuming flux only attachment): 0.2mm finished hole in
a 1.6mm thick board needs 0.05mm3 of solder to fill it. A BGA ball
of, say, 0.75mm dia (std. size for 1.27mm pitch BGAs) contains
0.166mm3 of solder. Because of the high surface area to volume
ratio of the hole, a capillary is formed that _will_ fill with
solder in preference to the ball remaining intact. So 30% of the
available solder disappears into the via.

Let's assume that a normally collapsed solder ball after reflow
sits at 80% of it's original diameter (this is a guess). Then the
full-volume balls (those without vias in pads) will end up 0.6mm
high. The reduced volume balls, however, have a _natural_ diameter
of about 0.6mm, i.e. without collapse. Hence you'll end up with an
unreliable joint at best, which won't stand up to temperature
cycling, even if you are able to get an initial connection.

I would suggest that you proceed as Philip Decker suggested - i.e.
fill the vias with solder before attaching the BGA. Better still,
use a mini squeegee (a carefully shaped popsicle stick will do) to
fill the vias with solder paste prior to placing the BGA - a little
practise will show you how much paste is required, and it's usually
fairly simple to remove any squeezed through excess paste.

With regards to the first board, I'd remove the BGA entirely before
cleaning the PCB and starting again. If the BGA itself is important,
reballing kits are readily available (for large pitch BGAs st least).

I still wouldn't use locating feet...

(But then again, I've never put through-vias in pads, either :-)

Cheers,

JohnH


> -----Original Message-----
> From: Dimiter Popoff [mailto:tgi_earth@yahoo.com]
> Sent: Friday, 24 November 2000 4:22 PM
> To: pci-sig@znyx.com
> Subject: Re: BGA Assembly
> 
> 
> John,
> thanks!
> 
> >What I can tell you is that there should be no need for any
> >form of limiting feet, etc. The surface tension of solder
> >balls on BGAs is sufficient to provide self centreing during
> >reflow.
> 
> Are you sure this would be the case also with the holes I have
> under each ball? I suspect it would be OK, but this was (is)
> my first board with BGAs and on top of this there is no room
> for errors... so I put the feet just to make sure I would not
> end up with a flat block of solder :-).
> 
> >Your problem (below) sounds very much like a lack of flux.
> 
> No, there was plenty of it. And the BGAs did not float in it
> because it did run through the holes. They did not solder well
> also the second time because the feet I had put were a bit too high
> (because of remnants of the glue, I suppose).
> 
> I am pretty sure I'll solder the second board free of problems, yet
> I need the first one working as well.
> 
> I suppose I'll go on with the feet for another while, because I noticed
> that through those holes, which have a pad for a decoupling capacitor
> at the opposite board side, more solder from the ball had flown just
> wetting the metal area; and having all the solder of such a ball
> run through the hole and destroy the connection (quite
> unlikely, I guess) is a mess I don't even want to think of :-).
> 
> Please advise (or speculate) on the above. I will be also very
> grateful if you have some other advice on that topic.
> 
> Dimiter
> 
> -----------------------------------------------------------------------
> Dimiter Popoff
> Transgalactic Instruments, Gourko Str. 25 b, 1000 Sofia, Bulgaria
> http://transgalactic.freeyellow.com
> Email: tgi@bulnet.bg, tgi_earth@yahoo.com
> Phone: 00359/2/9923340, 00359/2/9805997, Fax: 00359/2/9540384
> 
>