Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Making a BGA stand up

JohnW

#22976

Making a BGA stand up | 13 January, 2003

Hi folk's,

I'm looking for some help full suggestions here...

I've got a ceramic BGA with tn/lead / bismuth balls on it that is over collapsing due to the weight of the momponnet vs the ablity of theball to suppot them. We've spok eto the component vendor to get them to put copper stand off's on them going forward bu they ahve already made a fair number that I'll have to use for one of our customers. I'm trying to find a pad or something to place under the device since it's a not a full array. The pad has to be non electrically conductive and cannot absorbe moisture....any one any thoughts? Alternatives to this process would also be welcome, we've thought of applying some glue dot's to act at stand off's but there are concern's if you ever have to repair these things and alo if you miss a glue dot (not that this would happen you understand) you could skew the device in one corner.

all sensible answers welcome....

John

reply »


RDR

#22978

Making a BGA stand up | 13 January, 2003

Could you provide the specific alloy info (percentage od each) OR provide the melting point of this alloy? Please also provide the component specs (pitch, ball diam.) And the pad size and design used on the board.

Russ

reply »


RDR

#22979

Making a BGA stand up | 13 January, 2003

one more thing, how many of these are you planning on using? It might be advisable to have the supplier strip and reball with a more appropriate alloy to prevent ball collapse.

reply »

JohnW

#22986

Making a BGA stand up | 13 January, 2003

Russ,

unfortunately at the moment the preferred option of hi melt isn't available due to copyright issues. The current ball is made of 46/46/8 tin / lead / bismuth. ther ball's are 0.8mm in dia. The alloy actually has a plastic region around 170 and fully melts at 180 deg C, fun right....?

reply »


RDR

#22988

Making a BGA stand up | 13 January, 2003

Youch!

I know this is going to sound stupid but.... I have done it in the past in limited quantities. We placed capton (polyimide) tape at each of the four corners for a similar situation. It did work but the operator has to be careful in placing the tape as not to smudge paste, cover pads, etc... The tape had to be layered to the proper thickness and then cut to proper size. This process is much easier if you have the option of placing the BGA without paste and reflowing onto bare pads (I don't know if this is even feasable with this alloy however).

Sorry i don't really have any better ideas. I think the glue process may be a good option even with the stated concerns Glue can be softened/removed with both solvents and a soldering iron if rework was necessary.

Just out of curiousity, why was this alloy used instead of a hitemp or even a 63/37 alloy?

You might also post this in the SMT production forum as there are some very experienced and knowledgeable people that contribute there (they may not be looking in the design forum).

Good luck to you Russ

reply »

#22994

Making a BGA stand up | 13 January, 2003

Haaaaaa!!!! You contractors get all the fun don't ya???

Consider: * Replacing the balls on the corners of the devices with balls composed of a "hard solder", yano something like 90/10. * Shimming to prevent excessive collapse of a heavy BGA during reflow: ** Dispense adhesive dot to prevent more than 30-35% ball collapse. ** Snip off some stainless steel shim stock and just tack them to the PWB with a drop of flux. Remove after reflow, but before clean, so they don't get wedged somewhere they should not. ** Use glass spheres (adhesive staked), but they are a nuisance to get hold of the right size. ** Use aluminum wire. ** Stack two defect marking arrows under each corner (should equal 0.020" standoff).

Finally, you could affect the ability of that heat spreader to dissapate power, if you place a thermal insulator in the space between the board and the thermal pad under the die on the bottom of the device.

reply »

JohnW

#23033

Making a BGA stand up | 16 January, 2003

1st off for the forum folk, I don't like the fact that the new system only let's you see the posting your replying too, I'd rather be able to see the thread.....

Anyhoo, why this alloy?, it's because it's the device manufacturers preferred and the customer has all the life testing data for it and non for std tin/lead ot tin/lead/silver.

Dave, like the idea's, similar to our train of thought, the problems with these are: 1. can't use hi melt as this crosses soem patent with IBM which would add to the cost.... we have get to component vendor to add copper stand off's for future buildss and reduce the heat spreader thickness to reduce the overall component weight. 2. shimming the corners, yep work's great when your doing rework but not when your doing an sizable SMT run...lot's of smudged paste...it's similar for the stainless stuff plus as is the PCB designers want there are lot's of 0603's arranged nicely around the 4 sides to limit any space you use to have to play with Glass spheres may be an idea, I actually have some of various sizes for some work we were doing calibrating xray equipment.... I liked the repair arrows idea, not that we have many of them around....the downside is that again we'd nbeed to put these in the middle of the BGA rather than the corners which mean's leaving them in, the arrows are paper and the adhesive is some what hydroscopic so where there's humidity issues for some countries these wouldn't go well.

it's looking like some glue dot's in the middle of the decice, we're gonna run some trials to figure out how much glue you need to sdeispence to get a cured glue dot of X height, on the fun of it all.....

reply »

iman

#23069

Making a BGA stand up | 17 January, 2003

try using thermal pads under the area without the balls.

yeah, its a pain but we did it before for lot run of 500pcs BGA.

Not feasible for mass production run, you say? well, thats why there's a thing called DMF, and why its the designer guys responsibility to choose materials with DMF concession from the SMT process guys review of the whole board.

listen to the the tiny voice of the wunderful people who have to work long hour shifts running boards output, designed by guys with cushion-padded jobs seated in front of a brilliant dazzle-razzle PC. For the rare designer exceptions from my little O'world, keep up the good designer work ya'all been doing.

reply »

PCB Soldering Tools

Non-heated dispensing system