Can We Add Weights to Components During Reflow?

Phil
Welcome to Board Talk with Joe Belmonte, Jim Hall and Phil Zarrow of  ITM Consulting coming to you today from Board Talk Towers on mount Mount Rialto in the state of insanity ... oh, New Hampshire.  Today we have a question about reflow.

Jim
This question comes to us from C.B. "We have issues with small, light BGAs. They often have opens after reflow. We have no problems with larger BGAs or fine pitch BGAs with higher ball counts. Do you think it makes sense to add a weight to the top of small BGAs during reflow to improve soldering and reduce opens? Any advice would be appreciated."

Phil
It sounds like any advice would be desperate at this point. Adding weights? I haven't run into that one, yet.

Jim
How about no?

Phil
I would venture to guess a possible contributing factor could be to look at your solder paste deposits. Perhaps they're too thick for the pad size and you're getting the floating BGA syndrome. Also look at the Z down-force pressure on your pick and place machine. Are you putting it approximately halfway, one-third to two-thirds of the way into the solder paste as you should?

Jim
Since these are not fine pitch devices. They are light but, I assume, they're larger in dimensions, horizontal dimensions. You may have an incorrect re-flow profile that's causing excessive warpage, which is causing the opens to occur because the part is warping too much during the re-flow process.

Joe
I certainly agree with my colleagues. You should not do this. This is a situation where we really have to think about whether we really understanding or working towards the root cause? So often in our travels, we see people band-aiding processes and doing things, such as this, which are not the right thing to do as opposed to doing the work that's required to find the root cause of the problem, come up with a solution, verify the solution, and then institutionalize it.

That is the proper approach to a problem like this because this is not an industry-wide problem. This is unique to you.

We always say, when I worked at Motorola, the job isn't done until you institutionalize the solution, because if you walk away before the documentation is updated, the equipment's updated, the training completed, whatever it happens to be, then you haven't completed the task.

So what I would do here is I would certainly start brainstorming possible causes. Bring your folks together, the appropriate technical folks, even your supplier, even your customer, maybe 'cause it could have something to do with board design, and brainstorm possible causes and start running experiments and studies to identify the root cause, come up with solutions, test the solutions, and then institutionalize them and that's true in any problem such as this.

Phil
Couldn't have said it better myself. Well thanks for listening to Board Talk and this is Jim Hall, Phil Zarrow and Joe Belmonte.

Jim
Whatever you do, don't solder like my brothers.

Phil
And don't solder like my brothers, including Joe, even though he won't admit it.

I would be curious which board finish they are using. The issues we've had with opens on BGA's proved to be non-wettable pads from improperly processed ENIG boards. The nickel had oxidized and was rejecting solder resulting in an open.

In short, Absolutely NOT! Difficult to analyze without all the facts. Things to look at;

* Board finish, (ENIG, possible black pad?)
* Board storage, boards uncontrolled, no MBB, or desiccant over length of time could develop intermetallic growth rendering them (PCB) unsolderable, same applies to components. Are BGA's properly stored or baked prior to subjecting devices thru reflow, could components be warping due to moisture within the BGA package?
* Is package correctly defined, in drafting component library and on PCB.
* Stencil apertures are they 1 to 1 or is there an aperture reduction? Sometimes modifications can be made to apertures (shape and size) to adjust target solder paste volumes. However wouldn’t recommend this until reviewing all other alternatives as a root cause.
* Solder paste deposit volumes, verify printing accuracy and target paste volume across entire BGA area.
* Oven profile, should be verified to ensure desired temperatures are achieved. Thermocouples and mole required to verify, something like the Datapak is good tool for profile verification.
* And finally board warpage, how are boards transitioning in the oven. Is there a fixture or possible rail to tight? Board thickness may come into play, if board is thin, may need a fixture to support thru reflow.

Hope one of these items helps you resolve your issue, just remember what ever you do don't solder like either of the brother's.

Hector A. Maldonado

===

Just to add to my previous comment, the type of solderpaste used in your application is also a factor. Are you running lead free or Sn63/Pb 37? And BGA sphere composition also to be considered, the type of spheres would define temperatures needed for proper solder wetting, manufacturer should be able to provide temperature guidelines for the device in question.

And when printing is there sufficient PCB support under device as its being printed, in other words you need to ensure that board height does not fluctuate during printing this in combination with excess squeegee pressure could result in uneven paste deposits.