Why is NC the prefer process for BGA mounting? why not WS?

Mugen,

are you sure you get all of the WS residues out from underneath the BGA with your equipment ?

I think testing that will be quite difficult.

Wolfgang

reply »

A) The standoff height of some uBGA components does not allow water to penatrate under the component. B)Water soluable may dilute the flux and not remove it all together. This can be tested with Ion Chromatography. C) Water soluable in some cases needs more heat (How ever you want to call it slow conveyor, Higher temps, longer soak time) to activate the flux. D) Some balled packages are auto underfilled thus traping moisture, flux, contaminates under the component. Caldon W. Driscoll ACI USA 610-362-1200 cdriscoll@aciusa.org

reply »

1) why is No-clean (NC) the preferred process, for BGA mounting by SMT?

Probably the same reason NC is the preferred process for mounting non-area array SMT components.

2) why is water-soluble (WS) not a hot choice?

There�s not reason not to use WS. Cleaning of assemblies featuring BGA devices may seem at first consideration to be fairly easy due to the large standoff height achieved by the balls. The difference between fine pitch QFP and BGA devices is that it is initially difficult to get any cleaning solution under the parts and then remove it. This is due the leads preventing easy passage of the solution. The package standoff height from the board for a QFP is normally between 0.005"-0.015", which is not normally a problem. In the case of BGA the standoff height is 0.020", but the balls then obstruct the easy passage of the cleaning solution under the device and also impede its exit.

A simple trial to examine this effect involved mounting different types of BGA components on a glass plate then passing them through a batch cleaning unit. Visual examination then showed the effectiveness of solution penetration and removal. Trials repeating this experiment using a spray can show that flux can easily be flushed out from under the devices. Video taping this experiment provided a good indication of the cleaning potential. As the parts increase in size the effective drying of water based products may be more difficult but can easily be assessed using this technique. The use of cleanliness testing equipment to monitor ionic residues may also be used with very few problems. If required, parts may be attached to a piece of glass with flux placed under the device. This may then be tested to see the effectiveness of the test method. Using a specific quantity of sodium test solution would allow a specific reading to be obtained which could provide evidence of the equipment's ability to detect contamination. [Most of the fore-going is from comments made by Bob Willis, years ago.]

3) All I know, *is in layman terms*, that BGA has tendency to trap water, should WS process be the choice, and trapped water mingles with flux residue, to cause testing reliability problems, as reflected circuit shorts ......

Well, hopefully the water that�s loitering around under the BGA is not sitting in a puddle of flux res. Generally, the plan is to squirt enough water under the BGA that these flux res are flushed down the drain, euphemistically speaking of course.

Next, hopefully the air knife on the washer drying section moves the clean water that�s loitering around under the BGA out from under the BGA and reunites it with its aqueous brethren and sistern, not cistern, in the wash tank.

4) Being inquisitive bunch, and extremely technical in background, my end-customer is asking why NC? and they want to use WS process ......

Well, use NC if you like whining about voids in your BGA balls after soldering.

reply »

For one MUGEN Your original point number 5 asked for a technical point of veiw not any reference documentation. You get what you ask for.

Your appreciation is shallow, everyone that posted to your topic took time out formulate a responce.

So with that said I have lost the sudden urge to supply you with what you are looking for.

Cal

reply »

I can send you a DOE I did on WS vs NC on BGA�s, but the results were..

NC = Had significantly less voiding both in number of voids and size of voids. A longer soak time (reflow oven) resulted in less voiding both for NC and WS.

Mike

reply »