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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


solder balls

what is the main reasons that casuing solder ball during ref... - Jul 05, 2001 by

steven

#7171

solder balls | 5 July, 2001

what is the main reasons that casuing solder ball during reflow? and how to remove them? (mostly arround the connector leads or between the 2 0402 chips). solder paste is RMA type. need advise. thanks

reply »

#7173

solder balls | 5 July, 2001

Our solder ball problem is linked to moisture. If the paste sits on the board too long (especially during our humid summer months) we experience the problem. As far as the problem being specific to components or areas of the board, maybe some of the experts can help.

reply »

Dougie

#7174

solder balls | 5 July, 2001

Steven,

There is a load of info on this if you check through the archives. Quick pointer though are: Solder balls are caused by paste creeping under the component at placement. The part is placed and the paste is squashed under the component, when the paste reflows it forms a ball underneath the component and works it way out to the side of the component. Exactly the symptom you are seeing. Paste gets too far under the components for a number of reasons, amongst them are:

Paste slump Z axis placement height

Look at these first. Check your profile also.

Good luck.

Dougie.

reply »

Kenny Tan

#7181

solder balls - Possible root causes | 6 July, 2001

Hi there Steven ! Please allow me to share some of my problem we're facing here with you. Here are a few points that I suggest you should check on your current process:-

(1)PCB - Component land partern design ? It is per IPC spec ? - Masking been printed correctly on the board.

(2)Solder Paste - What type of ball size/pitch that you using ? Are the paste ball size over size per your product design - Expire date & thawing time ?

(3)Stencil - Type of stencil - Chemical edge Score = 3/10 - Laser cut Score = 7/10 - Nickle Plating Score = 9/10 - Correct stencil opening & design ? 1:1 (or) 0.9:1 - Correct stencil thickness ?

(4)Paste Printer - Cleaning Process effectiveness/Machanism functioning. - Are you ercycling the cleaning paper ? - Your wipping paper parallism to the bottom of your stencil. - No variances underneath of your stencil this might disturb your cleaning process. - Your stencil condition. How long it's been used ? - Printing speed , pressure & seperation speed setting. - Type of cleaning solvent per your solder paste ? Any Checmical reaction ? - Your machine humidity and temperature ? - Type of squeegee & squeegee condition. - Are the printer board directly transfer to chip mounter ? ( Please avoid operator handling ) - Are there any misprinted board been recyle ? DO NOT ALLOW MANUAL CLEAN! Suggesting using ultrasonic - ( Please get the components supplier consult before using )

(5) Chip Mounter - Correct Z height for your component placement ? Do not over press the solder paste. - Nozzle no leaking ? Which could blow away your solder.

(6) Reflow oven - Oven setting parameter. Proposed a slow activities profile at the preheat and short with enough temp at the peak zone.

Best Regard, Kenny Tan

reply »

steven

#7182

solder balls | 6 July, 2001

Thanks guys for all the input. i'm still struggling to find out the main cause. at time being, urgently need a suggestion of a effective way to remove all these solder balls. THANKS AGAIN.

reply »

#7186

solder balls | 6 July, 2001

We find no solder balls after washing our boards. It works GREAT!!! ;-)

Probably more to the point, look at "Circuits Assembly" 7/01, p 40, "Eliminating Solder Beads In No-Clean PCBs".

"Solder mask type & composition are the greatest contributor to solder balls." ("SMT" 6/98 p66)

reply »

Mag

#7202

solder balls | 10 July, 2001

Steven, Dave is right, wash them away. Since it is RMA paste, you may need to have a saponifier to go for Semi-aqueous washing. Or you may want to contact your supplier for a cleaner to wash it away with some agitation ( ultrasonic).

Mag.

reply »

Hussman

#7230

solder balls | 12 July, 2001

Wow, what a string of answers and nobody brought up the easiest way to eliminate solder balls. Yes, the paste being squished under the part does reflow and work it's way out to the side during reflow. I did a 5 week study on this (when I was a boring person) This result will occur most of the time no matter what paste, environment, or process parameters you use.

The best way to eliminate them for good is to use the homeplate design on your stencils. It just plain works. It gets rid of the extra paste that gets squeezed under the part, but also holds the part to the board. Does not affect your solder joint at all. No need to requalify your paste, take a year to reprofile all your product or add any steps to your process. Call your stencil manufacturer and ask them to reduce the aperture openings 10% and homeplate all your Rs and C's up 1206. If you have solder balls on larger devices, you can homeplate them too, but go 1:1 and watch your stencil openings. You want to leave enough past there to hold the part to the pad.

reply »

#7239

solder balls | 12 July, 2001

One possible reason no one brought-up "too much paste" is that suggestion has been proffered so many times that it�s in the archives like the layers of sediment making-up Kataden.

Another thing [we can argue about this], using home plate apertures with 0402 increases tombstoning, because it is very difficult to balance the temperature on both ends of the component with such narrow solder target to hit in placement. It might be better argued that a reverse home plate might work better, though.

reply »

Hussman

#7262

solder balls | 16 July, 2001

No argument here. But I've don't think I've ever seen solder balls on 0402s. Even when they're placed on narrow G dimension pads, where there is little toe fillet, I've never seen a solder ball. But your idea of reversed homeplates would work great if I did!

reply »

Michael Parker

#7264

solder balls | 16 July, 2001

Just received "Circuits Assembly", July 2001 issue. An article regarding solder balls and aperture design, beginning on page 40. This DOE had aperture, reflow profile and no clean pastes (3 types) as attributes. It is advocated that a U-shaped aperture eliminates solder beads. No mention of chip sizes in the article, just quantity of chips in the experiment. The article also claims solder fine defects are eliminated on fine pitch QFP's with U-shaped apertures.

Can the community at large make comment on U-shaped apertures? Does it work for all chip sizes? Is it a no-clean only design?

reply »

Steve

#7265

solder balls | 16 July, 2001

The solder balls you are talking about are caused by too much paste. Reduce the size of the stencil aperatures.

Concerning removing the solder balls, the first thing you need to ask yourself is, do I need to remove them. IPC-610, 12.4.10 states, "Acceptable: Entrapped or encpsulated solder balls that are within 0.13 mm of lands or conductors, or exceed 0.13 mm diameter." Also acceptable is if they do not violate minimum electrical clearance.

If they are encapsulated in no-clean residue, I would leave them.

If they are not encapsulated, a simple water wash will take them off.

reply »

steven

#7266

solder balls | 16 July, 2001

I have no choice, as per customer's request to remove the solder balls. Can i use water to clean away the solder ball since i'm using RMA solder paste not the water soluable paste.

reply »

steven

#7267

solder balls | 16 July, 2001

I hv no choice as per customer's request to remove the solder balls.

Can i just use water to clean away the solder balls since i'm using RMA solder paste and not the water soluable paste?

reply »

#7268

solder balls | 16 July, 2001

Sure, use of U-shape apertures is another way to reduce the amount of paste that you apply. That and similar schemes mentioned in the article are not a function of the flux used.

We have used U-shaped apertures for MELF and mini-MELF. They are suggested in SM-782A. We never did a good job of structuring a study, collecting data, bla bla bla to be able to understand if there was an advantage. Contrary to the results of the article you quote, our sense is there was no benefit, so we never pursed it. But then again, we don�t get ourselves in lather about solder balls. We were concerned about the fine MELF rolling off the fine board.

We cannot image on a QFP with a U-shaped aperture. Sounds wacked!!! But let�s explore this further by applying the guidelines from the article with a QFP44 * Yer pads are 71X16 [all units thou], say. * You whack that by 85% and get 62X14 for the outline of your aperture [which is OK]. * So in making the �U�, half of the aperture is 31X14 [which is OK], but the other half of the aperture is two strips 31X4 [which could not be OK, even with a can of Dr. Brown�s Celery Soda]. *These two strips are bad news from a paste release stand-point. Assuming a 6 thou stencil, the aspect ratio is 0.67 [SB GT 1.2 for laser cut] and the area ratio is 0.08 [SB GT 0.6 in order to be a bad news mutha bear. So, 0.08 is like a bad news mutha bear in need of a laxative.].

Referring back to the article, there is a bunch of running and confused sentences [which I think is the purpose of many of these trade journal articles, but that�s another thread.]. My take on their recommendations are: * Recommended pinching QFP apertures, but never directly suggested using U-shaped aperture for QFP. * Recommended using U-shaped apertures for chip components.

[Mike, I know you work at wuchacallit, but I don�t know the address. Pls e-mail me a post location so I can send you a front page NY Times article on some fine people hanging-out Hollister.]

reply »

#7269

solder balls | 16 July, 2001

Cleaning RMA with water is bad news. It will turn your solder connections white and remove none of the RMA residues.

There was a thread on removing RMA within the past two weeks on SMTnet. Check the fine SMTnet Archives. Additionally, look at Mag's post on this thread on July 10.

Certain RMA fluxes, such as Alpha 615-15, can be left on boards with good results if: * One isn't pin testing after rework. * DC or low frequency. No high speed digital or high frequency [over 1 Mhz ]. * Flux residue is not left on a switch contact. This can cause bad contact resistances. * A bit of slightly unattractive yellow residue is ok. Most RMA residues should not cause any corrosion or current leakage problems.

reply »

Michael Parker

#7278

solder balls | 17 July, 2001

Dave - I agree that the U shape for QFP's sounds a bit whacked, for the reasons you state. It may work OK for larger pitchs but then again there are a myriad of fixes that work before resorting to U shapes. With the U shape removing paste from under the lead I did wonder about the mechanical bond relying solely on heel and toe fillets. Maybe not such a big deal. It would certainly remove excess volume that wicks up the leads or bridges across leads. It might be worth exploring for long footprints to remove the middle section of the paste if bridging is prevalent. I too have used U shape for MELF's. It makes sense for keeping the buggers from rolling around. I did a direct e-mail to the authors of the article, they said the U shape is OK down to 0402's. 0201's have such small paste volume to begin with that the U shape would cause more problems.

reply »

Mag

#7286

solder balls | 18 July, 2001

Steve, To clean the RMA processed boards, u hv to make sure u talk to the supplier first, then get the right saponifier and do semi-aqueous cleaning.

reply »

Mag

#7287

solder balls | 18 July, 2001

Steven, You CANNOT just use water to clean away the solder ball when u hv used RMA paste.

reply »

David Chapman

#7296

solder balls | 19 July, 2001

Solder balls or mid chip beads? Remember IPC classification for Class 2 allows mid chip beads encapsulated in flux. Usually on R's and C's first thing I would do is verify .006 laser stencil 10% reduction and the most critical, The paste must be on the pads not on the mask. No matter what you do if you are printing off the pads the solder will ball and spatter. Look under the microscope and align the paste to pads as perfect as you can. Humidity control and good paste handling practices are critical. When we do have unacceptable balls we use our Smartsonic cleaner for removal.. Dave C.

reply »

#7309

solder balls | 20 July, 2001

David,

I've thought about doing this also with no clean boards that were poorly cleaned following a misprint (and subsequently reprinted, populated, and reflowed) but have shyed away because of the potential for damage to components from the ultrasonics. Have you done any testing to validate that process? What kind of power levels are you using, duration, etc.? Thanks.

reply »

#7312

solder balls | 20 July, 2001

Circuits Assembly magazine 09/99 "Defluxing with Ultrasonics" by Les Hymes. [There's more but I can't find it right now. I'll add to this as I find the junk, er very nice stuff.]

J-STD-001C states that its OK to use ultrasonics for cleaning so long as you run tests to make sure you aren't damaging anything [er, words to that effect]. Your very nice hard copy of TM-650 has a cuppla tests you could run on the cleaner. ;-)

We'd be concerned about damaging components that were not encapsulated [eg, crystals, metal can IC, ceramic packages, BGA with glued-on plastic covers, etc].

reply »

Win Source Online Electronic parts

IPC Training & Certification - Blackfox