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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


SMT voiding

Views: 5607

#67279

SMT voiding | 17 September, 2012

We are experiencing voiding on one of our lead-free assemblies on every part. The voiding only occurs underneath the parts and not in the solder fillet (see pic). We have tried baking, curing, washing, multiple pastes and everything else we can think of. We have come to the conclusion that it is outgassing from the ENIG pads themselves. Does anyone else have experience with this problem and found a solution?

Attachments:

reply »

#67349

SMT voiding | 25 September, 2012

Ken ... I resized your picture so that it fits on the page better.

Continuing with your outgassing through the solder pads theory, how thick is the copper on the pads? I don't expect to see outgassing through pad that are a thou or more in thickness.

I have questions ... * Are these voids in the solder connection or are they in the pads? * I'm confused by your picture. What the hell am I looking at? I assume the bluish thing in the center is a pad with voids under the component and no voids in the fillet [to the right and left] * What variations in the thermal recipe have you tried?

reply »

#67351

SMT voiding | 25 September, 2012

Thanks for fixing the picture. What you are looking at is a pad that the part was pried off of after reflow. The right side of the pad shows a good solder joint and underneath the part is where the voids occur. The outter solder joints show no voids on any part in x-ray, it only occurs underneath the parts. I will need to find out what our pad thickness is.

reply »

#67355

SMT voiding | 25 September, 2012

The more that I chew the idea of outgassing through the pad theory, the less I like it. Why would something, water???, choose to outgass through the pad when moving through the solder mask would be a path of much less resistance? And why wouldn't the whole pad have voids, rather than just the portion under the component leads?

OK. Let's step away from the board and think about it differently. What if there was a via or some other feature under the pad that was causing this? Not likely, because you said it affected all components on this board. So, each pad is unlikely to have similar features underneath.

Can't be the components, because it affects all components and they're supplied from different sources.

Can't be the solder mask, because the outgass is unlikely to find a path to get under the component pads.

All of this gets me back to solder paste and/or thermal recipe. I think the conveyor on your reflow over is running a warp speed and the nonmetal stuff in the solder paste just isn't getting enough time to release.

How about if you paste-up a board and reflow it with no components? My bet: No voiding.

reply »

#67389

SMT voiding | 30 September, 2012

We believe that Dave's analysis above is mostly accurate. Run a board with no components, and check for voids.

We'd also like a little more information, though, to clarify. What type of component is experiencing the voiding? You describe the voiding as "under" the component...are we talking about ground pads, or blind lands under components? The types of lands that are prone to voiding due to the mechanical nature of the part? Is the voiding isolated to any particular area of the board/components? Is the voiding isolated to a single assembly for you, or is it being experienced across your manufacturing line?

Cheers, ..rob

reply »

#67411

SMT voiding | 2 October, 2012

Have you come to a conculsion on this yet? If so please post for all to review.

reply »

#67417

SMT voiding | 3 October, 2012

Well what we have seen so far is that the voiding occurs on nearly every part, but only underneath the terminations. When we run a board with no parts we do not see any voiding as there are no terminations to trap the out-gassing from the pads. Next week we will be running trials with different pad finishes and solder pastes

reply »

#67426

SMT voiding | 3 October, 2012

Given that you're seeing it on nearly every part, we tend to lean towards Dave's explanation. We wouldn't expect outgassing from the pad to be an issue (at least, we've never heard of it before, which doesn't, necessarily, eliminate it as a potential).

We would also consider the following possibilities: 1. Reflow oven profile 2. Paste type/age 3. Component age/exposure 4. Board finish age/exposure

Cheers, ..rob

reply »

#67552

SMT voiding | 21 October, 2012

We have had this issue on several parts in the past, typically we use a grid of squares of solder paste on large flow areas, This helps reduce but not eliminate all voiding.

I would be very interested in hearing what you come up with.

O and what is the accutal part size?, Pad size?

reply »

#67558

SMT voiding | 23 October, 2012

I have attached a PDF overview of some test we have done and the results. (fixed oven profile)

Attachments:

reply »


aj

#67565

SMT voiding | 25 October, 2012

great report and good to see results

reply »

#67566

SMT voiding | 25 October, 2012

Ken, The photo looks like results I have seen because the gold flash was too thick on the board. This causes a condition known as gold imbrittlement. You may want to get an analysis of the plating thickness. I'm not real experienced with immersion silver but I can tell you that the few times I worked with it we had problems. Another thing I noticed is that your TAL on your profile seems high (for SAC). This could potentially worsen the condition if you have a gold thickness problem. The longer above liquidous the more the gold will dissolve into the intermetallic. Try lowering your TAL to like 60 seconds and see what happens.

What is the solder paste alloy ?

reply »

#67567

SMT voiding | 25 October, 2012

Here's a paper that may help http://www.ipcoutlook.org/pdf/assembly_challenges_bottom_terminated_ipc.pdf

Temperature Impact * Profile didn’t have a significant impact on voiding. * Voids slightly increased with higher temperature.

Reflow Atmosphere Impact * Nitrogen didn’t help to reduce voiding amount * Nitrogen could result in less voids but larger voids.

Orientation Impact * Orientation didn’t impact the voiding significantly.

SMD vs. NSMD Pad * There was no significant difference in voiding between solder mask defined pad and no solder mask defined pad.

Component Type vs. Voiding *Voids varied depending on the package type. [see Figure 11 below]

Design Impact * More solder balls were seen on LGA packages. Solder balls were rarely seen QFN packages.

Attachments:

reply »

#67568

SMT voiding | 25 October, 2012

Wait a minute...Your peak temps are PbFree like. Why is your profiler looking at time above 183 ? What is the time above 217 ?

I thought we were looking at time above 217 not 183.

reply »

#67569

SMT voiding | 25 October, 2012

Yes it looks like I attached the wrong reflow profile and I'll get the new one in there. Our TAL should be between 60-90 sec over 217. I'm not sure the embrittlement would be an issue as we have seen this issue across at least 3 different board vendors, but we will look into it.

reply »

#67577

SMT voiding | 1 November, 2012

Hello kkay, can I ask what type of solder paste the #4 sample was? Thanks!

reply »

table top X-ray inspection

StikNPeel™ Rework Stencils