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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Solder Resist Solder balls

John Watt

#13229

Solder Resist Solder balls | 22 December, 1998

Folk's, I'm in need of some help / direction. I'm working on my thesis for a masters based on looking at the effect's of the solder resist promoting / reducing the amount of solder ball's during wave. I'm looking at thing's like colour, texture, composistion and mix(cross hatched / colour variations etc.) Does anyone have thought's/idea's/experience in this or could suggest some good reading material ? Thanks in advance

reply »

Earl Moon

#13230

Re: Solder Resist Solder balls | 22 December, 1998

| Folk's, | I'm in need of some help / direction. I'm working on my thesis for a masters based on looking at the effect's of the solder resist promoting / reducing the amount of solder ball's during wave. I'm looking at thing's like colour, texture, composistion and mix(cross hatched / colour variations etc.) Does anyone have thought's/idea's/experience in this or could suggest some good reading material ? | Thanks in advance | Ah yes, another PCB scholar. How different my life would have been had it not been for solder mask and all the affects it can have on my solder ball formations!

Not making any assumptions, I must consider you will never be using dry film (DF) solder mask. Some still do, but foolishly as it was a bain to modern electronics for so long. It played a poor role, as solder mask over bare copper (SMOBC), until the advent of wet or liquid photoimageable types (LPI).

The reason for this discussion is that there never was a method to determine DF cure - as it had no such properties nor was there a test method available to determine it. Often, solder balls were effected using DF. Also, its adhesion characteristics were always suspect as it did not "bond" well to anything (resin/glass substrate or bare copper conductors).

LPI solder mask is composed of nothing more than an epoxy resin system. For cure to be effected and assured, molecular cross-linking must be effected. When formulation and application processes are properly managed, cure and adhesion characteristics will be as specified. This means cure not only is possible, but is measurable.

The primary issue, concerning your question about solder balls, is cure. If uncured, solder balls likely will form during most soldering operations as a result of outgassing. Also, entrapped process chemicals or moisture might contribute to the problem. Not only is this a problem as stated, but they become "entrapped" in the goo.

LPI usually is made up of two component parts as are most epoxies. Once properly formulated, application consists of "draping" or otherwise precisely coating (a specified amount or thickness) the PCB substrate surface. Cure is effected during a "drying" and/or baking operation. Part of the cure cycle is exothermic or endothermic, or a combination of both.

As the PI, in LPI, denotes photoimageable the coated substrate surface is imaged, polymerized, and stripped in specified patterns allowing precise openings to be formed around traces, pads, or other features as required. Essentially, this material is like any etch resist except it performs a different function.

Aside from the indicated process management requirements, problems may arise if the resist is not properly stripped. This can manifest itself as residue left on pads or in holes preventing solderability as another issue to consider.

As we continue living in a SMOBC/HASL (hot air solder leveled) world, mask adhesion is quite good as it is not present over melting metals - again when fully cured and cross linked. No solder mask works well over the best solder termination area finish - tin lead plating that hase been fused. It wrinkles and deteriorates readily thus reducing its effectiveness - and its ugly. This is why we have gone to alternative finishes.

Color is not an issue concerning solder mask performance unless, again, process management is poor as in formulation. X-hatching, or other shapes/openings are not an issue provided, again, processes are managed and the mask does what is supposed to do - prevent solder shorting between two conductors.

Just another note for one so serious about PCB's as you. The best solder mask is none. Pad only outer layered MLB's are the best way to go as no mask is required to prevent solder bridging.

And on it goes.

Earl Moon

reply »

Dave F

#13231

Re: Solder Resist Solder balls - 'N Surface Energy | 22 December, 1998

| Folk's, | I'm in need of some help / direction. I'm working on my thesis for a masters based on looking at the effect's of the solder resist promoting / reducing the amount of solder ball's during wave. I'm looking at thing's like colour, texture, composistion and mix(cross hatched / colour variations etc.) Does anyone have thought's/idea's/experience in this or could suggest some good reading material ? | Thanks in advance | John: Much of this street seems fairly well paved. Beyond that, wave soldered component insertions are declining at 13% annual rate.

Assuming the mask is cured, surface energy of mask finishes seems to be a driver of solder ball formation, both in wave and reflow soldering.

I don't know about reading material. Maybe some mask suppliers can help.

TTYL Dave F

reply »

Earl Moon

#13232

Re: Solder Resist Solder balls - 'N Surface Energy | 23 December, 1998

| | Folk's, | | I'm in need of some help / direction. I'm working on my thesis for a masters based on looking at the effect's of the solder resist promoting / reducing the amount of solder ball's during wave. I'm looking at thing's like colour, texture, composistion and mix(cross hatched / colour variations etc.) Does anyone have thought's/idea's/experience in this or could suggest some good reading material ? | | Thanks in advance | | | John: Much of this street seems fairly well paved. Beyond that, wave soldered component insertions are declining at 13% annual rate. | | Assuming the mask is cured, surface energy of mask finishes seems to be a driver of solder ball formation, both in wave and reflow soldering. | | I don't know about reading material. Maybe some mask suppliers can help. | | TTYL Dave F | Dave, How you do stir the holiday pot!

Surface energy issues mostly are resolved via the LPI solder mask composition, application, and post cleaning processes. Unlike DF types, epoxy based solder masks exhibit much less surface energy/tension, etc. - especially because of their porosity (except for issues concerning moisture absorbtion) and when properly "scuffed"/cleaned after processing.

You are exactly correct in providing this information. You also are correct concerning contacting each and every SM supplier (Ciba, Morton, Dynachem, or whoever they are now, etc.) to gain more in-depth information about how to more effectively manage the SM application process to ensure improved surface conditions to ensure minimal solder ball formation. Unfortunately, some suppliers are ill informed so fab suppliers, working closely with assemblers, must improve processes.

IPC-SM-840 provides many insights to solder mask design and acceptance requirements. It does not, as it never has, indicate how to better manage processes to improve specified effects. Also, IPC-TM-650 provides test methods, to some extent, but does not address issues concerning solder mask/solder ball formation - nor does any IPC/ANSI/J-STD. The experiment is ongoing but great strides have been made since the dry film days, but not so much over the days when solder mask was "silk screened" save photoimageability, accuracy, repeatability, and tolerances.

It is now up to our scholar, who probably already has an in depth understanding of PCB process management or lack of it, to provide us with more information gained both from experience and academia.

Earl Moon

reply »

John Watt

#13233

Re: Solder Resist Solder balls - 'N Surface Energy | 24 December, 1998

| | | Folk's, | | | I'm in need of some help / direction. I'm working on my thesis for a masters based on looking at the effect's of the solder resist promoting / reducing the amount of solder ball's during wave. I'm looking at thing's like colour, texture, composistion and mix(cross hatched / colour variations etc.) Does anyone have thought's/idea's/experience in this or could suggest some good reading material ? | | | Thanks in advance | | | | | John: Much of this street seems fairly well paved. Beyond that, wave soldered component insertions are declining at 13% annual rate. | | | | Assuming the mask is cured, surface energy of mask finishes seems to be a driver of solder ball formation, both in wave and reflow soldering. | | | | I don't know about reading material. Maybe some mask suppliers can help. | | | | TTYL Dave F | | | Dave, How you do stir the holiday pot! | | Surface energy issues mostly are resolved via the LPI solder mask composition, application, and post cleaning processes. Unlike DF types, epoxy based solder masks exhibit much less surface energy/tension, etc. - especially because of their porosity (except for issues concerning moisture absorbtion) and when properly "scuffed"/cleaned after processing. | | You are exactly correct in providing this information. You also are correct concerning contacting each and every SM supplier (Ciba, Morton, Dynachem, or whoever they are now, etc.) to gain more in-depth information about how to more effectively manage the SM application process to ensure improved surface conditions to ensure minimal solder ball formation. Unfortunately, some suppliers are ill informed so fab suppliers, working closely with assemblers, must improve processes. | | IPC-SM-840 provides many insights to solder mask design and acceptance requirements. It does not, as it never has, indicate how to better manage processes to improve specified effects. Also, IPC-TM-650 provides test methods, to some extent, but does not address issues concerning solder mask/solder ball formation - nor does any IPC/ANSI/J-STD. The experiment is ongoing but great strides have been made since the dry film days, but not so much over the days when solder mask was "silk screened" save photoimageability, accuracy, repeatability, and tolerances. | | It is now up to our scholar, who probably already has an in depth understanding of PCB process management or lack of it, to provide us with more information gained both from experience and academia. | | Earl Moon

Earl how can the sruface finish not have an effect ? All the boards that I'm working with are LPI. I've run a few trials so far using the following finishes : dark green gloss, light green gloss, light green matt & dark green matt. The different finishes were applied to the same PCB so remove any design influences in layout design, the same oven etc..the usual change 1 thing at a time stuff (the design of experiments bit comes later...) So far the best results have come from the light matt finish. I'm thinking that 1) there is less of a surface area on the matt finish for a solderball to 'cling' to, 2) the matt should allow the retension of more flux. As for the light green / dark green bit, I keep thinking back to my school physic's class with the light cube / dark cube, where one reflect's more heat so stay's cooler and thother get's hotter, if this is the case in the PCB then the lower heat retension of the PCB surface should reduce spattering of the solder ?

reply »

Earl Moon

#13234

Re: Solder Resist Solder balls - 'N Surface Energy | 24 December, 1998

| | | | Folk's, | | | | I'm in need of some help / direction. I'm working on my thesis for a masters based on looking at the effect's of the solder resist promoting / reducing the amount of solder ball's during wave. I'm looking at thing's like colour, texture, composistion and mix(cross hatched / colour variations etc.) Does anyone have thought's/idea's/experience in this or could suggest some good reading material ? | | | | Thanks in advance | | | | | | | John: Much of this street seems fairly well paved. Beyond that, wave soldered component insertions are declining at 13% annual rate. | | | | | | Assuming the mask is cured, surface energy of mask finishes seems to be a driver of solder ball formation, both in wave and reflow soldering. | | | | | | I don't know about reading material. Maybe some mask suppliers can help. | | | | | | TTYL Dave F | | | | | Dave, How you do stir the holiday pot! | | | | Surface energy issues mostly are resolved via the LPI solder mask composition, application, and post cleaning processes. Unlike DF types, epoxy based solder masks exhibit much less surface energy/tension, etc. - especially because of their porosity (except for issues concerning moisture absorbtion) and when properly "scuffed"/cleaned after processing. | | | | You are exactly correct in providing this information. You also are correct concerning contacting each and every SM supplier (Ciba, Morton, Dynachem, or whoever they are now, etc.) to gain more in-depth information about how to more effectively manage the SM application process to ensure improved surface conditions to ensure minimal solder ball formation. Unfortunately, some suppliers are ill informed so fab suppliers, working closely with assemblers, must improve processes. | | | | IPC-SM-840 provides many insights to solder mask design and acceptance requirements. It does not, as it never has, indicate how to better manage processes to improve specified effects. Also, IPC-TM-650 provides test methods, to some extent, but does not address issues concerning solder mask/solder ball formation - nor does any IPC/ANSI/J-STD. The experiment is ongoing but great strides have been made since the dry film days, but not so much over the days when solder mask was "silk screened" save photoimageability, accuracy, repeatability, and tolerances. | | | | It is now up to our scholar, who probably already has an in depth understanding of PCB process management or lack of it, to provide us with more information gained both from experience and academia. | | | | Earl Moon | | Earl how can the sruface finish not have an effect ? | All the boards that I'm working with are LPI. I've run a few trials so far using the following finishes : dark green gloss, light green gloss, light green matt & dark green matt. The different finishes were applied to the same PCB so remove any design influences in layout design, the same oven etc..the usual change 1 thing at a time stuff (the design of experiments bit comes later...) | So far the best results have come from the light matt finish. I'm thinking that 1) there is less of a surface area on the matt finish for a solderball to 'cling' to, 2) the matt should allow the retension of more flux. As for the light green / dark green bit, I keep thinking back to my school physic's class with the light cube / dark cube, where one reflect's more heat so stay's cooler and thother get's hotter, if this is the case in the PCB then the lower heat retension of the PCB surface should reduce spattering of the solder ? | As I too recall those physics lessons, radiant light energy, more efficiiently and effectively than other types, is what kept the reflective heat process going as with the black/white rotating motor. Is it as much the case with convective or conductive heat? After all, uncontrolled radiant heat is one of the biggest reasons we no longer use focused or direct IR and have gone almost completely to convection heating in reflow ovens.

As in my (and Dave's) above discussion, I agree the surface condition has much to do with solder ball formation as a function of surface energy/tension. The surface area function certainly a part of the equation concerning energy/tension. However, again, when properly processed, there should be little difference between the two types (matte or smooth) in terms of solder ball formation.

The primary reason there are matte finishes is to reduce glare and aid visual inspection after soldering. DF solder mask surfaces always had very high surface energy and could not be applied (laminate), yielding anything but a very smooth, shiny surface making solder ball formation prevalent as were complaints concerning inspection capabilities.

Part of another brief experiment might include discussions with your fab capability, provided it objectively understands solder mask process management beyond reading and applying their supplier instructions, to determine differences in how the two types, regardless of color, are composed, applied, and cleaned. This discussion also should include process consistency from one lot/batch or even board to another. This means composition, imaging, developing, stripping, and cleaning before introduction to your surface finish of choice (HASL, OSP, etc.).

Once you have determined solder mask process management is effective, determine how well solder surface finish process management has been effected. As you well know, this issue is paramount to solder wetting, solder ball formation, and solderability.

Another point you make is interesting concerning color or finish. Perhaps you or Dave could provide more insight. I wonder what differences in reflow solder profiles, as an example, might need such fine tuning relative to one solder mask finish/color to another. I would think, based on some experience and the fact radiant heat is not so much a factor anymore, once in liquidous, very little effect would be realized as equilibrium would have been reached almost equally (between solder mask finishes) at that point. I mean, how many degrees one way or the other is a real factor unless matte (as it has more surface area) heats and cools more quickly (how much more could it be at those temperatures?).

The same question might be posed for wave soldering. Here the affect is somewhat different as another major contributor to solder ball formation is the delta T from final preheat and boards contacting the wave at a much higher temperature than would be realized during reflow. Of course, poor fab process management would be magnified here as moisture or chemical entrapment as with all else discussed.

I guess I'm saying - to run an effective experiement, all that goes into it must be consistent, with exception of surface condition/color but they too must be consistent. They may be consistently good or bad, but consistent none the less. This starts with solder mask composition, application, imaging, developing, stripping, cleaning, and solder termination area coating selections, their process management and subsequent solder process management.

It is not a simple issue, but the experiment can be made relatively so. This really is an area one such as you can make a difference to us all with the same questions but without the will or means to determine cause and effect.

Respectfully,

Earl Moon

reply »

SMT Services - STI Electronics

Six Sigma Training