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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Help! 2-layered Thru-Hole PCB Solder Mask/Soldering Problem

Dreamsniper

#9238

Help! 2-layered Thru-Hole PCB Solder Mask/Soldering Problem | 29 September, 1999

Hi everyone,

Greetings!

I�ve got a 2-layered through-hole PCB�s having the following problems:

Solder Mask peels after wave soldering at random locations and only at the bootom side of the PCB Insufficient Solder and some solder never make it to the top-side of the pcb�s. Poor Wetting Blow-hole Some Joints have voids No Solder

These problems happen only to one of our pcb vendor and the rests are okay. The solder mask type is Liquid Photoimageable and I think that the Brand is Finedel DSR-2200 C-7BSX (if I�m not wrong). Our process requires to bake the PCB�s for 4 hrs at 80 deg. C and after that operator applies spot-on on some of the holes that do not need to be soldered and then returned back into the oven for another 1 HR baking to dry the spot on.

Had sent the pcb to one of the local manufacturer near our plant for analysis and the report says:

Slightly under plate. Weak point were detected in some holes (~15 um copper) possibly due to smear/debris from drilling. Two-part epoxy screening type solder mask resulting exposed track and uneven coating.

Does anyone have encountered the same problem like this? And can anyone help me understand the analysis report more clearly. Any counter measusre you recommend regarding wave solder process? ALL INPUTS ARE WELCOME !

Thank you very much, Dreamsniper

reply »

Brian

#9239

Re: Help! 2-layered Thru-Hole PCB Solder Mask/Soldering Problem | 30 September, 1999

| Hi everyone, | | Greetings! | | I�ve got a 2-layered through-hole PCB�s having the following problems: | | Solder Mask peels after wave soldering at random locations and only at the bootom side of the PCB | Insufficient Solder and some solder never make it to the top-side of the pcb�s. | Poor Wetting | Blow-hole | Some Joints have voids | No Solder | | These problems happen only to one of our pcb vendor and the rests are okay. The solder mask type is Liquid Photoimageable and I think that the Brand is Finedel DSR-2200 C-7BSX (if I�m not wrong). Our process requires to bake the PCB�s for 4 hrs at 80 deg. C and after that operator applies spot-on on some of the holes that do not need to be soldered and then returned back into the oven for another 1 HR baking to dry the spot on. | | Had sent the pcb to one of the local manufacturer near our plant for analysis and the report says: | | Slightly under plate. Weak point were detected in some holes (~15 um copper) possibly due to smear/debris from drilling. Two-part epoxy screening type solder mask resulting exposed track and uneven coating. | | Does anyone have encountered the same problem like this? And can anyone help me understand the analysis report more clearly. Any counter measusre you recommend regarding wave solder process? ALL INPUTS ARE WELCOME ! | | Thank you very much, | Dreamsniper | | Hi!

1st comment is negative. Drop your vendor from the qualified list, illico presto. He is probably your cheapest-to-purchase one, beloved of purchasing departments, but the most expensive one overall. That he makes one error on one occasion is bad enough: that he blows it on plating thickness, drilling quality, solder mask application and solderability is REALLY unforgiveable.

As for action to take, your cheapest and quickest thing to do is to have the batch remade by somebody professional who knows what he is doing and find a trash can for what you already have. There is nothing you can do to save what you have. If you try to use them, expect returns of the finished product because the quality will NEVER be up to speed.

I would normally expect about 25 �m of copper in through-holes which should be clean, without debris. It sounds like the solder mask is probably applied over a contaminated surface and is improperly cured. No way can you decontaminate what is under it.

And don't pay the guy's invoice!

Brian

reply »

Scott Cook

#9240

Re: Help! 2-layered Thru-Hole PCB Solder Mask/Soldering Problem | 30 September, 1999

| Hi everyone, | | Greetings! | | I�ve got a 2-layered through-hole PCB�s having the following problems: | | Solder Mask peels after wave soldering at random locations and only at the bootom side of the PCB | Insufficient Solder and some solder never make it to the top-side of the pcb�s. | Poor Wetting | Blow-hole | Some Joints have voids | No Solder | | These problems happen only to one of our pcb vendor and the rests are okay. The solder mask type is Liquid Photoimageable and I think that the Brand is Finedel DSR-2200 C-7BSX (if I�m not wrong). Our process requires to bake the PCB�s for 4 hrs at 80 deg. C and after that operator applies spot-on on some of the holes that do not need to be soldered and then returned back into the oven for another 1 HR baking to dry the spot on. | | Had sent the pcb to one of the local manufacturer near our plant for analysis and the report says: | | Slightly under plate. Weak point were detected in some holes (~15 um copper) possibly due to smear/debris from drilling. Two-part epoxy screening type solder mask resulting exposed track and uneven coating. | | Does anyone have encountered the same problem like this? And can anyone help me understand the analysis report more clearly. Any counter measusre you recommend regarding wave solder process? ALL INPUTS ARE WELCOME ! | | Thank you very much, | Dreamsniper

I am a bit confused. You say the mask is LPI, yet the report mentions two part screened on mask. This doesn't make sense to me. But the exposed and uneven coating is a scare, and may very well indicate poor application / process control at the board house. That could mean improper curing, as well. I'd do an immediate audit of this house. Look at their cleanliness, their tanks, their process control mechanisms. A good road trip, you will do yourself a big favor by learning about the manufacture of bare boards. It's an art, combined with science. Real voodoo. You think assembling the board is tough? Go see how the bare boards are done! A good multilayer house is a fascinating tour / study in process controls.

In addition, what is the bare board material? Has this board house been audited / certified by you? If not, why not? The pain will continue if you let purchasing lead the way in PCB's. Their agenda is much different than yours. They are not driven to view the TOTAL COST of Ownership of bare boards; so they go for the cheap bid.

Although Brian makes some good points, my approach would be a bit different than Brian's (sorry here, Brian). I'd give the existing vendor another chance; mainly because he's got the artwork, and can turn around product fastest. Call him. Ask him to come immediately to your facility, and show him the issues and the report. If he's credible, he'll turn around replacement boards for you in a flash.

Scott

reply »

#9241

Re: Help! 2-layered Thru-Hole PCB Solder Mask/Soldering Problem | 30 September, 1999

| Hi everyone, | | Greetings! | | I�ve got a 2-layered through-hole PCB�s having the following problems: | | Solder Mask peels after wave soldering at random locations and only at the bootom side of the PCB | Insufficient Solder and some solder never make it to the top-side of the pcb�s. | Poor Wetting | Blow-hole | Some Joints have voids | No Solder | | These problems happen only to one of our pcb vendor and the rests are okay. The solder mask type is Liquid Photoimageable and I think that the Brand is Finedel DSR-2200 C-7BSX (if I�m not wrong). Our process requires to bake the PCB�s for 4 hrs at 80 deg. C and after that operator applies spot-on on some of the holes that do not need to be soldered and then returned back into the oven for another 1 HR baking to dry the spot on. | | Had sent the pcb to one of the local manufacturer near our plant for analysis and the report says: | | Slightly under plate. Weak point were detected in some holes (~15 um copper) possibly due to smear/debris from drilling. Two-part epoxy screening type solder mask resulting exposed track and uneven coating. | | Does anyone have encountered the same problem like this? And can anyone help me understand the analysis report more clearly. Any counter measusre you recommend regarding wave solder process? ALL INPUTS ARE WELCOME ! | | Thank you very much, | Dreamsniper | | Armin: I�d like more information on:

� Extent of the problem in terms of new/old product, new/old problem, and all/some boards in lot. � Has the fabricator been contacted, provided samples of failed items, and provided a response? � Board fabrication: Clarification of the solder mask type and finish, SMOBC / tin lead fusing under the solder mask, differences between this board and others that don�t exhibit these problems, and date of fabrication. � Pre-Assembly Handling: What�s the acceptance solderability testing of the boards and storage conditions and time prior to assembly. � Wave soldering process set-up and flux type and application method.

That being said, let�s take the angle that the fabricator is not at fault. Although, the thin plating thickness and drilling debris in through holes makes me wonder why I�m taking this tact. But then again, I understand a competitor analyzed the boards and "who knows the evil that ..." And the "standard process" of baking the boards before soldering indicates a broader barrel plating problem than the one with this board. While I understand that the post temporary solder mask baking cures the mask, I would argue that a thinner mask application would eliminate this "necessity."

What the heck!!! Let�s push on and consider alternatives, before we run these hard folks working otta town. But, before I can do that, I�ve got this baking thing stuck in my throat and I�ve got to get it out!!!! Baking before temporary mask does nothing positive. If the purpose of baking is to:

� Prepare the board to take the mask: it is way too long, you shouldn�t be doing it all, and better mask suppliers don�t require pre-baking. Live large, spend a penny replacing that cheap mask, and save a buck in total cost. � Remove moisture prior to wave soldering: you shouldn�t be doing it at all, the baking at 80�C for 4 hours doesn�t do dip to remove moisture. Weigh a stack of boards before and after baking to prove its benefits. And more telling, indicates (1) that you have barrel plating problems with multiple suppliers and (2) there is room to improve your inventory control and storage methods.

Now, let�s talk about the problems and something other than the fabricator as the potential problem, if we can:

1. "Solder mask peels after wave soldering at random locations and only at the bottom side of the PCB." Obviously the heat of the wave solder is causing this peeling. If your board has fused solder over the copper, reflowing this solder can cause the mask over the traces to wrinkle. If the solder mask is not LPI and is an epoxy screened mask, it could be becoming brittle and flaking/peeling, if it was not cured just right. 2. "Insufficient solder and some solder never make it to the top-side of the PCB�s." First, complete barrel fill is not required by 610, although I recognize that your local requirements may superceed 610. Second, I�ve seen situations where insufficient solder and poor barrel flow were caused by poor wave soldering. In fact in this situation, I�d look to poor wave set-up and control before I looked at the board. 3. "Poor wetting." I wonder if this could be caused by (1) board storage in a humid environment, over long (GT 6 months) periods, (2) corrosion forming on pads during baking, and / or (3) poor wave solder set-up and control? 4. "Blow-hole." Owww, most often caused by poor barrel plating on moisture laden boards. We almost made it!!!! Yes, inadequate storage is a factor, but plating thickness is the main issue. 5. "Some joints have voids." See 3 above. 6. "No solder." I�ve seen situations where no solder was caused by poor wave soldering. I�d think about the same things as "poor wetting," but on the other hand, I�d be wondering what�s on the pads that�s preventing them from taking on solder.

Now, on to the failure analysis:

1. I like to see 1 mil of plating thickness, which corresponds with Brian�s 25um. So, in the thin areas your supplier is short about 40% of the necessary plating thickness. Plating is usually the thinnest at the edge of the barrel. 2. Definitions: (1) Smear is a fused epoxy deposit left on copper from drilling heat. Proper drill control during board fabrication prevents smearing. Desmearing is the board fabrication process that removes smearing. (2) Drilling debris could be fiberglass threads that weren�t cut during drilling or any number of types of crud that drilling produces that the fabricator did not clean before plating. Drill sharpening and cleaning remove drilling debris. 3. Smear and drilling debris prevent proper plating and fusing. 4. As Scott mentioned, LPI is not a "two-part epoxy screening type solder mask." So, someone is confused. It�s possible that you didn�t receive what you asked for. 5. Boards with wet (screened) solder masks are more likely to have "exposed track and uneven coating" than boards with film solder masks.

Ta

Dave F

reply »

Dreamsniper

#9242

Re: Help! 2-layered Thru-Hole PCB Solder Mask/Soldering Problem | 30 September, 1999

| | Hi everyone, | | | | Greetings! | | | | I�ve got a 2-layered through-hole PCB�s having the following problems: | | | | Solder Mask peels after wave soldering at random locations and only at the bootom side of the PCB | | Insufficient Solder and some solder never make it to the top-side of the pcb�s. | | Poor Wetting | | Blow-hole | | Some Joints have voids | | No Solder | | | | These problems happen only to one of our pcb vendor and the rests are okay. The solder mask type is Liquid Photoimageable and I think that the Brand is Finedel DSR-2200 C-7BSX (if I�m not wrong). Our process requires to bake the PCB�s for 4 hrs at 80 deg. C and after that operator applies spot-on on some of the holes that do not need to be soldered and then returned back into the oven for another 1 HR baking to dry the spot on. | | | | Had sent the pcb to one of the local manufacturer near our plant for analysis and the report says: | | | | Slightly under plate. Weak point were detected in some holes (~15 um copper) possibly due to smear/debris from drilling. Two-part epoxy screening type solder mask resulting exposed track and uneven coating. | | | | Does anyone have encountered the same problem like this? And can anyone help me understand the analysis report more clearly. Any counter measusre you recommend regarding wave solder process? ALL INPUTS ARE WELCOME ! | | | | Thank you very much, | | Dreamsniper | | | | | Armin: I�d like more information on: | | � Extent of the problem in terms of new/old product, new/old problem, and all/some boards in lot. | � Has the fabricator been contacted, provided samples of failed items, and provided a response? | � Board fabrication: Clarification of the solder mask type and finish, SMOBC / tin lead fusing under the solder mask, differences between this board and others that don�t exhibit these problems, and date of fabrication. | � Pre-Assembly Handling: What�s the acceptance solderability testing of the boards and storage conditions and time prior to assembly. | � Wave soldering process set-up and flux type and application method. | | That being said, let�s take the angle that the fabricator is not at fault. Although, the thin plating thickness and drilling debris in through holes makes me wonder why I�m taking this tact. But then again, I understand a competitor analyzed the boards and "who knows the evil that ..." And the "standard process" of baking the boards before soldering indicates a broader barrel plating problem than the one with this board. While I understand that the post temporary solder mask baking cures the mask, I would argue that a thinner mask application would eliminate this "necessity." | | What the heck!!! Let�s push on and consider alternatives, before we run these hard folks working otta town. But, before I can do that, I�ve got this baking thing stuck in my throat and I�ve got to get it out!!!! Baking before temporary mask does nothing positive. If the purpose of baking is to: | | � Prepare the board to take the mask: it is way too long, you shouldn�t be doing it all, and better mask suppliers don�t require pre-baking. Live large, spend a penny replacing that cheap mask, and save a buck in total cost. | � Remove moisture prior to wave soldering: you shouldn�t be doing it at all, the baking at 80�C for 4 hours doesn�t do dip to remove moisture. Weigh a stack of boards before and after baking to prove its benefits. And more telling, indicates (1) that you have barrel plating problems with multiple suppliers and (2) there is room to improve your inventory control and storage methods. | | Now, let�s talk about the problems and something other than the fabricator as the potential problem, if we can: | | 1. "Solder mask peels after wave soldering at random locations and only at the bottom side of the PCB." Obviously the heat of the wave solder is causing this peeling. If your board has fused solder over the copper, reflowing this solder can cause the mask over the traces to wrinkle. If the solder mask is not LPI and is an epoxy screened mask, it could be becoming brittle and flaking/peeling, if it was not cured just right. | 2. "Insufficient solder and some solder never make it to the top-side of the PCB�s." First, complete barrel fill is not required by 610, although I recognize that your local requirements may superceed 610. Second, I�ve seen situations where insufficient solder and poor barrel flow were caused by poor wave soldering. In fact in this situation, I�d look to poor wave set-up and control before I looked at the board. | 3. "Poor wetting." I wonder if this could be caused by (1) board storage in a humid environment, over long (GT 6 months) periods, (2) corrosion forming on pads during baking, and / or (3) poor wave solder set-up and control? | 4. "Blow-hole." Owww, most often caused by poor barrel plating on moisture laden boards. We almost made it!!!! Yes, inadequate storage is a factor, but plating thickness is the main issue. | 5. "Some joints have voids." See 3 above. | 6. "No solder." I�ve seen situations where no solder was caused by poor wave soldering. I�d think about the same things as "poor wetting," but on the other hand, I�d be wondering what�s on the pads that�s preventing them from taking on solder. | | Now, on to the failure analysis: | | 1. I like to see 1 mil of plating thickness, which corresponds with Brian�s 25um. So, in the thin areas your supplier is short about 40% of the necessary plating thickness. Plating is usually the thinnest at the edge of the barrel. | 2. Definitions: (1) Smear is a fused epoxy deposit left on copper from drilling heat. Proper drill control during board fabrication prevents smearing. Desmearing is the board fabrication process that removes smearing. (2) Drilling debris could be fiberglass threads that weren�t cut during drilling or any number of types of crud that drilling produces that the fabricator did not clean before plating. Drill sharpening and cleaning remove drilling debris. | 3. Smear and drilling debris prevent proper plating and fusing. | 4. As Scott mentioned, LPI is not a "two-part epoxy screening type solder mask." So, someone is confused. It�s possible that you didn�t receive what you asked for. | 5. Boards with wet (screened) solder masks are more likely to have "exposed track and uneven coating" than boards with film solder masks. | | Ta | | Dave F | Hi Scott, Brian and Dave F,

I'll get back to this issue by monday as i will be evaluating 100 pcs. of bd. from the supplier of this problematic pcb. Then I will provide you better views on the problem including types of materials and processes. Meantime thank you very much for all this good and helpful advise you provided me.

'til next week...

armin

reply »

SMT Equipment

SMT in-printer dispensing