Surface Mount Solder Balling

We are getting solder balls after reflow of surface mount components. We are using a no-clean paste and a new jar seemed to help for a little bit. Any ideas? Thanks in advance

reply »

| We are getting solder balls after reflow of surface mount components. We are using a no-clean paste and a new jar seemed to help for a little bit. Any ideas? Thanks in advance Hello Ryan, There's a quite a few things that can cause solder balls during reflow, but there's specific things that need to be done to correct the problem depending on the type of solder balls that you're seeing...could you be a little more descriptive about what you're seeing, and tell us some more information about this particular board? For instance: -Where exactly are the solder balls occuring? All over? Only on resistors and capacitors? -How big are they? -What brand of paste are you using? -What are your stencil specifications? -What kind of oven are you using? -What is the reflow profile set-points? With that information I think you'll get some good input on what to do to fix it... C-ya, -Steve Gregory-

reply »

| Thanks Steve- Here is the whole story: The solder balls are appearing all over, similar to a splatter. They are relatively small. We are using H-Technologies S-HQ no-clean solder paste. The stencil is 6 mil and the boards are running in a ABW Systems TSC 1008 oven. The original profile was a smooth ramp up to peak, and the operator is trying a ramp to 150c-180c then holding the profile at a plateau to soak the board before peak. The current set points are: Top - 180 185 170 210 335 Bottom -= 180 185 170 205 330 Does this help?

| | We are getting solder balls after reflow of surface mount components. We are using a no-clean paste and a new jar seemed to help for a little bit. Any ideas? Thanks in advance | Hello Ryan, | There's a quite a few things that can cause solder balls during reflow, but there's specific things that need to be done to correct the problem depending on the type of solder balls that you're seeing...could you be a little more descriptive about what you're seeing, and tell us some more information about this particular board? For instance: | -Where exactly are the solder balls occuring? All over? Only on resistors and capacitors? | -How big are they? | -What brand of paste are you using? | -What are your stencil specifications? | -What kind of oven are you using? | -What is the reflow profile set-points? | With that information I think you'll get some good input on what to do to fix it... | C-ya, | -Steve Gregory-

reply »

| | Thanks Steve- | Here is the whole story: | | The solder balls are appearing all over, similar to a | splatter. They are relatively small. We are using H-Technologies | S-HQ no-clean solder paste. The stencil is 6 mil and the boards | are running in a ABW Systems TSC 1008 oven. The original profile was | a smooth ramp up to peak, and the operator is trying a ramp to 150c-180c | then holding the profile at a plateau to soak the board before peak. | The current set points are: | Top - 180 185 170 210 335 | Bottom -= 180 185 170 205 330 | Does this help? Ryan, Yep...it does. I've not used H-Technologies solder paste before, so I won't pretend that I know exactly what that paste likes. But it does seem to me that you're hitting the assembly with a lot of heat initially, which could explain your solder balls. I'm pretty sure you know about what's supposed to happen during the first ramp from ambient temperature to your preheat dwell temperature...that's where you're trying to drive off all the volatile components of the paste. If you have too fast of a initial ramp what happens is that the solvents will boil instead of smoothly evaporating like you want them to do...this will actually blow the paste off the pads...literally. When the paste goes liquidous, all of it will try as best it can to coalesce into one main fillet, but if it's too spread out, what you wind up with is the little balls you're seeing. I worked at a memory company where we had ABW 1008's, and I just so happen to have the set-points I used there for my no-clean profiles...(talk about being a pack-rat, I hardly ever throw anything away...GRIN) What I did was to have eight basic profiles; a single sided PCMCIA, low mass, medium mass, and high mass profile, and a double sided PCMCIA, low mass, medium mass, and high mass profile. We were using Qualitek no-clean paste. From the looks of the set-points you have now, the board must be a little on the dense side. So I'll give you what I used for my high mass boards. The high mass profile I used was for panelized SIMM modules that had DRAM pretty packed together on the board. The set points I used were: Profile name: High1 (single side, high mass, 1st side) Top: 180 160 160 220 330 Bottom: 180 160 160 220 330 Belt speed: 18 in/per/min Profile name: High2 (double side, high mass, 2nd side) Top: 180 160 180 230 330 Bottom: 180 160 160 210 280 Belt speed: 16 in/per/min I used these as starting point profiles, and then looked at the board when it came out of the oven. If everything looked half-way decent, I would go ahead and plot one so I could really dial it in. Hope this helps ya! -Steve Gregory-

reply »

| | | | Thanks Steve- | | Here is the whole story: | | | | The solder balls are appearing all over, similar to a | | splatter. They are relatively small. We are using H-Technologies | | S-HQ no-clean solder paste. The stencil is 6 mil and the boards | | are running in a ABW Systems TSC 1008 oven. The original profile was | | a smooth ramp up to peak, and the operator is trying a ramp to 150c-180c | | then holding the profile at a plateau to soak the board before peak. | | The current set points are: | | Top - 180 185 170 210 335 | | Bottom -= 180 185 170 205 330 | | Does this help? | Ryan, | Yep...it does. I've not used H-Technologies solder paste before, so I won't pretend that I know exactly what that paste likes. But it does seem to me that you're hitting the assembly with a lot of heat initially, which could explain your solder balls. | I'm pretty sure you know about what's supposed to happen during the first ramp from ambient temperature to your preheat dwell temperature...that's where you're trying to drive off all the volatile components of the paste. | If you have too fast of a initial ramp what happens is that the solvents will boil instead of smoothly evaporating like you want them to do...this will actually blow the paste off the pads...literally. | When the paste goes liquidous, all of it will try as best it can to coalesce into one main fillet, but if it's too spread out, what you wind up with is the little balls you're seeing. | I worked at a memory company where we had ABW 1008's, and I just so happen to have the set-points I used there for my no-clean profiles...(talk about being a pack-rat, I hardly ever throw anything away...GRIN) What I did was to have eight basic profiles; a single sided PCMCIA, low mass, medium mass, and high mass profile, and a double sided PCMCIA, low mass, medium mass, and high mass profile. We were using Qualitek no-clean paste. | From the looks of the set-points you have now, the board must be a little on the dense side. So I'll give you what I used for my high mass boards. The high mass profile I used was for panelized SIMM modules that had DRAM pretty packed together on the board. The set points I used were: | Profile name: High1 | (single side, high mass, 1st side) | Top: 180 160 160 220 330 | Bottom: 180 160 160 220 330 | Belt speed: 18 in/per/min | Profile name: High2 | (double side, high mass, 2nd side) | Top: 180 160 180 230 330 | Bottom: 180 160 160 210 280 | Belt speed: 16 in/per/min | I used these as starting point profiles, and then looked at the board when it came out of the oven. If everything looked half-way decent, I would go ahead and plot one so I could really dial it in. | Hope this helps ya! | -Steve Gregory- | Steve, I must applaud you, Justin, Dave, Chrys, and all the rest of the SMisfits participating in this forum. I believe, as you clearly show, there are no secrets. There are only lessons to be learned and sharred to advance our technology. Thanks, for all your PAST input, though I'm not going away (even though some would hope) unless you GIVE ME ANSWERS TO CCGA. I'll only look for a new line of work if you all don't talk to me about CCGA. Heck, Solcectron can't be leading Flextronics or that Sunnyvale startup just because all their top management came from, and still works closely with, IBM. Can they? Earl Moon

reply »

| | | | | | Thanks Steve- | | | Here is the whole story: | | | | | | The solder balls are appearing all over, similar to a | | | splatter. They are relatively small. We are using H-Technologies | | | S-HQ no-clean solder paste. The stencil is 6 mil and the boards | | | are running in a ABW Systems TSC 1008 oven. The original profile was | | | a smooth ramp up to peak, and the operator is trying a ramp to 150c-180c | | | then holding the profile at a plateau to soak the board before peak. | | | The current set points are: | | | Top - 180 185 170 210 335 | | | Bottom -= 180 185 170 205 330 | | | Does this help? | | Ryan, | | Yep...it does. I've not used H-Technologies solder paste before, so I won't pretend that I know exactly what that paste likes. But it does seem to me that you're hitting the assembly with a lot of heat initially, which could explain your solder balls. | | I'm pretty sure you know about what's supposed to happen during the first ramp from ambient temperature to your preheat dwell temperature...that's where you're trying to drive off all the volatile components of the paste. | | If you have too fast of a initial ramp what happens is that the solvents will boil instead of smoothly evaporating like you want them to do...this will actually blow the paste off the pads...literally. | | When the paste goes liquidous, all of it will try as best it can to coalesce into one main fillet, but if it's too spread out, what you wind up with is the little balls you're seeing. | | I worked at a memory company where we had ABW 1008's, and I just so happen to have the set-points I used there for my no-clean profiles...(talk about being a pack-rat, I hardly ever throw anything away...GRIN) What I did was to have eight basic profiles; a single sided PCMCIA, low mass, medium mass, and high mass profile, and a double sided PCMCIA, low mass, medium mass, and high mass profile. We were using Qualitek no-clean paste. | | From the looks of the set-points you have now, the board must be a little on the dense side. So I'll give you what I used for my high mass boards. The high mass profile I used was for panelized SIMM modules that had DRAM pretty packed together on the board. The set points I used were: | | Profile name: High1 | | (single side, high mass, 1st side) | | Top: 180 160 160 220 330 | | Bottom: 180 160 160 220 330 | | Belt speed: 18 in/per/min | | Profile name: High2 | | (double side, high mass, 2nd side) | | Top: 180 160 180 230 330 | | Bottom: 180 160 160 210 280 | | Belt speed: 16 in/per/min | | I used these as starting point profiles, and then looked at the board when it came out of the oven. If everything looked half-way decent, I would go ahead and plot one so I could really dial it in. | | Hope this helps ya! | | -Steve Gregory- | | Steve, | I must applaud you, Justin, Dave, Chrys, and all the rest of the SMisfits participating in this forum. I believe, as you clearly show, there are no secrets. There are only lessons to be learned and sharred to advance our technology. | Thanks, for all your PAST input, though I'm not going away (even though some would hope) unless you GIVE ME ANSWERS TO CCGA. I'll only look for a new line of work if you all don't talk to me about CCGA. | Heck, Solcectron can't be leading Flextronics or that Sunnyvale startup just because all their top management came from, and still works closely with, IBM. Can they? | Earl Moon We have experienced a solderball problem in the past, with several different solder paste manufacturers. No matter what paste, we still got solder balls. We reduced the stencil aperture, 20%, and the problem went away. Now we can run any paste without problems.

reply »

Thanks everybody for the help! We are still working on the bottom side, but the top side solder balling seems to have gone away for now. We flattened out the middle soak zone of the profile, so that rather than an even ramp up, the board heats quickly and then soaks at about 140c-160c for about two minutes. We haven't run the bottom side yet, but will do that soon. The wave residues are slowly diminishing with a change to water-based flux and adjustment to the flux volume applied. Pallets seem to make the residue more difficult to get rid of. Is this a common problem? Thanks again.

Ryan Jennens Phoenix Engineering Design

reply »

If you mean solder beads at the side of chip components then download the FAQs and the process guide on solder beads which may be of interest. The documents are on my web site www.bobwillis.co.uk

| We are getting solder balls after reflow of surface mount components. We are using a no-clean paste and a new jar seemed to help for a little bit. Any ideas? Thanks in advance

reply »