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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Re: Solder Beading Solder Balling

Tom B.

#10289

Solder Beading Solder Balling | 4 August, 1999

Hello Netters,

Can anyone provide me with info or where to get info on acceptability of solder beads and solder balls?

Thanks for any help

Tom B.

reply »

#10290

Re: Solder Beading Solder Balling | 4 August, 1999

| Hello Netters, | | Can anyone provide me with info or where to get info on acceptability of solder beads and solder balls? | | Thanks for any help | | Tom B. | | Tom, The IPC sets the standards for acceptability. But in a nutshell if the ball is not perminantly attached it's not acceptable. Of course there is "customer criteria" as well but we always go back to IPC.

reply »

#10291

Re: Solder Beading Solder Balling | 5 August, 1999

| | Hello Netters, | | | | Can anyone provide me with info or where to get info on acceptability of solder beads and solder balls? | | | | Thanks for any help | | | | Tom B. | | | | Tom, | The IPC sets the standards for acceptability. But in a nutshell if the ball is not perminantly attached it's not acceptable. Of course there is "customer criteria" as well but we always go back to IPC. | | Tom, even if you know the standards it�s hard work to control. Best look at possible causes and try to eliminate them, make them the exception.

Wolfgang

reply »

#10292

Re: Solder Beading Solder Balling | 5 August, 1999

| | |

Hello, If you are using no-clean and they are entrapped in the flux it is also IPC accepted but i havent found a customer yet who agrees with that. Im sure you have checked your profile. I had the same problem with chip components but after changing aperture shapes to trapezoids it disappeared.

Hello Netters, | | | | | | Can anyone provide me with info or where to get info on acceptability of solder beads and solder balls? | | | | | | Thanks for any help | | | | | | Tom B. | | | | | | Tom, | | The IPC sets the standards for acceptability. But in a nutshell if the ball is not perminantly attached it's not acceptable. Of course there is "customer criteria" as well but we always go back to IPC. | | | | | Tom, | even if you know the standards it�s hard work to control. | Best look at possible causes and try to eliminate them, make them the exception. | | Wolfgang |

reply »

#10293

Re: Solder Beading Solder Balling | 5 August, 1999

| Hello Netters, | | Can anyone provide me with info or where to get info on acceptability of solder beads and solder balls? | | Thanks for any help | | Tom B. | Hi Tom Wolfgang is right. We think you should correct your process to reduce ball formation to the level that you don't have to worry about acceptance. That said, tell us whether it's a problem with reflow or wave. |

reply »

#10294

Re: Solder Beading Solder Balling | 5 August, 1999

| | | | | | Hello, | If you are using no-clean and they are entrapped in the flux it is also IPC accepted but i havent found a customer yet who agrees with that. Im sure you have checked your profile. I had the same problem with chip components but after changing aperture shapes to trapezoids it disappeared.

Hello Wayne, interesting point, I heard of it and there were some evaluations on this subject here in germany with the result that it helps prevent solderballs. We actually never tried it because our designfolks didn�t want to put to much work in it, we just adjusted our square pads, stencil data and the printing rules to control the amount of paste and it also did the job. Another cause of balls I noticed in the wavesoldering process with noclean flux and nitrogen at the site of one of our subcontractors, they had to add an extra rotating brushcleaning system right after the soldermachine but it wasn�t much help. I think changing the flux and the solderresist were more successful.

Wolfgang

reply »

#10295

Re: Solder Beading Solder Balling | 5 August, 1999

| | | | | | | | | Hello, | | If you are using no-clean and they are entrapped in the flux it is also IPC accepted but i havent found a customer yet who agrees with that. Im sure you have checked your profile. I had the same problem with chip components but after changing aperture shapes to trapezoids it disappeared. | | Hello Wayne, | interesting point, I heard of it and there were some evaluations on this subject here in germany with the result that it helps prevent solderballs. | We actually never tried it because our designfolks didn�t want to put to much work in it, we just adjusted our square pads, stencil data and the printing rules to control the amount of paste and it also did the job. | Another cause of balls I noticed in the wavesoldering process with noclean flux and nitrogen at the site of one of our subcontractors, they had to add an extra rotating brushcleaning system right after the soldermachine but it wasn�t much help. I think changing the flux and the solderresist were more successful. | | Wolfgang

We had a lot of solder balls when we first started SMT. Turned out these were "squeeze balls" caused when the placement pressure squeezed paste toward the center of chip components where it was over the mask rather than the pad. As the paste melted and the flux volitalized it blew this displaced paste out the side as balls along the side of the component.| We changed the aperture shape for chip components to a baseball "home plate" shape with the points toward the center. The reduced paste volume at the center eliminated the formation of balls. Many CAD systems and most stencil manufacturers can make this modification with little effort. John Thorup

reply »

#10296

Re: Solder Beading Solder Balling | 5 August, 1999

| | | | | | | | | | | | Hello, | | | If you are using no-clean and they are entrapped in the flux it is also IPC accepted but i havent found a customer yet who agrees with that. Im sure you have checked your profile. I had the same problem with chip components but after changing aperture shapes to trapezoids it disappeared. | | | | Hello Wayne, | | interesting point, I heard of it and there were some evaluations on this subject here in germany with the result that it helps prevent solderballs. | | We actually never tried it because our designfolks didn�t want to put to much work in it, we just adjusted our square pads, stencil data and the printing rules to control the amount of paste and it also did the job. | | Another cause of balls I noticed in the wavesoldering process with noclean flux and nitrogen at the site of one of our subcontractors, they had to add an extra rotating brushcleaning system right after the soldermachine but it wasn�t much help. I think changing the flux and the solderresist were more successful. | | | | Wolfgang | | We had a lot of solder balls when we first started SMT. Turned out these were "squeeze balls" caused when the placement pressure squeezed paste toward the center of chip components where it was over the mask rather than the pad. As the paste melted and the flux volitalized it blew this displaced paste out the side as balls along the side of the component.| We changed the aperture shape for chip components to a baseball "home plate" shape with the points toward the center. The reduced paste volume at the center eliminated the formation of balls. Many CAD systems and most stencil manufacturers can make this modification with little effort. | John Thorup | | If I understand you right, you didn�t change the pad itself. In that study I remember they changed the padshape itself but applying less solder at the innerside might work as well. We print with 15% reduction and 0.15mm thick stencil. Today we almost don�t have any problems with that effect you described so perfectly, when it happens it�s always a new part, a new engineer and a failure of our design controlsystem (the right person on vacation). As for our subcontractors they don�t have any problems with our SMD stuff but still some problems in wavesoldering and it�s hard to get them on right track for they refer to the IPC and we should proof the amount and size of the balls doesn�t meet the spec. Lausy job!

Best wishes

Wolfgang

reply »

#10297

Re: Solder Beading Solder Balling | 5 August, 1999

| | | | | | | | | | | | | | | Hello, | | | | If you are using no-clean and they are entrapped in the flux it is also IPC accepted but i havent found a customer yet who agrees with that. Im sure you have checked your profile. I had the same problem with chip components but after changing aperture shapes to trapezoids it disappeared. | | | | | | Hello Wayne, | | | interesting point, I heard of it and there were some evaluations on this subject here in germany with the result that it helps prevent solderballs. | | | We actually never tried it because our designfolks didn�t want to put to much work in it, we just adjusted our square pads, stencil data and the printing rules to control the amount of paste and it also did the job. | | | Another cause of balls I noticed in the wavesoldering process with noclean flux and nitrogen at the site of one of our subcontractors, they had to add an extra rotating brushcleaning system right after the soldermachine but it wasn�t much help. I think changing the flux and the solderresist were more successful. | | | | | | Wolfgang | | | | We had a lot of solder balls when we first started SMT. Turned out these were "squeeze balls" caused when the placement pressure squeezed paste toward the center of chip components where it was over the mask rather than the pad. As the paste melted and the flux volitalized it blew this displaced paste out the side as balls along the side of the component.| We changed the aperture shape for chip components to a baseball "home plate" shape with the points toward the center. The reduced paste volume at the center eliminated the formation of balls. Many CAD systems and most stencil manufacturers can make this modification with little effort. | | John Thorup | | | | | If I understand you right, you didn�t change the pad itself. | In that study I remember they changed the padshape itself but applying less solder at the innerside might work as well. | We print with 15% reduction and 0.15mm thick stencil. | Today we almost don�t have any problems with that effect you described so perfectly, when it happens it�s always a new part, a new engineer and a failure of our design controlsystem (the right person on vacation). | As for our subcontractors they don�t have any problems with our SMD stuff but still some problems in wavesoldering and it�s hard to get them on right track for they refer to the IPC and we should proof the amount and size of the balls doesn�t meet the spec. Lausy job! | | Best wishes | | Wolfgang Correct - the pad shape was not changed - only the shape of the paste deposition. As far as your subcontractors - it's what you specify that you want - not what the IPC says is acceptable. If the IPC minimum acceptable is not sufficient for your needs you have every right to specify more. Just be sure that you need more and that it is buildable or you will be throwing your money away. John Thorup|

reply »

#10298

Re: Solder Beading Solder Balling | 6 August, 1999

| | | | | | | | | | | | | | | | | | Hello, | | | | | If you are using no-clean and they are entrapped in the flux it is also IPC accepted but i havent found a customer yet who agrees with that. Im sure you have checked your profile. I had the same problem with chip components but after changing aperture shapes to trapezoids it disappeared. | | | | | | | | Hello Wayne, | | | | interesting point, I heard of it and there were some evaluations on this subject here in germany with the result that it helps prevent solderballs. | | | | We actually never tried it because our designfolks didn�t want to put to much work in it, we just adjusted our square pads, stencil data and the printing rules to control the amount of paste and it also did the job. | | | | Another cause of balls I noticed in the wavesoldering process with noclean flux and nitrogen at the site of one of our subcontractors, they had to add an extra rotating brushcleaning system right after the soldermachine but it wasn�t much help. I think changing the flux and the solderresist were more successful. | | | | | | | | Wolfgang | | | | | | We had a lot of solder balls when we first started SMT. Turned out these were "squeeze balls" caused when the placement pressure squeezed paste toward the center of chip components where it was over the mask rather than the pad. As the paste melted and the flux volitalized it blew this displaced paste out the side as balls along the side of the component.| We changed the aperture shape for chip components to a baseball "home plate" shape with the points toward the center. The reduced paste volume at the center eliminated the formation of balls. Many CAD systems and most stencil manufacturers can make this modification with little effort. | | | John Thorup | | | | | | | | If I understand you right, you didn�t change the pad itself. | | In that study I remember they changed the padshape itself but applying less solder at the innerside might work as well. | | We print with 15% reduction and 0.15mm thick stencil. | | Today we almost don�t have any problems with that effect you described so perfectly, when it happens it�s always a new part, a new engineer and a failure of our design controlsystem (the right person on vacation). | | As for our subcontractors they don�t have any problems with our SMD stuff but still some problems in wavesoldering and it�s hard to get them on right track for they refer to the IPC and we should proof the amount and size of the balls doesn�t meet the spec. Lausy job! | | | | Best wishes | | | | Wolfgang | Correct - the pad shape was not changed - only the shape of the paste deposition. As far as your subcontractors - it's what you specify that you want - not what the IPC says is acceptable. If the IPC minimum acceptable is not sufficient for your needs you have every right to specify more. Just be sure that you need more and that it is buildable or you will be throwing your money away. | John Thorup| | | What I meant is that if there are some balls detected on a sample you have to decide wether it�s in the spec or not, means measure the size and count the numbers and then inspect all boards or reject and have them reworked at subcontractor site. That normally goes along with your measurement isn�t correct or it was just unluckily the sample or ... I don�t like this kind of stuff, takes time, money and nerves so getting into the brushing and remove the balls in less the time and effort is what I prefer and it saves money too. Although there�s still the feedback to the sub. to improve things and it�s getting better every day.

Wolfgang

reply »

JohnW

#10299

Re: Solder Beading Solder Balling | 7 August, 1999

| | | | | | | | | | | | Hello, | | | If you are using no-clean and they are entrapped in the flux it is also IPC accepted but i havent found a customer yet who agrees with that. Im sure you have checked your profile. I had the same problem with chip components but after changing aperture shapes to trapezoids it disappeared. | | | | Hello Wayne, | | interesting point, I heard of it and there were some evaluations on this subject here in germany with the result that it helps prevent solderballs. | | We actually never tried it because our designfolks didn�t want to put to much work in it, we just adjusted our square pads, stencil data and the printing rules to control the amount of paste and it also did the job. | | Another cause of balls I noticed in the wavesoldering process with noclean flux and nitrogen at the site of one of our subcontractors, they had to add an extra rotating brushcleaning system right after the soldermachine but it wasn�t much help. I think changing the flux and the solderresist were more successful. | | | | Wolfgang | | We had a lot of solder balls when we first started SMT. Turned out these were "squeeze balls" caused when the placement pressure squeezed paste toward the center of chip components where it was over the mask rather than the pad. As the paste melted and the flux volitalized it blew this displaced paste out the side as balls along the side of the component.| We changed the aperture shape for chip components to a baseball "home plate" shape with the points toward the center. The reduced paste volume at the center eliminated the formation of balls. Many CAD systems and most stencil manufacturers can make this modification with little effort. | John Thorup | | My wo pennies worth.....The common theme with the home plate / trapizoid design's is the redusction of the paste available, I run a std 4 thou reduction of chip component's and 2though on fine pitch to 20 thou on rectangular appetures and have no problems. Mid ship solderball's are as John say's caused by excess paste being 'squeezed' under the component at placement and colelesting in the middle of the component, again it can be removed by better stencil design. The other thing that causes the solderballing is poor profiling / ramp rates, excess moisture in material ( including the solderpaste) so there's the dicipline eliment there too and it's so easy to fix when you find the problem...past experience talking there! a bad batch of paste can have similar effect's. Wave soldering under N2 does produce more solderball's than in air but then you got dross to worry about so it's swing's and roundabouts, as someone else pointed out flux and resist have a big influence but then if your prheat's are wrong you'll get them, or again excess moisture in the boards. IPC std's and customer standard's will give target's like 5 ball's of .25mm dia per square inch or what ever but it's got to be zero or as close to zero as we can get.

And if you think it's bad now...wiat till you try it with lead free solder!!!!!!! ooooo deep joy

JohnW

reply »

Thermal warpage test services

Reflow Ovens thermal process improvement