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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


#9916

Wave solder mystery | 26 August, 1999

I am building a board with T1/34 LED's in a 10X10 array there are 12 arrays in a panel. The LED's are body to body in the array, .1 spaced. The LEDS have been autoinserted on a Universal RAD III, so they are cut and clinched. The trouble I am having is in wave. Occasionally throughout the board I have solder joints that the solder has been pulled out of the joint. I know this because some of the joints have solder at the top of the fillet but not at the bottom. This ALWAYS happens on joint where the clinched leads trail out of the wave. If I re-run the board a second time in the reverse direction the bad joint will become good and good joints will become bad. I have profiled to my no clean chemistry and have tried 6 thru 7 degree wave angle. Anyone have any thoughts! Thanks in advance.

reply »

Brian Larson

#9917

Re: Wave solder mystery | 27 August, 1999

If your through holes are overlarge or the anular rings too small, the surface tension of the wave may exceed that of the joint. I would look at the design. You may want to put a strip of kapton tape on the following edge of the panel to test the "surface tension" theory.

| I am building a board with T1/34 LED's in a 10X10 array there are 12 arrays in a panel. The LED's are body to body in the array, .1 spaced. The LEDS have been autoinserted on a Universal RAD III, so they are cut and clinched. The trouble I am having is in wave. Occasionally throughout the board I have solder joints that the solder has been pulled out of the joint. I know this because some of the joints have solder at the top of the fillet but not at the bottom. This ALWAYS happens on joint where the clinched leads trail out of the wave. If I re-run the board a second time in the reverse direction the bad joint will become good and good joints will become bad. I have profiled to my no clean chemistry and have tried 6 thru 7 degree wave angle. Anyone have any thoughts! | Thanks in advance. | |

reply »

#9918

Re: Wave solder mystery | 27 August, 1999

Brian, Thanks for the input. The hole size is good but the anular ring is small. I have helped the condition by reducing the clinch on the LED's. This I believe is reducing the pull out path. Thanks again

| If your through holes are overlarge or the anular rings too small, the surface tension of the wave may exceed that of the joint. I would look at the design. You may want to put a strip of kapton tape on the following edge of the panel to test the "surface tension" theory. | | | | | I am building a board with T1/34 LED's in a 10X10 array there are 12 arrays in a panel. The LED's are body to body in the array, .1 spaced. The LEDS have been autoinserted on a Universal RAD III, so they are cut and clinched. The trouble I am having is in wave. Occasionally throughout the board I have solder joints that the solder has been pulled out of the joint. I know this because some of the joints have solder at the top of the fillet but not at the bottom. This ALWAYS happens on joint where the clinched leads trail out of the wave. If I re-run the board a second time in the reverse direction the bad joint will become good and good joints will become bad. I have profiled to my no clean chemistry and have tried 6 thru 7 degree wave angle. Anyone have any thoughts! | | Thanks in advance. | | | | | |

reply »

JohnW

#9919

Re: Wave solder mystery | 28 August, 1999

Doug,

Have you tried running the boards at an angle to the wave of between 15 - 30 degree's ? I've done this on a few boards to eliminate short's and I believe it will work on your problem very sucessfully. Best thing to do is get your wave solder carrier manufacturer to make you a carrier with a circular insert removed and get a circular holder to match your board , then run it through the wave at 5 degree intervlas until you get the best result's. Normally you get a range of good angles so pick the middle to make it a good process window.

John

Brian, | Thanks for the input. The hole size is good but the anular ring is small. I have helped the condition by reducing the clinch on the LED's. This I believe is reducing the pull out path. | Thanks again | | | If your through holes are overlarge or the anular rings too small, the surface tension of the wave may exceed that of the joint. I would look at the design. You may want to put a strip of kapton tape on the following edge of the panel to test the "surface tension" theory. | | | | | | | | | I am building a board with T1/34 LED's in a 10X10 array there are 12 arrays in a panel. The LED's are body to body in the array, .1 spaced. The LEDS have been autoinserted on a Universal RAD III, so they are cut and clinched. The trouble I am having is in wave. Occasionally throughout the board I have solder joints that the solder has been pulled out of the joint. I know this because some of the joints have solder at the top of the fillet but not at the bottom. This ALWAYS happens on joint where the clinched leads trail out of the wave. If I re-run the board a second time in the reverse direction the bad joint will become good and good joints will become bad. I have profiled to my no clean chemistry and have tried 6 thru 7 degree wave angle. Anyone have any thoughts! | | | Thanks in advance. | | | | | | | | | | | |

reply »

#9920

Re: Wave solder mystery | 30 August, 1999

| I am building a board with T1/34 LED's in a 10X10 array there are 12 arrays in a panel. The LED's are body to body in the array, .1 spaced. The LEDS have been autoinserted on a Universal RAD III, so they are cut and clinched. The trouble I am having is in wave. Occasionally throughout the board I have solder joints that the solder has been pulled out of the joint. I know this because some of the joints have solder at the top of the fillet but not at the bottom. This ALWAYS happens on joint where the clinched leads trail out of the wave. If I re-run the board a second time in the reverse direction the bad joint will become good and good joints will become bad. I have profiled to my no clean chemistry and have tried 6 thru 7 degree wave angle. Anyone have any thoughts! | Thanks in advance. | | Doug: Could you slow-down your conveyor? Good luck. Dave F

reply »

Brian Conner

#9921

Re: Wave solder mystery | 31 August, 1999

| I am building a board with T1/34 LED's in a 10X10 array there are 12 arrays in a panel. The LED's are body to body in the array, .1 spaced. The LEDS have been autoinserted on a Universal RAD III, so they are cut and clinched. The trouble I am having is in wave. Occasionally throughout the board I have solder joints that the solder has been pulled out of the joint. I know this because some of the joints have solder at the top of the fillet but not at the bottom. This ALWAYS happens on joint where the clinched leads trail out of the wave. If I re-run the board a second time in the reverse direction the bad joint will become good and good joints will become bad. I have profiled to my no clean chemistry and have tried 6 thru 7 degree wave angle. Anyone have any thoughts! | Thanks in advance. | | Doug,

Been there done that, cursed at it, and tried somthing else!!! Seriously, the clinch is the issue. You can try slowing the conveyor down, cooling down the assembly, heating up the assembly, speeding up the assembly - nothing eliminate this problem.

We were running a board with 100's of zip components(20-24 legs)that were auto-inserted. I had the same results that you did. One day, the inserter died on the 2nd to last board. We prepped the zips and hand placed them - all 256 of them(luckily, one of the smaller boards!!) on the last board!!! The partially loaded assembly was populated and sent through the wave - not a single defect on the hand-loaded zips had a problem!!! The assembly that was completely hand-loaded with zips had no defects!!!

Reccomendation...if you can hard tool your auto-insert machine to only clinch one outside lead on each side, you will reduce the amount defects immediately!!

Good Luck,

Brian Conner

reply »

reflow oven profiler