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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


unsoldered joints

jacqueline coia

#14157

unsoldered joints | 22 September, 1998

need advice to point me in the correct direction for solution to the problem of unsoldered joints which look soldered under normal visual inspection, but when pull strength test is performed the leads are coming away from pad, no strong intermetallic joint has been produced, although pad/lead are reflowing. Not sure if problem relates to reflow profile, batch of solder paste or just the temperamental character of water soluble paste. advice and comments much appreciated.

reply »

Steve Gregory

#14158

Re: unsoldered joints | 22 September, 1998

Hi Jacqueline! Do these joints that you're having problems with happen to be on fine pitch solder joints? Are the fillets staying attached to the foot and separating from the pads cleanly? If that is true, are there many vias concentrated around the QFP footprint and attached to the pads by a very short trace? If that is true and the board is being wave soldered, then what you have is that the pads are becoming hot enough to almost reflow the solder joints again when passing across the wave, and losing the original bond created from the initial reflow. The boards need to be corrected by tenting those vias that are so close to the pads, or to move those vias further away from the pads. To get through building those boards until they get corrected, you can use either kapton tape, or a liquid mask and cover those vias before you stuff the through hole parts. That should prevent the thermal transfer during wave that causes those solder joints to lose their bond. I hope this helps! -Steve Gregory-

reply »

Mike C

#14159

Re: unsoldered joints | 22 September, 1998

| Hi Jacqueline! | Do these joints that you're having problems with happen to be on fine pitch solder joints? Are the fillets staying attached to the foot and separating from the pads cleanly? | If that is true, are there many vias concentrated around the QFP footprint and attached to the pads by a very short trace? If that is true and the board is being wave soldered, then what you have is that the pads are becoming hot enough to almost reflow the solder joints again when passing across the wave, and losing the original bond created from the initial reflow. The boards need to be corrected by tenting those vias that are so close to the pads, or to move those vias further away from the pads. | To get through building those boards until they get corrected, you can use either kapton tape, or a liquid mask and cover those vias before you stuff the through hole parts. That should prevent the thermal transfer during wave that causes those solder joints to lose their bond. | I hope this helps! | -Steve Gregory- Steve I have had that problem also Whereas the joint looks great, but you can move the pin left and right and the solder joint is attached to the leg and not the board. The problem was tracked to the wave solder and it was discovered the profile was to hot and the top side was seeing reflow temp's (even without Via's in the area). But could this symptom also be an indication of intermetalic growth? Mike

reply »

Chrys

#14160

Re: unsoldered joints | 22 September, 1998

| | Hi Jacqueline! | | Do these joints that you're having problems with happen to be on fine pitch solder joints? Are the fillets staying attached to the foot and separating from the pads cleanly? | | If that is true, are there many vias concentrated around the QFP footprint and attached to the pads by a very short trace? If that is true and the board is being wave soldered, then what you have is that the pads are becoming hot enough to almost reflow the solder joints again when passing across the wave, and losing the original bond created from the initial reflow. The boards need to be corrected by tenting those vias that are so close to the pads, or to move those vias further away from the pads. | | To get through building those boards until they get corrected, you can use either kapton tape, or a liquid mask and cover those vias before you stuff the through hole parts. That should prevent the thermal transfer during wave that causes those solder joints to lose their bond. | | I hope this helps! | | -Steve Gregory- | Steve | I have had that problem also | Whereas the joint looks great, but you can move the pin left and right and the solder joint is attached to the leg and not the board. The problem was tracked to the wave solder and it was discovered the profile was to hot and the top side was seeing reflow temp's (even without Via's in the area). | But could this symptom also be an indication of intermetalic growth? | Mike Sounds like topside reflow at the wave to me. Best thing to do is attach a couple of thermocouples and run a profile. Get the fastest conveyor speed possible while keeping preheat temps in the 220F zone at the end of preheat. It's usually the wave contact that's melting the joints. And the chip wave pumps more heat into the board than the smooth wave. It seems counterintuitive, since the contact time is much longer on the smooth wave, but the chip actually pumps a higher volume of hot solder acros the board. It's the heat available for transfer on the chip wave that kills ya. Remember that old Q-dot from heat transfer class? If the wave isn't the problem, check the plating on the circuit board. It is highly possible on HASL board to have them plated with non-eutectic solder. I see it a lot on parts that don't wet. If the solder is 80%tin, 20% lead, it's melting temp is about 210C. So if you're peaking at 215, your paste will reflow beatifully, but the solder on the pads never made it into the solution.

reply »

fluid dispensers

IPC Training & Certification - Blackfox