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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


bulbous joint

Northern Coatings

#27319

bulbous joint | 19 February, 2004

Has anyone ever seen a bulbous joint (on 1 side of 1 designator) on every board? I checked the preheat profile at the wave and have played with dwell times but can't get rid of it. I checked the archives and didn't find much. Any help would be appreciated.

reply »


RDR

#27320

bulbous joint | 19 February, 2004

This is a pretty common sympton when wave soldering chip components. As a note IPC allows for this in wave process. The problem is in the pell off of the wave and you may not be able to fix. Is your back flow off of the wave only present when a board is running over it? If not you may want to adjust this flow and see if it goes away. What is your solder temp?

Personally I wouldn't worry about it but if you want to play try the wave flow dynamics.

Russ

reply »

Northern Coatings

#27323

bulbous joint | 19 February, 2004

What would cause this? Too much pre-heat, not enough, too much dwell, not enough? I am staying within the flux specs for preheat and can't seem to affect it much. Can board/parts contamination lead to this? I think that is a reach but I'm runnin out of ideas. I have a Vectra with Contour wave so the wave is flowing fine. The solder is 500f.

reply »

#27326

bulbous joint | 19 February, 2004

We're inclined to buy Russ's stuff about peel from the wave. Look here, but shield your eyes first: [ http://www.thepdfshop.co.uk/ppm/defects/Wave/asp/50.asp ] [A respectful bow to BW for the pix]

Could you tell us more about: * Board direction through the wave [relative to the problem bulbousity] * Component type, primary /secondary side, leads relative to wave, etc * Conveyor angle, flux, solder alloy, direction of solder flow on the wave surface, etc

What do you bet that if you ran the board 180* [from the way you run it now] that the bulbs would move?

reply »

Northern Coatings

#27330

bulbous joint | 20 February, 2004

Dave, you are correct they DID move (blows my theory on bad parts)! I will update with the info. you brought up but I have 1000 of these boards on the line as we speak, any ideas on a quick fix? Inquiring managers want to know!

thanks

reply »

#27331

bulbous joint | 20 February, 2004

Lucky guess, eh? [Sorry to blow your theory on bad parts]

A 1000 of these boards on the line, ummm. Two ideas to consider: * Increase the conveyor angle [don't get carried away, just keep the angle below 8*]. * Balance the flow of solder so that the solder at the center of the pot is not moving.

reply »


RDR

#27335

bulbous joint | 20 February, 2004

Ah, the old contour wave. I don't really know how to adjust those to get correct backflow to reduce your solder volume. I need to ask why you are worried about this. Do you have a QC requirement in excess of IPC class3 wich allows for this joint to be 1.5 times the height of the component? this is what is happening, The solder if you will, is being pulled/snapped back onto the termination of the component as opposed to being stripped off by the wave. This is where the wave flow dynamics come imto play. Ibelieve if my memory serves me right that the contour wave is a 50/50 flow off of the front and back. We use laminar flow waves with a flow of 90-95% front flow and back flow only very slightly present that can be stopped by just touching the solder. Or backflow only when the board is actually rding over the wave. Your pre-heat is not critical here since these little chip parts when they hit the wave are going to be at solder temp in a very short time. pre-heat is used to activate flux and prevent thermal shock when the assembly hits the wave along with getting leads and such to a closer temp to promote solder wicking/wetting. This would make me believe that it is not dwell either since dwell is usually a factor in getting T.H. top solder, the extra dwell allows for more heat to be spread through the lead and board to promote better wetting/solderwicking through the board. Changing the conveyor angle as Dave suggests could take care of your problem. Have you checked your immersion depth and parallelism of the wave to PCB? As far as immersion depth I do not know if you would want to go deeper or shallower but this may have effect on this also. Parralelism is very critical in succesful wave operation you want to make sure that you don't have a "deep side" this greatlt affects the peel off of the wave.

Hope this helps a little

Russ

reply »

Northern Coatings

#27339

bulbous joint | 20 February, 2004

I talked QC into accepting the bulbous joint, the problem was it was often shorting to a well placed via next to it. I say WAS because I reduced the dwell time in the main wave (you are correct 50/50 flow) to approx. a nanosecond, which reduced it enough that it no longer shorts. Parellelism is good so I am going to increase the conveyor angle Monday (per Dave) and we'll see how it goes from there.

thanks for all the help

reply »

#27349

bulbous joint | 21 February, 2004

Here's another idea to try: Rotate the boards 90*

reply »

greg york

#27360

bulbous joint | 23 February, 2004

> Has anyone ever seen a bulbous joint (on 1 side > of 1 designator) on every board? I checked the > preheat profile at the wave and have played with > dwell times but can't get rid of it. I checked > the archives and didn't find much. Any help would > be appreciated.

Use a low solids no clean rosin flux 1 - 1.5% (Higrade 3541 or similar)and keep preheat around 90C topside not too high or flux will move away. This will give you fix immediately. Then look out for too much convection as very dangerous to fluxes. Finally fix wave properly as either too deep or shallow in wave so poor drainage.Adjust back plate to ensure solder sits on turn on then flows over back once board hits the wave, look for 2 - 2.5 seconds dwell and ensure parralell wave very important

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

Essemtec Spider SMT Fluid Dispenser

Six Sigma Training