Bad solder joints on surface mount connector

[Jimbo7]

Hi, I have a problem with a surface mount connector. .. nearly half of the 40 pins have bad solier joint after reflow. They look OK but they fail after a few weeks (cracking). Have tried touchup by reflowing the bad joints using solder iron but the parts fail again a few weeks later. Using sn63 leaded solder. Connector pins have Ni Au flash finish. Any ideas what could be the cause?
Thanks.

 

[barry]

They fail just sitting there, or is there something you're not telling us? Temperature cycling? Mechanical stress? Environmental?

Perhaps inadequate cleaning before soldering.

 

[dick_freebird]

Inspecting the joints more closely would probably tell
you (or more likely, a trained inspector or assembler)
what the cause is. Whether the solder itself is cracking,
or failing to adhere in the first place, "cold joint" etc.
are things that people get trained to judge. I'd bet you
could find descriptive pictures on the Web, but do you
have the microscope, time and confidence? Or better
to tap someone who does?

 

[Jimbo7]

Thanks Barry, thanks dick
Barry,
no deliberate vibration or temperature cycling applied. However the parts will have seen some baking at 70oC for a few hours to cure some glue used elsewhere. Actually, we try to fit a brass shell to the connector after connector is fitted and when we solder that shell in place some of the connector pin solder joints fail. Maybe heat is causing the problem ?

Dick
thanks. I have tried to add a photograph to see if anyone can spot whether these are cold joints or cracks.

 

[BradtheRad]

I have tried to add a photograph

Sorry, no image is visible. Try the 'Add an Image' button, it's reliable and straightforward.

 

[barry]

If i had to guess, I'd say it's contamination. Try degreasing before soldering.

 

[Jimbo7]

Now with image added

 

[dick_freebird]

To my untrained eye (one that has only seen plenty of
my own work) it looks like the joints were moved while
still (partially) molten, crystallized ("cold") rather than
shiny, dragging material and developing voids as a result.
Might look to the fixturing of the upper piece and the
thermal ramp rates, keep things firmly in place until the
assembly has truly solidified.

Might need more than gravity, to do it. Especially if the
upper piece has thermal flexure of its own, and is large.

 

[FvM]

Can give a link to the connector data sheet? I don't understand the pin form and how it's touching the pad. Why we see a "huge" block of solder?

 

[BradtheRad]

Hairline fractures are visible in the solder under each pin.

 

[Jimbo7]

Thanks Dick, good idea about holding it in place.
FvM, I don't have full datasheet sorry. It's a samtec ZHR5-20-02-F-RA though. Drawings available online. Incidentally, we had some problems with the plastic connector warping during reflow. Anyone had similar problem before? I wonder if that movement is causing cold joint?

 

[FvM]

Thanks, I can see the pin form. My impression of an excessive gap between pin and pad is probably wrong. However, what's the actual gap after reflow? Did you make a microsection?

Incidentally, we had some problems with the plastic connector warping during reflow. Anyone had similar problem before? I wonder if that movement is causing cold joint?

I think, it can well explain it. Wrong temperature profile?

 

[marce]

I concur, the joints look overcooked and as stated previously some movement during the soldering process. I also think there is far to much solder on the joints, check the footprint and actual PCBs especially the locating holes for the connector. Also check that it is flat to the board pre-reflow.

 

[Jimbo7]

FvM, yeah microsection shows a small gap between pad and tail, maybe 0.1mm or so. On bad joints this is quite full of voids and some are full cracks.

Marce, thanks. We are trying different stencils. .. can too much solder be a bad thing (assuming we don't get shorts?