2
\$\begingroup\$

I'm running into an issue with a specific RJ45 SMT connector where the I'm getting at least 1 pin lifting on >50% of the boards that run through our reflow oven. The rest of the components on the PCB do not have any issues with soldering, only this RJ45 connector.

The issue I'm having is with pins 9 or 12 in the image below.

RJ45 Footprint

The stencil that we are using has a thickness of 5 mils.

Most pins that are lifting are still partially soldered. The issue is that often with a slight amount of pressure on the pin the joint will crack and separate. It's like the paste has melted and then the pin lifts. With the pin lifted, the surface tension of the solder maintains the connection but it is very weak. If the pin lifts far enough the surface tension will break and the pin will no longer have a connection.

The things I have tried in order to resolve this problem:

  1. Double check hole drill sizes, I've also tried enlarging the holes with a drill.
  2. Adjusted height of connector for more or less pressure from the pick and place machine when placing the component.
  3. I've double checked the amount of pressure applied by the pick and place machine when picking the part. No pins appear to be lifted prior to entry to the oven.
  4. Reflow profile has been adjusted to bring in to line with the manufacturer specified profile while not deviating from the allowable/recommended paste settings. For reference, we are using Indium 8.9hf as our paste.
  5. I've tried cutting off the plastic posts so that nothing is protruding through the holes and all of the strain relief is happening on the tabs on the front of the connector. This issue does not appear to be tied to the posts.
  6. I've tried adjusting placement of the part on the pad forward/backward to see if relative placement on the footprint has any affect.

Any additional suggestions on things I can try to resolve this issue are appreciated. I think I've covered most of the basics, but maybe somebody has some additional experience with a similar issue and can lend me the benefit of their experience.

\$\endgroup\$
1
  • \$\begingroup\$ It seems that you've investigated it quite well, so at this point I'd suspect a defect part rather than something in the solder process. Please post the part number used. Is it from a well-known vendor or from some "Aliexpress"? If the plastic isn't actually suitable for reflow then obviously all manner of strange stuff will happen. \$\endgroup\$
    – Lundin
    Commented Sep 13, 2023 at 9:10

2 Answers 2

2
\$\begingroup\$

I think its probably due to a bad part. If its just a few pins that are lifting during soldering then it's most likely from internal mechanical stress (that shouldn't happen). Pins should not move or only move less then the solder has tension. Just the fact that you can see the solder bridges means that it is the part that is moving during reflow.

I would probably take one of these RJ45 parts and put them through the reflow oven and see if the pins move during reflow not attached to the board.

If the pins are moving on their own during reflow then that would be some good evidence to give to the manufacturer that their part is not good.

\$\endgroup\$
-1
\$\begingroup\$

The pins that lift are the corner pins, the corners will heat faster during reflow, check that your reflow profile is compatible with this part.

\$\endgroup\$
6
  • \$\begingroup\$ "The pins that lift are the corner pins, these pins will heat faster during reflow" Eh? It kind of depends what signals that are connected there, yeah? There's no mentioning of what signals that are used, so you can't assume that some pins heat faster than the others. Grounded pins will obviously heat slower and so on. You can't regard a component as some isolated entity complete detatched from the PCB design, when the whole purpose of soldering is to attach something to a PCB. \$\endgroup\$
    – Lundin
    Commented Sep 13, 2023 at 9:13
  • \$\begingroup\$ no, it;s simply that the corners have more suface area near them hold a cube of cheese over a grill (on a skewer) and see which bits melt first.. \$\endgroup\$ Commented Sep 13, 2023 at 14:11
  • \$\begingroup\$ We are talking about connectors soldered to a PCB here. If your argument against "grounded pins will obviously heat slower" is just "no, grill some cheese" then I think we are done talking here... \$\endgroup\$
    – Lundin
    Commented Sep 13, 2023 at 14:20
  • \$\begingroup\$ we're talking about reflow, it's not just metal that gets hot in there. \$\endgroup\$ Commented Sep 14, 2023 at 3:11
  • \$\begingroup\$ Your "answer" (could have been a comment) explicitly speaks of pins and of nothing else. \$\endgroup\$
    – Lundin
    Commented Sep 14, 2023 at 6:49

Not the answer you're looking for? Browse other questions tagged or ask your own question.