0
\$\begingroup\$

I have designed a power supply card using V7805-2000. Copper thickness is 70 microns, board size is 5 cm × 5 cm.

When this underwent vibration test the regulator came off, broken into two pieces(regulator leads are in the holes).

enter image description here

I have used 40 mils drill size, 130 X 70 regular pad size. 135 X 75 mils anti pad, thermal pad and solder mask top and bottom.

enter image description here

In the above image, you can see that there is an 0.5 mm notch from the surface of the regulator.

I have used vias on one side, where this notch gets mounted/placed on the PCB. Hence the regulator moves/shakes slightly when moved with a hand after assembling.

Even there are copper shapes on one side of the notch.

Is this might be one of the reasons why the regulator broke?

\$\endgroup\$
2
  • 1
    \$\begingroup\$ You need a better description of exactly how it broke. Pictures would help. An example board layout as well if you think that's a concern. If all else about the board is well, some additional fixative may be required. \$\endgroup\$
    – K H
    Commented Mar 4, 2021 at 8:12
  • 3
    \$\begingroup\$ I recommend using the laydown variant of that regulator module and using an adhesive tape to hold it down. Going through vibration tests with top-heavy upright components is very hard. Even TO-92 cases can give you a headache. \$\endgroup\$
    – Janka
    Commented Mar 4, 2021 at 8:48

2 Answers 2

Reset to default
3
\$\begingroup\$

If the module isn't mounted flush with the board and can actually moved by hand it can surely be a reason for a vibration failure. Copper is horrible when repeatedly flexed.

I'd suggest to strain relief the assembly with some elastic compound to the board, if possible (silicone, silastil, whatever is called around you). It's also important to evaluate if the module is 'top heavy', since the regulator body will leverage to the board.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Elastic compound? Wouldn't that still allow (some of) the movement and over time cause metal fatigue in the pins? \$\endgroup\$
    – Arcatus
    Commented Mar 4, 2021 at 8:46
  • 1
    \$\begingroup\$ there are various grades of elastic. If your part is really heavy it could break the substrate. If you are worried you can even epoxy it to the board. Ask the module manufacturer if they have tips \$\endgroup\$
    – Lorenzo Marcantonio
    Commented Mar 4, 2021 at 8:59
3
\$\begingroup\$

You will find on all power supplies for PC’s large components like these with polyurethane adhesive to provide structural support and damping.

I can find some sources but there are many other answers on this site for this same issue. They come in various ratings of shore hardness and I would search for the higher ratings >50 and not the lower levels. That will solve your problem.

This is a common problem for aerospace vibration level tests with 15g acceleration and a common solution.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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