2
\$\begingroup\$

Let's say I have a square PCB with mounting holes near each corner and that the PCB will be mounted to an enclosure with 4 standoffs. Is there a rule of thumb about how much tolerance I can have on the height of those standoffs? If the standoffs are not all the same height, then the PCB is going to flex to some degree, so perhaps another way to ask this is how flat does a PCB have to be when it is mounted?

\$\endgroup\$
  • 1
    \$\begingroup\$ That all depends on your components, how they are mounted, etc, and the type of PCB it is. \$\endgroup\$ – Majenko Jul 10 '14 at 18:41
  • \$\begingroup\$ ^this, anyway I'd say that "more than you think" is a quite safe assumption. Have you ever mounted an heatsink on the CPU on a motherboard? PCBs can flex a lot. \$\endgroup\$ – Vladimir Cravero Jul 10 '14 at 18:42
  • \$\begingroup\$ Does rather depend on the size of the PCB and the magnitude of the error. Some further information would be useful. \$\endgroup\$ – Brian Drummond Jul 10 '14 at 18:48
  • 2
    \$\begingroup\$ 1mm or so on a 4" or larger 1.6mm board won't likely cause any problems. The standoffs I usually specify (PEM) have a tolerance of +.05 –.13mm so there's lots of margin. Thinner boards can easily flex more, but you have to be concerned about the integrity of solder joints to the largest parts in particular, and especially if the package has no "give" (BGA, QFN, 2512 etc.). \$\endgroup\$ – Spehro Pefhany Jul 10 '14 at 18:53
1
\$\begingroup\$

As the comments have mentioned it depends on components, size of board, etc.

As for the board itself, if the standoffs are that irregular then your own strength is going to let you down well before you reach the board's breaking point. A quick look at the datasheet for the boards I get fabricated (from pcbtrain.co.uk, see link below) shows the warp strength is in the region of 50 Kpsi (400+ MPa).

Surface mount components are going to give you the most trouble with bending boards, as others have said. But don't forget things like large electrolytics or bulky connectors. A flexed board is going to have a weak spot. And you'll end up with an oversized capacitor coincidentally placed right on that weak spot. And that oversized cap will sit very close to the chassis that's housing your lovely, new (flexed, but who cares, right?) PCB. And you're going to drop it. And, like a piece of electronic toast, it's going to land upside down, right where that capacitor is pressing against the board. And the board is going to say...

Nothing, because it's broken. At my previous job I have seen it happen regularly in certain mass-produced small electronic devices like pocket and portable radios; the break is ALWAYS in the same place for a given product, and it's invariably a catastrophic failure. In some cases, if the flex is that bad on the board (which you really only get if it's screwed into standoffs, not clipped) you're actually better off not using the anchor point that causes the greatest anomaly.

Datasheet ref: http://www.pcbtrain.co.uk/cp/uploaded/VT481TDSrevA18_9_13.pdf

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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