4
\$\begingroup\$

I am looking for a standard or guideline which describes how much torque should be used to fasten a PCB to an enclosure using different types of fasteners.

For example, we use M3 hex nuts in a system. There are ~25 people doing assembly. We do not want everyone to guess how tight the nut should be turned. We would rather give an appropriate tool to everyone and make sure each PCB is fastened the same way.

\$\endgroup\$
  • \$\begingroup\$ Not that I know of. A lock washer and / or thread locking adhesive is a good idea. \$\endgroup\$ – George Herold Jun 18 '15 at 14:31
  • \$\begingroup\$ Way more important than torque would be if the PCB is flat and how much it flexes due to tightening. And most importantly where the SMD components are located and oriented (particularly SMD ceramics). Always better to under-tighten, than over-tighten. \$\endgroup\$ – Indraneel Dec 25 '18 at 19:38
6
\$\begingroup\$

It depends most on what and how you are fastening. Many different materials will have different properties.

For example, the compressive strength of FR4 material is at least 460MPa, or 460N/mm^2.

That number is one you want to stay well clear of, especially if you have multi-layer boards (since you don't know the exact specs of the prepreg your fab uses). So let's say "definitely don't go over 100N/mm^2".

Now here's the rub. That's the compression strength, but it does mean you have to have proper rings to prevent extra warping forces when the nut is tightened and even then some warping may still happen if tightened too far.

Then how that translates to a Nm reading depends on:

  1. The surface area of the rings you use: Smallest surface determines the maximum number of downward Newtons.
  2. The threading on the bolt; It's neatness and the number of threads per unit length.

Doubtless there's conversion charts for that on the internet for certain types of screws, but I couldn't find them just now (though I did only spend about 1 minute searching, since that's also what it took to get the compression strength of FR4).

All that said, I do think the tiny surface area of the threads on your screws, unless you use 13.9 grade metal, maybe, will be the weakest link in your system. The cheap DIY store screws will definitely break before up to 8 layers of PCB. So will, as I know from experience, RVS A2 screws.

That is, I did once split a 4 layer board with an M4 screw and two washers, but the RVS A2 one broke on hand-tightening, then I took out a military grade 13.9 one and needed a power wrench to tighten it enough to crack open.

But if you want to stay super safe, get the data of a brass spacer in the thread size you use and then apply the maximum allowable torque for that, even if you use harder types of fasteners, a metalised PCB hole will certainly survive quite well. Brass spacers are relatively weak compared to FR4 compression strengths and usually with brass spacers you can still tighten stuff well enough to mount PCBs and such.

| improve this answer | |
\$\endgroup\$
4
\$\begingroup\$

I use a TackLife 1-4 Nm adjustable electric torque screwdriver.

1 Nm is too much torque for standard PCB material to handle before cracking if your screws don't strip. I'm assuming most PCB material is similar in composition.

0.6 Nm torque adjuster is required at work when assembling PCB components, 1 Nm for body/frame connections (metal on metal). Our electronics are exposed to vibration. Lock-Tite (blue) is used in critical components subjected to excessive vibration. Most of our fans are mounted using rubber screws as they are optical instruments sensitive to vibration and even BLDC fans are a problem.

These are imperical statistics, not out of a datasheet. Most people don't have a < 1 Nm torquer so the "hobby rule" is 1/4 turn past "hand tight."

| improve this answer | |
\$\endgroup\$
3
\$\begingroup\$

You can find maximum torque values for various strengths of M3 fasteners, for example Holo-Krome recommends 1-3 N-m depending on the type of fastener. If your fasteners are crappy soft metal that might be too tight. You can get a skilled assembler to do a few (as Andy has suggested!) and then compare with a torque-controlled electric driver if nothing else- then they'll be consistent. Or deliberately break a few and use a fraction of that torque.

Using a nut with integral lockwasher is not a bad idea if there is vibration, or you can use Locktite type thread locking (or buy fasteners with it already applied).

enter image description here

| improve this answer | |
\$\endgroup\$
  • 1
    \$\begingroup\$ At a previous employer that made automotive electronics, we attached PCBs into plastic housings (with molded bosses) with Hi-Lo screws all day without using any sort of threadlocker. The drivers were all push-start with calibrated torque limits. \$\endgroup\$ – Nick T Jul 22 '15 at 21:31
1
\$\begingroup\$

It will depend on PCB and fastener materials so my advice is to get the best guy in the assembly shop to decide how tight to turn the nuts and measure that torque. Nothing beats experience.

Try a few other guys too to see what they say. Lock-nuts will need a bit more torque of course.

| improve this answer | |
\$\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.