5
\$\begingroup\$

I have an inertial measurement unit PCB that picks up changes in acceleration. I want to secure it inside a plastic box/casing. The products (and pictures) I want to use are below:

PCB https://www.sparkfun.com/products/13303

Plastic box/casing https://www.aliexpress.com/item/1-piece-plastic-standard-box-76-35-20mm-2-99-1-38-0-79-inch-box/1439035089.html (The PCB fits inside easily, it measures about 2.5cm x 2.3cm)

Now it's incredibly important that the PCB - which has the sensor - is very rigid to the plastic box/casing. I want to measure motion of the box as precisely as possible. I'm not sure if using the 2 corner holes with screws/nuts is sufficient as it's only on one side and therefore provides some leeway for the other side to flex slightly.

So I was thinking why not just glue the damn thing? Glue doesn't conduct electrical signals (as far as I know) so that's not an issue. And the part being glued, the PCB, is extremely light, so the holding force need not be too high.

Will this work? If so, what glue/adhesive product should I use? Are there better solutions? This is a hacked-together project, so professionalism isn't the highest priority. Just the requirement that the PCB is very rigidly attached to the bottom half of the plastic case. Thanks for any help!

\$\endgroup\$
2
  • 2
    \$\begingroup\$ The flexing is on the order of 1 to 0.1 millimeters, with sufficient force. It's a non issue unless the sensor is a, on a rocket, and b, has fractions of a millimetre resolution. \$\endgroup\$
    – Passerby
    Jul 27 '17 at 17:54
  • 2
    \$\begingroup\$ An alternative to gluing it would be to submerge the PCB in a pool of potting compound filling about half the case. Advantages : near perfect mechanical coupling, greater moisture resistance. Disadvantages : difficult (well nigh impossible!) maintenance. \$\endgroup\$
    – user_1818839
    Jul 27 '17 at 19:54
4
\$\begingroup\$

Finding a glue that reliably adheres to plastics can be difficult. As such you just can't beat good old screws, spacers, and nuts for solid mechanical connection.

In your case, since you want a direct transferral of motion information to the board from the box, I might be inclined to augment the screws with an epoxy filler that fills the bottom of the box and covers the PCB. As a bonus, the PCB will then also be protected from ay sources of moisture, contaminants etc.

BTW: If this is a motion sensor you will want the board attached to what looks like the lid in your picture (The bit on the right) since that has the mounting lugs, assuming you plan on using those. The deeper part of the box would then be the cover.

\$\endgroup\$
7
  • 2
    \$\begingroup\$ I agree with screws, and especially the epoxy filler. Only issue with that is that maintenance is impossible. However, I would also use a short standoff as well, that would firm mount it to the standoffs then expoxy in. Or standoffs with just expoxy underneath to secure the bottom. \$\endgroup\$
    – biggi_
    Jul 27 '17 at 18:26
  • 1
    \$\begingroup\$ @biggidvs yup, its a one time throw away on failure solution \$\endgroup\$
    – Trevor_G
    Jul 27 '17 at 18:58
  • 1
    \$\begingroup\$ @JDS the filler will do exatly that, fill the spaces. \$\endgroup\$
    – Trevor_G
    Jul 27 '17 at 19:16
  • 1
    \$\begingroup\$ @JDS.. btw, if you do use an epoxy, the question becomes..do you really need a box at all... You could just screw and epoxy bond it to a metal plate. \$\endgroup\$
    – Trevor_G
    Jul 27 '17 at 19:17
  • 1
    \$\begingroup\$ @JDS there are epoxies specifically designed to pot electronics. Ones that insulate electrically while passing heat. Not sure you want the local hardware store variety. \$\endgroup\$
    – Trevor_G
    Jul 27 '17 at 19:52
3
\$\begingroup\$

The problems with glue are:

  • Depending on formulation, it may eventually degrade into a more conductive substance, absorb water, etc.

  • It can interfere with testing/servicing the board compared to screws which are easily removable.

In this case, neither should be a problem since you are using a standard board you won't need to debug itself and the overall system is a prototype.

I would suggest using a hot-glue gun and standard high-temperature glue sticks. Hot glue is not all that strong on smooth surfaces, as glues go (good in case you want to undo it — I have also heard that isopropyl alcohol can help remove it) and can fill gaps and is fairly rigid compared to, say, double-sided foam tape (good for transmitting vibration).

Avoid getting the glue on or near the board's wire holes so you can replace the wire to the board as needed, but do let it squeeze into the mounting holes for a stronger joint than just being between two flat surfaces.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks for your answer, good to know I can fall back on this for now. \$\endgroup\$
    – JDS
    Jul 27 '17 at 19:15
  • \$\begingroup\$ I'd add to "score" and "rough up" a plastic surface first. The glue will adhere to a rough plastic surface much better. \$\endgroup\$
    – rdtsc
    Jul 28 '17 at 1:29
1
\$\begingroup\$

Personally I would use bolts and standoffs. If you just want a temporary fix, I would use VHB tape rather than glue.

\$\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.