Author Topic: thermally conductive spray adhesive?  (Read 840 times)

0 Members and 1 Guest are viewing this topic.

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 3556
  • Country: us
  • 💎
thermally conductive spray adhesive?
« on: November 22, 2018, 02:51:52 am »
So my LED project hit a bit of a snag because I realized that when you recycle LED lights, they don't actually have insulator behind them, so I had exposed copper, and I was forced to use kapton tape as a insulation layer over my aluminum strips.

I then realized I don't have a thermally conductive adhesive spray.

Does such a project exist for future reference? I completed the project with 77 3M adhesive  :-X

I have my doubts about its thermal conductivity but it seems to work OK for a salvage job, even though I feel like a pirate.
 

Offline DaJMasta

  • Super Contributor
  • ***
  • Posts: 1815
  • Country: us
    • medpants.com
Re: thermally conductive spray adhesive?
« Reply #1 on: November 22, 2018, 03:36:05 am »
Does it need to be a thermally conductive electrical insulator?  The insulation could be tricky because the spray isn't going to be even/uniform thickness outside of a very controlled environment, but I'd imagine that a decent thermal conductor is probably manageable.
 

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 3556
  • Country: us
  • 💎
Re: thermally conductive spray adhesive?
« Reply #2 on: November 22, 2018, 01:34:05 pm »
no I don't trust a spray film for a electrical insulator, thats why I put kapton. I would put kapton even if the spray was certified because I don't have a super controlled environment to apply it, dry it and test it properly. I think you would need to do process development to use that kind of technology properly. 
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 8188
Re: thermally conductive spray adhesive?
« Reply #3 on: November 22, 2018, 03:06:43 pm »
The usual way to get a guaranteed glue line thickness is to use epoxy with a pinch of glass microbeads of the desired diameter mixed in.   As long as there are a few microbeads in each area of the joint, and the surfaces are reasonably hard, their diameter sets the glue line thickness.  The amount of microbeads to glue is very small - 1% by volume would be on the high end.  Don't use microbaloons - they are too fragile and you don't want little pinhole voids.

Its not spray glue, but a dab of thermally conductive epoxy with glass microbeads mixed in per LED could provide mechanical and thermal bonding and also electrical insulation.   Pick a slow cure epoxy to give you enough time to use up the minimum mix volume before it starts curing.

N.B. regular epoxy doesn't do well at high temperatures so tempting as it is to load up dollar store 24H epoxy or laminating epoxy with alumina powder or preferably a more thermally conductive insulator in powder form, you'd need genuine thermally conductive epoxy sold for heatsink bonding.
 
The following users thanked this post: cdev

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 3556
  • Country: us
  • 💎
Re: thermally conductive spray adhesive?
« Reply #4 on: November 22, 2018, 03:17:22 pm »
how do you press it over LED strips? Do you need a special jig? Some kinda special roller (maybe with indents over the LED/Resistors/Joints), foam, etc? I just did it by hand carefully and thoroughly.

Or do you make a special squeege?

I only have fiber-glass additive to epoxy but I like the idea, maybe I will purchase some microbeads for good measure eventually.

I could see it being done with a special wooden roller with the correct face to roll over the components with minimum pressure but it seems like you would need a multi-axis lathe to do it right to make a single bead layer. Ok realistically you can do it with a dremel, fine point and alot of time and care, but damn that seems like a pain in the ass unless LED strips are highly standardized... I guess quality is hard. Or do a mold over a set length then make a inverse mold then wrap it on something. I was hoping for quick and easy  :wtf:

Can you dip it in a ultra slick adhesive with ultra low surface tension, dust it with the beads, blow it out, then do the epoxy? It seems like it MIGHT work so long the glue is ridiculously slick. Maybe ultra thin super glue dip and inert gas blow out to make a minimum thickness layer? Still a completely disgusting process!




I am overthinking this, its 1%.
« Last Edit: November 22, 2018, 03:33:54 pm by coppercone2 »
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 8188
Re: thermally conductive spray adhesive?
« Reply #5 on: November 22, 2018, 05:03:16 pm »
Anyone who uses a Dremel on uncured epoxy deserves a weird science award!  :scared:  Their hair will certainly look the part.  :popcorn: :-DD

Just dab the mixed epoxy, already containing the glass spacer microbeads, where its needed, or load it into a disposable syringe with a tapered plastic nozzle you can cut to size, so you can make a thin bead along the strip.

Then all you've got to do is press it onto the lightly sanded and freshly degreased heatsink surface hard enough that the epoxy squeezes out till the spacer beads are in a mono-layer touching both surfaces.    Some sort of a press jig with one side covered in thick closed cell polyurethane foam to protect the components on the strip, with clingfilm over it to aid cleanup would be a big help when applying pressure.

Its also a smart idea to put masking tape (the good blue kind) either side of where the strip's got to go, to make it easier to clean up  excess epoxy with the corner of a plastic scraper or squeegee, then peel off the tape before the epoxy gels to remove the resulting smear.
« Last Edit: November 22, 2018, 05:15:19 pm by Ian.M »
 

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 3556
  • Country: us
  • 💎
Re: thermally conductive spray adhesive?
« Reply #6 on: November 22, 2018, 05:04:58 pm »
yea I love the new making tape home depot sells.

I need to see how well jb weld compares to thermal epoxies ( i think it is a thermal epoxy).

BTW possible trap for young players: test your epoxy conductivity when its cured. For some reason the MGchemicals silver epoxy I have does not measure low conductivity until its cured, like if you stick probes into a liquid pool.

ok, its .58 vs .68 jbweld vs mgchemicals. 3m is 0.72. ,loctite is 0.6.


seems like jb weld is a fair choice for LED light strips that no one actually bothers to heat sink anyway.

nmv, this is best so far
https://www.mouser.sk/ProductDetail/MG-Chemicals/8329TCM-6ML?qs=tfZGHB2PWd2xZF14WWhlNQ==&gclid=EAIaIQobChMI-Zzt_r7o3gIVlYjICh3V3AakEAQYBCABEgJV-PD_BwE

Does anyone know of better figures then 1.4W/c?
« Last Edit: November 22, 2018, 05:11:21 pm by coppercone2 »
 
The following users thanked this post: cdev

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 8188
Re: thermally conductive spray adhesive?
« Reply #7 on: November 22, 2018, 05:13:08 pm »
JBweld is not a thermal epoxy, but some of their product range are heat-resistant (or at least what's regarded as heat resistant in a 'consumer' epoxy).

Make sure you get ceramic filled JBweld, not the ordinary steel filled JBweld the company is famous for.   Although they *claim* the steel filled stuff is electrically non-conductive, I wouldn't bet on that if high voltages are involved.
 
The following users thanked this post: cdev


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf