Author Topic: How thermally conductive is potting compound?  (Read 4281 times)

0 Members and 1 Guest are viewing this topic.

Offline Razor512Topic starter

  • Regular Contributor
  • *
  • Posts: 156
  • Country: 00
How thermally conductive is potting compound?
« on: January 21, 2013, 09:29:43 pm »
I was wondering how thermally conductive is potting compound, I was planning on doing an outdoor access point installation  (but do not want to spend the outdoor access point money).

The CPU temperature with no heatsink can reach 65-70C. my plan is to install the router in a weatherproof enclosure and then simply have the antenna sticking out of the enclosure, then using power over Ethernet to connect the SP to the switch as well as have it suspended  on a steel cable in order to cover a large area.

I have not worked with potting compound before, but my plan was to remove the router's internal components from it's original case, and then mount it into the weatherproof case. then after that, attaching a small heatsink to the CPU, and wifi radio RF shield (the hot components in the wifi radio are connected to the RF shield using thermal pads).  After doing all of that, I would like to cover those areas of the device in potting compound (about 3-3.5cm thick), and then thick a thermal pad on the top (at which point when the weatherproof enclosure is closed, the enclosure will act as a heatsink where it meets the potting compound.

Only issue I am unsure of is if the potting compound can conduct enough heat away to keep the router's CPU and wifi radio from overheating (I really want to avoid getting an outdoor unit as they tend to cost nearly 5-7 times as much for similar transmit power as a router that is designed for indoor use. (600-650mw)
 

Offline ftransform

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: 00
Re: How thermally conductive is potting compound?
« Reply #1 on: January 21, 2013, 09:35:13 pm »
How about putting it inside of a heat sealed bag instead?

I left a un-sealed 1980's CCTV camera (with .5 cm holes all over the case) hanging under a 6 inch awning outside for over 6 years and it never gave me any problems (even in the rain. snow. wind. cold).

I think wrapping your router in a garbage bag would probably be enough. Ive seen functional setups like that before lol

Put it in a heat sealed bag and then put a harness around it imo. Potting is overkill.
« Last Edit: January 21, 2013, 09:37:51 pm by ftransform »
 

Offline AndyC_772

  • Super Contributor
  • ***
  • Posts: 4228
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: How thermally conductive is potting compound?
« Reply #2 on: January 22, 2013, 08:05:39 am »
What's the massive thickness of potting compound for?

Why not just stick a heat sink on the CPU, then spray the whole board with a thin layer of conformal coating?

Offline ptricks

  • Frequent Contributor
  • **
  • Posts: 671
  • Country: us
Re: How thermally conductive is potting compound?
« Reply #3 on: January 22, 2013, 12:33:21 pm »
Both potting compound and conformal coating are very poor conductors of heat when compared to a heat sink applied directly.
I would mask off an area around the main cpu big enough for the heat sink, apply the coating, then attach the heat sink . Most outdoor mounted boards don't even bother with conformal coating, they simply mount the device in a weather proof enclose.
 

Offline Razor512Topic starter

  • Regular Contributor
  • *
  • Posts: 156
  • Country: 00
Re: How thermally conductive is potting compound?
« Reply #4 on: January 22, 2013, 01:23:43 pm »
my main issue is with somehow getting the heat generated by the wifi radio and CPU outside of the case without opening the internals to the outdoor weather.

I originally wanted to get a chunk of copper or other metal and have it cut to shape to press against the CPU, wifi radio and finally using a thermal pad, against the door of the enclosure to more efficiently get the heat out of the case, but could not find any place that actually sold chunks of any metal cut to a specific size (I like the idea of placing the router in a bag, (especially if I can safely vacuum seal it or just having it in a bag (I have extra heatsinks that I can also attach, just a little worried about  it getting damaged (it will be installed in an outdoor location in Florida which tends to have high temperatures in the summer and lots of rain and wind the rest of the year)

Other than placing the entire unit outside, I am stuck having the business run about 100 feet of coax from the router to the antenna and having just the antenna outside but I feel that will greatly degrade the signal compared to having the entire unit suspended and about 100 feet of ethernet running to the unit.
 

Offline UPI

  • Regular Contributor
  • *
  • Posts: 139
  • Country: us
Re: How thermally conductive is potting compound?
« Reply #5 on: January 22, 2013, 02:25:12 pm »
Thinking out loud here:

Drop it in a sealed bag or case of mineral oil to disperse the heat widely.

Pot the entire bag or case.

 

Offline grumpydoc

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: How thermally conductive is potting compound?
« Reply #6 on: January 22, 2013, 06:51:45 pm »
What about something like this
 

Offline ptricks

  • Frequent Contributor
  • **
  • Posts: 671
  • Country: us
Re: How thermally conductive is potting compound?
« Reply #7 on: January 22, 2013, 09:51:32 pm »
my main issue is with somehow getting the heat generated by the wifi radio and CPU outside of the case without opening the internals to the outdoor weather.

You need to determine how much heat is generated by the devices and what the operating range is for the parts you are using. Chances are you don't need to do anything to dissipate the heat from inside the case, the heat sink and the air inside the casing might be enough.
In the past I have set up things like wifi repeaters using a watertight enclosure and burying the device about 18 inches in the ground, the antenna cable came out the ground and up the pole to the antenna but the device itself stayed at about room temp no matter the weather and I didn't have to use any extra cooling.  This was basically a wifi router with an external antenna buried in a box with a solar panel on the pole for power.  That setup has been running about 6 years now without any maintenance at all. It sits in the middle of a cow pasture and links a farmers home to the dairy behind the home about 1 mile away.


 

Offline Razor512Topic starter

  • Regular Contributor
  • *
  • Posts: 156
  • Country: 00
Re: How thermally conductive is potting compound?
« Reply #8 on: January 23, 2013, 12:33:34 am »
both components reach the mid to upper 60C and the CPU overheat at 85c (not sure about the wifi radio)

And this is with the room temperature at 20C
 

Offline ve7xen

  • Super Contributor
  • ***
  • Posts: 1193
  • Country: ca
    • VE7XEN Blog
Re: How thermally conductive is potting compound?
« Reply #9 on: January 23, 2013, 01:15:41 am »
I really doubt by the time you've built an enclosure, bought conformal coating of some sort, sorted out some sort of PoE or bought long power cables, and figured out how to get cabling into the enclosure that this is going to be any more cost effective, and it'll certainly be a lot more work and much more likely to fail. Products designed for outdoor use are really not all that expensive. E.g. Ubiquiti Picostation M2HP is a 600mW radio, it costs about $80. Edimax, and probably some other vendors have similar products.
73 de VE7XEN
He/Him
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf