Page 1 of 1

Thermally conductive Polyimide (Kapton)

PostPosted: Wed Feb 27, 2008 8:18 pm
by mdburke-K9MB
Hi,
Thought I would pass along some info I just got today.
Just doing some research on Polyimide (Kapton) and found some interesting bits of information on the Dupont site
I was not aware of it, but Dupont makes umpteen versions of Kapton film for a lot of purposes.
Not sure what the guys are using, but they make a special version that has almost three times the thermal conductivity of standard light amber Kapton. The thermally conductive stuff is either the "MT" or "MTB" versions.
I have links below to Dupont's site and a link to the MT stuff

http://www2.dupont.com/Kapton/en_US/tec ... index.html

http://www2.dupont.com/Kapton/en_US/ass ... 8497-1.pdf

McMaster Carr Supply company sells a 2mil Kapton film that is thermally conductive. Since the stuff has a dielectric strength of about 3000volt/mil at 300degrees, the 2 mil stuff should work fine unless you have some burrs on your spreader plates.
You can search for Kapton Film at:
http://www.mcmaster.com/
Not cheap, about $12 for a 12" X12" piece, but at least you don't have to buy a ton of it.
Maybe this is an old topic, but Ithought I read that people were buying the stuff on EBAY and the stuff I saw there was the standard stuff and it has a thermal resistance almost 3 times higher than the MT or MTB stuff and that could be critical if you are thinking about efficient thermal transfer, specially in linear applications.
73,Mike

Re: Thermally conductive Polyimide (Kapton)

PostPosted: Wed Feb 27, 2008 9:22 pm
by frank carcia
Mike,
I was on that site a couple months ago and found I used the wrong dielectric constant when I did my spreaders which explained why I had to add 1000 pf per phase. I'm just using the normal stuff 3 mils thick. My transformer cores got warmer than the FETs at full strap so the thermal transfer must be fine with all that surface area and copper spreader.
I work with a real good thermal guy who laughed at me when I showed him one of my new paper designs. He is quite conservative with his designs but after I told him I was going to use 1/4 inch copper he just shook his head.
I think the amber stuff is 7000 volts/mil.

Re: Thermally conductive Polyimide (Kapton)

PostPosted: Wed Feb 27, 2008 10:41 pm
by mdburke-K9MB
Hi Frank,
I didn't know if it made any difference or not. I have not checked the difference in the dielectric constant, but I think they are both about 4.3, so it must be a function of the thickness, which determined capacitance.
As you say, you are not even close to overheating those FETS in Class E, but I had wondered if someone decided to positive bias them and use them as a Class AB2 or class B push pull linear amplifier. I suppose that there would be a lot more heat there than in Class E so the heat transfer gradient would become important. There must be some applications where it is important, or they would not make and advertise the stuff.
You certainly have a lot of thermal mass there and physical mass too, I guess with quarter inch thick copper. Probably looks like a black hole to any thermal activity. :)
Mike

Re: Thermally conductive Polyimide (Kapton)

PostPosted: Wed Feb 27, 2008 11:34 pm
by kf1z
I got the stuff from ebay.... being sold there for audio amplifier uses.
For my rig, works fine... of course as you say..class-e may not matter...

My copper heat-spreaders never more than warm.......and the heat-sinks not far behind.


I stay clear of McMaster Carr...sure, they have a lot.... but they sure charge a lot too!!

Re: Thermally conductive Polyimide (Kapton)

PostPosted: Thu Feb 28, 2008 12:07 am
by mdburke-K9MB
Hi Bruce,
I buy a lot of a few items from McMaster Carr for my second business, making Irish Pennywhistles (http://www.burkewhistles.com)
Mostly, I buy O'rings from them, but some other odds and ends like silicone for sealing, etc. I bought O'rings from a distributor online and the prices were not too much different from the prices at McMaster Carr. I buy mill runs of aluminum and brass tube and no way would I buy an inch from McMaster Carr, so it really depends on what you need. They are not bad for prototyping, because we get stuff the next day UPS ground, but for production quantities of most things, they won't budge on pricing. Same is true with Digikey. For small amounts, they are fine, but if I want to buy thousands of a semiconductor, I go to Arrow, or Avnet or Future and for passives, RS Electronics and Newark is pretty good in larger amounts, but they are really expensive on small lots and I use Digikey and Mouser for those.
You are right, you simply cannot beat the pure capitalism of EBAY for getting a good deal, unless it is a flea market at Dayton or something like that where you can bargain. There is some talk about the government getting in and taxing internet purchases. That will screw it up because the people that have businesses, are small time outfits and obviously operate on a shoestring many times. There are some bigger outfits, but most of the deals are from a small business or an individual
Well, to bed.
This discussion is fascinating. I am graduating to a higher level of ignorance after a couple of days here. :)
Good night.
Mike