Using Kapton Tape instead of Mica washers

[madswitcher]

Hi All,

has anyone tried using Kapton tape instead of mica washers to insulate power transistors on heatsinks. I am thinking in particular of TO-3 devices? 

Experiences good and bad?

Thanks

Mike

 

[scott2000]

I wonder if it transfers heat well.....I looked real quick and it's resistant but, that really doesn't tell if it's a good conductor maybe????....Thermal conductivity??? Is that something??

 

[JohnRoberts]

Yes thermal resistance is a thing...

I have heard of TO3 insulators made from different material, but never kapton tape.

I just googled it and Kapton has high thermal resistance, the opposite of what you want for a power device heat sink washer,,,

Look for a dedicated heat sink insulation washer that will have low thermal resistance, but high electrical resistance..

JR

 

[madswitcher]

Hi Guys,

thanks for the replies.  I was of the same opinion as yourselves on this as I normally use the standard mica washer for this job and wondered about its thermal conductivity, but looking on Rod Elliot's site, he suggested using Kapton instead as long as it was of 25um thickness.  This was for his 25 watt amp in the 'Construction 'section via the link below and I thought this might be a new idea to try out. Maybe some experimentation is called  for. 

http://sound.whsites.net/project3b.htm

Cheers
Mike

 

[Newmarket]

I guess the idea is that although it has a higher thermal resistance per unit thickness, it is very thin to more than compensate ?
You'll need a smooth heatsink surface  - extruded not stamped/pressed to avoid risk puncturing the thin film.
I use Polyimide (Kapton) tape a lot in the day job but for isolating heatsink interfacing it's Laird TGARD500.
There are also interesting types from Bergquist etc that require compression to be effective but are then very good.
It does sound as if forced air cooling might be required ?
Water cooling is great - and quiet. I've used it in scientific instrumentation previously.
However it can be complicated to implement if you go all out for it - Water chiller / pump etc...

 

[JohnRoberts]

I had to search polyimide because kapton is a trademark name... yes apparently they make some heatsink washers from it.

Back in the day I only used mica washers for TO3, but those (TO3) are rarely used anymore.

JR

 

[PRR]

I have seen Kapton tape used under TO-126 devices in a mil-grade amplifier in the 25W range.

I saw it in "surplus" so maybe it was rejected for dubious insulation?

In general if you are not working the transistors at a large fraction of their ideal Pdiss, the insulator is not a primary bottleneck.

These days, if you have to ask, it is perhaps wise to look for Fully Insulated Package. Make the manufacturer responsible instead of trusting the technician. (If you know your technician is you and you are more obsessive than the manufacturer, have at it any way you like.)

 

[Gene Pink]

... yes apparently they make some heatsink washers from it.

From left to right:

Mica .002" (have .005" also)
Sil-pad .007"
Kapton (I think) .002"
Beryllium oxide ceramic .062"

Gene

 

[JohnRoberts]

I have seen Kapton tape used under TO-126 devices in a mil-grade amplifier in the 25W range.

I saw it in "surplus" so maybe it was rejected for dubious insulation?

In general if you are not working the transistors at a large fraction of their ideal Pdiss, the insulator is not a primary bottleneck.

I spent most of my engineering career worrying about every last 0.1'C/W of thermal resistance (junction to sink or ambient). Even subtle things like 2 screw TO3 packages vs one screw plastic packages (yes everything makes a difference in the margin). It was only near the end of my amp design involvement when spring clips replaced screws, and plastic displaced metal power device packaging in wide use.

Lots of our design rules of thumb suggested that amps using plastic power devices would need 50% more devices to handle the same power dissipation as metal devices. History suggests that these design rules were too conservative, because plastic amps have demonstrated excellent reliability.

Modern class D power amps reduce the massive dissipation, but still must manage some power dissipation. Since the power is not spread across a dozen devices, but one or two the thermal resistance matters. I recall one semi-custom device development program we tried at Peavey (last century), two put two switching devices in a common package. The program failed miserably because the thermal resistance junction to case was inferior to using separate devices. 

These days, if you have to ask, it is perhaps wise to look for Fully Insulated Package. Make the manufacturer responsible instead of trusting the technician. (If you know your technician is you and you are more obsessive than the manufacturer, have at it any way you like.)

Yup... in earlier times device heatsinks were always electrically hot, causing all kinds of problems from insulation failures.  When working on legacy gear, or in doubt, ASSume the heatsink is electrically hot.

JR

 

[rackmonkey]

Useless fact about the thermal qualities of Kapton: the insulation around the Apollo era lunar lander was made from layers of kapton sheet over Mylar. Just a few mil of that sandwich was all there was between the sensitive equipment inside and 100 degrees C day and -173 degrees C at night. It’s the orange-y foil-looking stuff you see in pictures of the lander.

No charge for that one.

 

[madswitcher]

Useless fact about the thermal qualities of Kapton: the insulation around the Apollo era lunar lander was made from layers of kapton sheet over Mylar. Just a few mil of that sandwich was all there was between the sensitive equipment inside and 100 degrees C day and -173 degrees C at night. It’s the orange-y foil-looking stuff you see in pictures of the lander.

No charge for that one.

...... and thanks to the other folks for their comments.  I might have a play with both, providing I can get some genuine Kapton tape and not some eastern knock-off.

Best regards

Mike

 

[JohnRoberts]

Useless fact about the thermal qualities of Kapton: the insulation around the Apollo era lunar lander was made from layers of kapton sheet over Mylar. Just a few mil of that sandwich was all there was between the sensitive equipment inside and 100 degrees C day and -173 degrees C at night. It’s the orange-y foil-looking stuff you see in pictures of the lander.

No charge for that one.

That suggests a high thermal resistance, exactly the opposite of what we want for a heat sink insulator.

We want the washer to effectively conduct device heat to the heat sink, not insulate it.

Kapton is the trademark/brand name and there are likely different flavors of polyimide.

JR