Aluminum Oxide vs. Sil Pad Insulators

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
VEry Important question for you who already used this two kind of insulators:

Which of this two are beter to use Aluminim Oxide or Silicon Pads insulators?

I am getting ready to install output transistors on heatsink and after I researched this topic could not make decision.

Aluminum Oxide pads are 2mm thick made for TO-247.

Silicon Pad Insulators I would need to make (cut) from 6" x 6" sheet I have.

Answer from your personall experience ,please!!

Regards
 
Aluminum Oxide has superior conductivity compared to silicone and mica. The only real caveat is the brittle nature of the aluminum oxide. The heatsink needs to be flat and without burrs (Aaron or Raymond or otherwise). You also need to be careful about the torque when you tignten down on aluminum oxide, or you will crack it for sure.
 
zox2003 said:
VEry Important question for you who already used this two kind of insulators:

Which of this two are beter to use Aluminim Oxide or Silicon Pads insulators?

I am getting ready to install output transistors on heatsink and after I researched this topic could not make decision.

Aluminum Oxide pads are 2mm thick made for TO-247.

Silicon Pad Insulators I would need to make (cut) from 6" x 6" sheet I have.

Answer from your personall experience ,please!!

Regards

Use the aluminum oxide. It is better.

John
 
I have problem with understanding this whole temperature thing on Aluminum Oxide PAds.This is a sentence I found about Al Oxide: "Aluminium oxide is an excellent thermal and electrical insulator".

So fact that Aluminum Oxide is "excellent thermal insulator" is making him BAD or GOOD for temperature transfer.

In Amplifier Transistor is source of heat that needs to be transfer to heatsink.Between transistor and heatsink I need to insert Aluminum Oxide which is claimedn as "excellent thermal insulator".How will heat get transfered from transistor to heatsink thru thermal insulator.

Another spec.:The thermal conductivity of aluminum oxide is 15.06 W/mK at 75° (167° F)

?????

PLease help to clear this up!!!!
 
Thermal conductivity of Aluminium oxide is 18W/m/°C.
Thermal conductivity of Aluminium nitride is 180W/m/°C.

But they usually come in 1mm thickness, 0.5mm perhaps if you are lucky. And because they are rigid, you still need a grease (or arctic silver, or Aavid Ultrastick, or .....) to fill the gaps on both sides of the insulation to transistor / heatsink. Which increases the thermal resistance.

You may therefore consider this :

Keratherm 86/90
www.kerafol.com
http://kerafol.de/jml/pdfdocs/therm...0er_serie_e.pdf

Thermal conductivity 10W/m/°C (no big deal).
Standard thickness 0.1mm (hence factor of 5 better than 1mm Al2O3).
It is flexible (to an extent) like Silpad, so no need for additional thermal compound.
Manufacturer: Austerlitz
Aluminium oxide washers have a significantly lower thermal resistance than both mica, kapton and silicone rubber. The insulation voltage is also lower. Thickness: 3 mm for ALO-3 and ALO-P3-3, and 1.5 mm for ALO- 220.
Thermal resistance: <0.3 °/W
Isolation voltage: 30 kV
Temperature resistance: >400 °C
I think these two extracts from previous postings disagree.
Can someone fathom out the data and confirm if they are better than the cheaper Kapton/Mica alternatives.
What thickness of alu oxide causes it to perform worse than 0.002inch Mica?
 
AndrewT said:

I think these two extracts from previous postings disagree.
Can someone fathom out the data and confirm if they are better than the cheaper Kapton/Mica alternatives.
What thickness of alu oxide causes it to perform worse than 0.002inch Mica?

Thermal conductivity, percent of copper:

Copper 100%
Gold 78%
Beryllia 61%
Aluminum 53%
Aluminum nitride 27-42%
Alumina (99%) 9.4%
Kapton, quite low.
Mica, no data on file...sorry

Cheers, John

ps..almost forgot..

To reallly compare the two accurately, you need the dimensions of the silicon die, the die attach thickness (you can assume 3 mils as a default), and the case plate thickness of copper. The interface to the insulators will be the same on the die side of the insulator, but the thicker alumina spreads the heat horizontally more at the heatsing interface, giving an overall advantage to the alumina..simply comparing the total thermal resistance of the material is inadequate, you have to either use a 45 degree spread model, or mount and test the same device with both methods.

The alumina is better than mica/silpad thermally, but I typically use the latter(s).

Cheers, John
 
Still no answer on main question:

"I have problem with understanding this whole temperature thing on Aluminum Oxide PAds.This is a sentence I found about Al Oxide: "Aluminium oxide is an excellent thermal and electrical insulator".

So fact that Aluminum Oxide is "excellent thermal insulator" is making him BAD or GOOD for temperature transfer.

In Amplifier Transistor is source of heat that needs to be transfer to heatsink.Between transistor and heatsink I need to insert Aluminum Oxide which is claimedn as "excellent thermal insulator".How will heat get transfered from transistor to heatsink thru thermal insulator.

Another spec.:The thermal conductivity of aluminum oxide is 15.06 W/mK at 75° (167° F)

?????

PLease help to clear this up!!!!
 
zox2003 said:
Still no answer on main question:

"I have problem with understanding this whole temperature thing on Aluminum Oxide PAds.This is a sentence I found about Al Oxide: "Aluminium oxide is an excellent thermal and electrical insulator".


That statement is incorrect.

Aluminum oxide is an excellent electrical insulator, which also has a very high thermal conductivity (high with respect to other electrical insulators).

Aluminum nitride can be better, and beryllia is much better. But they are expensive, BeO has machining related toxicity issues.

Cheers, John
 
Hi,
from the Wiki Pedia link
Aluminium oxide is an excellent thermal and electrical insulator.
I see the problem you are having.
WikiPedia is wrong.
That's the problem with the open database that can be written and altered by all.

Alu Oxide is an excellent electrical insulator.
Alu Oxide is an excellent thermal conductor.

The difficulty is that as a sheet thermal transfer layer it is usually quite thick, resulting in a significant thermal resistance as the thickness becomes even greater.

I would still like to know what thickness alu oxide can approach before it becomes worse than 0.002inch Mica.

I understand JN and the heat spreader complication, but I would like real numbers from someone that knows.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.