Heatsink compound

Keeping output transistors (BJT or MOSFET) cool can be a difficult task. The goal is to allow the heat to flow away from the device. Anything that slows down the heatflow is bad.

With that in mind, has anybody tried the high-tech heatsink compounds used in PCs? The one that comes to mind is Artic Silver II. If this is a good as they say it is, it should be able to drop the junction temperature by several degrees. For someone running a Class-A amp, this could go a long way to keeping the transistors cool. It's also a lot cheaper than a BIG heatsink.
 
AudioFreak said:
Arctic Silver is very good ... just be careful it is electrically conductive :)

But your still gonna need that BIG heatsink.

The manufacturer claims that is isn't conductive, but that it can cause capacitance issues. If you're using a mica with TO-247 or TO-220 cases, this doesn't come into play.

I was thinking that it might make a heatsink that's a little too small be usable. Alternatively, it could lower the junction temperature of the output transistors, allowing them to pump out more power, or just last longer.

Thanks.
 
As soon as you put an isolation wafer between it doesnt matter much what kind of thermal grease you use, most of the resistance lies in the isolation wafer. I wouldnt trust manufacturers sales pitch to much...

Best you can get lies around 0.1 C/W without any isolation wafers(of any sort) I use Al2O3 if I can.

Most PC-processor heatsinsk are fan-cooled today (my old P-133 doesnt use fan) , that is the only way to get rid of 40 W in such a small place. And everybody now how much a Pc- sounds.
They arent that cheap, considering they are only designed for a load around 40 W so you need many of them to cool a class-A!

I d rather put a silent(expensive) fan on a ordinary heatsink , and lower the thermal resistance with a little as maybe 25 % but it can go as low as a third of the ordinary thermal resistance.

/rickard
 
Arctic silver is advertised as conductive "under pressure" and "in a thin layer." It was not ever intended for use in a situation where conduction was critical, so there isn't a great deal of information available, although the website does have some figures.

I'll include my question and their reply:
=============================

Tyler,

There is a good chance that the AS2 WILL be conductive under those
conditions. Whether this is critical given the mica washer is something you
will have to determine.

We also are releasing an innovative new ceramic based thermal compound that
is an electrical insulator. This might suit your purposes bettter;

Here is some additional information on the new compound including the early
reviews...

http://www.arcticsilver.com/arctic_alumina.htm

On, and I'm sorry about taking so long to reply to your inquiry. Remind me
never to release 2 new products in the same month again!

Nevin House
Arctic Silver, Inc.

At 05:16 AM 10/26/01, you wrote:
>I'm writing in the hopes that someone at your company will nice enough to
>answer a couple of questions for me.
>
>I'm interested in using your product to interface transistors to a heatsink
>in an audio amplifier. However, the electrical conductivity seems to be a
>bit of an unknown; the voltages from die to ground will be significant
>(42V) compared to electrical logic voltages so this is a concern. Because
>of the machining of my parts the layer would likely be ~0.005" thick and
>under firm but not extreme compression. I will be using a mica washer and
>would like to use Arctic Silver II on both the die-to-washer surface and the
>washer-to sink surface.
>
>===DIE===
>Artic Silver II
>===Mica===
>Artic Silver II
>===Sink===
 
Hmmm...sounds like this would be of use for, say, a front-end transistor where you could dedicate a heatsink to that one transistor and simply bypass insulators and all concerns about the heatsink going live.
The 'conductive under pressure' part worries me for any other application, though. I am assuming (perhaps incorrectly, granted) that if this stuff is conductive under pressure that it will also be conductive when <i>not</i> under pressure, although presumably to a lesser extent. Supposing that you were to use it with a mica insulator and some slopped over the edge to contact the heatsink directly. Would this lead to a high-resistance (hundreds of k to perhaps M ohm range) connection? I can see where you might have to be scrupulous about applying the compound; not allowing any to come into contact with the heatsink. High PIA factor if that's the case.

Grey