Don, and all,
while it's very true that a THIN layer of thermal compound is desirable,
it's ESSENTIAL for the layer to be continuous, that is, to avoid having
air pockets there. And the techniques to achieve these goals vary.
Most people recommend applying a very thin layer, spread evenly over the
mounting surface. But I happen to disagree with that method! The fact is
that it's impossible to apply a perfectly even and flat layer. So, at
the moment of contact when you mount the part, some areas will touch
while some others will remain with air between them. When you tighten
down the device, hopefully this air will be expelled and the tiny space
between the device and the heatsink will completely fill up with thermal
compound. But if the applied layer is too thin, this might not happen,
and you can end up with air trapped under your expensive power transistor!
For that reason I use a radically different technique: I apply a single
thick round blob of thermal compound to the middle of the mounting
surface, or if the device is long, I apply a single nice smooth string
of compound to it. Then I mount it, pressing it down strongly but
evenly, to make the compound flow out and fill the whole space. In this
way the formation of air pockets is avoided. When thermal compound
appears squeezed out all around the device, I can be sure that no air
remains anywhere under the part. From this point on, additional pressure
can be used to thin out the layer of thermal compound. This pressure
should be well distributed over the device, to avoid deforming it. Just
tightening the mounting screws is not a good idea. A clamp in the middle
of the device is ideal.
If necessary, the device and heatsink can be warmed up to soften the
thermal compound and make it flow out more easily. Otherwise it's a good
idea to re-tighten the mounting screws during the first test run, when
the equipment and specially the power devices get warm.
Since I'm a hopeless cheapskate, I scoop up the excess thermal compound
that squeezes out, and use it for the next device.
This method has worked well for me, in 35 years building electronic
equipment, and is easy and quick.
Manfred
========================
Visit my hobby homepage!
http://ludens.cl
========================
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|