On Sat, 31 Mar 2007 05:57:30 -0500, davy
<davy.2oc3oh@no.email.invalid> wrote:

>
>Speaking from experience I have seen TV's and video's where the failure
>of power transistors and FETs in the switchmode supply due to the
>actual white stuff from drying out...

While it could "somewhat" be true, not directly so.
If the grease is merely average quality and the part being
'sunk isn't overdriven onto the point of ultimately drying
out the grease because it was excessively hot all along, it
will be sufficient for the life of the product and even 20
years longer. In other words, a spot of grease can't make
up for poor engineering or later penny-pinching in a design,
nor can a better grease than that "white stuff".


>infact one manufacturers
>modification was to remove this white stuff and replace it with their
>recommendations (usually their own stuff costing that little bit more)..
>I literally had to scrub the old white stuff off from Tv's and video's
>less than 12 months old. When things are under warrenty you have to
>repair them to the manufacturers specs or they don't wanna know should
>they have to have it back.

It is not an indictment of "white stuff" in general,
certainly some of it is really poor, but that doesn't weight
against the average grade compound.



>
>Common wheel bearing grease.. well thats new just remember there's
>wheel grease and wheel grease but you don't say what brand it is...

Right, because it doesn't matter. Synthetic based wheel
bearing grease is essentially similar formula as the
infamous arctic silver, except the particles are not
abrasive and are moderately larger. It is beside the point
though, that there is no reason to use it unless one were
stranded on a deserted island with everything (computer,
power, etc,) but thermal grease - an unlikely scenario.
Once the thought of vaseline came up it was just a silly
daydream thought, not a serious use.



>
>but grease also forms a film and why don't all manufacturers use it...
>there's a heck a lot if it about, tonnes of it..?

You seem to suggest a "film" is some bad thing. It is the
goal. However, the surfaces are typically nonplaner enough
that the gaps in many areas are larger than one particle in
heatsink compound, which is essentially a thick oil or
grease, with the solids added to be more thermally
conductive than the "grease" carrier alone would be. Plus,
adding the solids reduces the pumping of the carrier by not
only making it more viscous, but by the greater thermal
conduction reducing the interface temp.

>
>This heat sink stuff contains oxides which comes in all proportions as
>to brand... you gotta admit that the white stuff is far more thicker
>than the stuff you see on CPUs and takes more pressure to spread.

If you have gotten ahold of a product that is too viscous
for your particular use it is not at all indicative of all
"heat stink stuff". It's not as though there is only one
vat of the stuff from one manufacturer, just as there are
variations of many many different products, not all
so-called generic white thermal compound is a reasonable
quality. If you bought a poor quality car, do you swear off
all cars and buy a horse?



>
>All we are interested in is filling the gaps the mating surface with a
>heat conducting agent such as a metal oxide.


Yes, which any normal quality white grease will do fine.
There is a minor gain in some exotic compositions, and in
very high heat density situations there is a further gain in
going with a synthetic base oil/grease instead of silicone.
That situation is not present on today's CPUs because they
have heat spreaders. The thing to avoid is any very low
quality grease, or any very low quality factory applied
thermal interface material, or any very low quality
heatsink, motherboard, CPU, etc, etc.