Thermal paste or Sudocrem?
R D S
essential, I came across this,
http://63.74.115.230/tips1214/
[quote]Ambient temp was roughly 23ºC for these tests, running Prime 95 for
five hours:
a.. Arctic Silver 5, fully burned in: 43~45ºC
b.. Toothpaste + Vaseline: 46~49ºC
c.. Desitin Diaper Rash Cream: 46~48ºC [/quote]
Now I do have Sudocrem so I will be adding to it's already widespread uses.
Any opinions?
Rick
Mike Tomlinson
writes
>Now I do have Sudocrem so I will be adding to it's already widespread uses.
>
>Any opinions?
It's not such a crazy idea as it first appears, but I'd recommend
getting the proper stuff.
Alternatives may work well at first but will dry out, leaving you with a
powder doing the job of heat transfer instead of the intended-for-
purpose gloop, which will stay fluid for longer.
Synapse Syndrome
>
> Poking about the net wondering whether thermal paste (I have none handy)
> was essential, I came across this,
>
> http://63.74.115.230/tips1214/
>
> five hours:
> a.. Arctic Silver 5, fully burned in: 43~45C
> b.. Toothpaste + Vaseline: 46~49C
> c.. Desitin Diaper Rash Cream: 46~48C [/quote]
> Now I do have Sudocrem so I will be adding to it's already widespread
> uses.
>
> Any opinions?
No doubt Sudocrem is mostly made from water (aqua), and so will not have the
same thermal properties within minutes of it being used in this way, or am I
stating the bleeding obvious like some sort of fucking idiot.
ss.
Roger Hunt
Actually, the application instructions are so close to those of HS paste
they could be the same thing :
"The general rule is to use a small amount and apply in a thin layer.
Massage in small circular movements until the Sudocrem has disappeared
leaving a translucent film. If this does not cover the affected area
apply a little more. However, if the area stays white after application,
too much has been applied"
<http://www.sudocrem.co.uk/using-sudocrem.html>
--
Roger Hunt
Daniel James
> Any opinions?
Yeah, don't.
Things like toothpaste and nappy cream are water-based, and will dry
out. Toothpaste is a particularly bad idea as it contains an abrasive
powder, which will get between the CPU and heatsink and prevent them
making direct contact -- the reason it works well when freshly applied
is that water is a good heat conductor (until it evaporates). Something
with no solids -- like, say, Mayonnaise -- would probably be even
better, for a short while.
Vaseline becomes much more fluid at high temperatures, and will seep
out -- that might be OK while the CPU is hot but when you turn the
system off and it cools down the CPU and heatsink may deform slightly
and break the thermal contact. When you turn it on again there may no
longer be enough vaseline in place and the CPU may overheat.
What probably would work as well as the likes of Arctic Silver is
high-temperature silicone grease -- the stuff they use for sealing
joints in lab glassware -- it should work as well as vaseline but not
change consistency so much with heat. Probably cost as much as Arctic
Silver, though ...
Cheers,
Daniel.
TMack
It al depends on whether you intend to use it on your processor or on your
piles.
--
Tony
'04 Ducati ST3, '08 DL650GT, '95 LS650, OMF#24
Roger Hunt
<tonyREM...@REMOVECAPS.dsl.pipex.com> writes
--
Roger Hunt
TMack
The "fundamental" choice was between thermal paste or Sudocrem
bill
"TMack" <tonyREM...@REMOVECAPS.dsl.pipex.com> wrote in message
news:uNydnQ5gMacaBd_U...@pipex.net...
unnecessarily expensive way of treating them though (unless one had a chip up
ones anus of course)
Roger Hunt
<tonyREM...@REMOVECAPS.dsl.pipex.com> writes
--
Roger Hunt
PeterC
OK for redheads.
--
Peter.
You don't understand Newton's Third Law of Motion?
It's not rocket science, you know.
R D S
"Roger Hunt" <nos...@nospam.demon.co.uk> wrote in message
news:1pwoPIA5...@nospam.demon.co.uk...
Don't apply it with a toothbrush.
Trev
<rsa...@gmail.com> R D S bashed on keyboard and typed:
And if you do dont use your own brush
--
Trev
Nobody is perfect.
But Being a Yorkshire man is as close as you can get.
Roger Hunt
nAT.dsl.pipex.COM@?.?> writes
--
Roger Hunt
Steve Walker
The Ring Of Confidence.
Roger Hunt
tr...@beeb.net> writes
Athlon X2 5000+ Low power, with P95 running 24/7/365 (both cores of
course), and Silver-based HS paste : 32ºC CPU temp. Oh no - 31º!
Is that Cool or what?
--
Roger Hunt
Johnny B Good
from Roger Hunt <nos...@nospam.demon.co.uk> contains these words:
Well, here's a novel thought, how about using _NO_ heatsink paste?
Looking at that table of thermal tests involving items a, b and c, I'm
rather struck by the absence of an option d in which no thermal paste is
used. Unless the heatsink contact surface is severely rough or bowed,
I'm willing to bet it won't be any worse than the arctic silver
alternatives used in options b and c (and possibly better, perhaps even
a near match for the arctic silver).
--
Regards, John.
Please remove the "ohggcyht" before replying.
The address has been munged to reject Spam-bots.
Roger Hunt
<jcs.comp...@plugzetnet.co.uk> writes
heat spreader, rather than the bare chip (like with my old XP 2800+).
Without HS paste the heat has to travel through the many points of
contact between CPU & heatsink. With a thin smear of HS paste the heat
travels through the points of contact and through the paste as well.
It can only be better with the HS paste surely?
--
Roger Hunt
Mike Tomlinson
<nos...@nospam.demon.co.uk> writes
>Athlon X2 5000+ Low power, with P95 running 24/7/365 (both cores of
>course), and Silver-based HS paste : 32ºC CPU temp. Oh no - 31º!
That is very good! Do you use CnQ?
--
(\__/)
(='.'=) Bunny says Windows 7 is Vi$ta reloaded.
(")_(")
Roger Hunt
<mi...@jasper.org.uk> writes
>In article <oemfiVAj...@nospam.demon.co.uk>, Roger Hunt
><nos...@nospam.demon.co.uk> writes
>
>>Athlon X2 5000+ Low power, with P95 running 24/7/365 (both cores of
>>course), and Silver-based HS paste : 32ºC CPU temp. Oh no - 31º!
>
>That is very good! Do you use CnQ?
>
This is just a standard uninteresting XP install on a Biostar TA-690G,
with the processor above, and I rather like it.
--
Roger Hunt
johannes
Well, Intel officially recommend replacing the thermal paste at regular
intervals. Imagine what 'Computer Maintenance' would be like. Not only
upgrading various software and defragmenting disks, IT staff also have
to go round ripping out every CPU of every office computer and renew the
thermal paste. LOL!
Mike Tomlinson
<nos...@nospam.demon.co.uk> writes
>Cool & Quiet? No, never tried it. (Don't think I need it!)
Well, no with those temperatures you don't. I'm using an older X2
4800+, but I think this is the 95W one, so CnQ helps to keep the machine
a bit cooler and, um, quieter. CnQ was a disaster on Win2k, it made the
machine very unstable, but it's been fine since an upgrade to eXpee.
The mobo temp widget says my cpu's running at 31C, bit it's not doing
much. Like you, I prefer to leave the machine on 24/7.
Roger Hunt
<mi...@jasper.org.uk> writes
>In article <0Ah4VIA5...@nospam.demon.co.uk>, Roger Hunt
><nos...@nospam.demon.co.uk> writes
>
>>Cool & Quiet? No, never tried it. (Don't think I need it!)
>
>Well, no with those temperatures you don't. I'm using an older X2
>4800+, but I think this is the 95W one, so CnQ helps to keep the machine
>a bit cooler and, um, quieter. CnQ was a disaster on Win2k, it made the
>machine very unstable, but it's been fine since an upgrade to eXpee.
>
years. I'm half tempted to install it ... but maybe I'm not that much of
a masochist.
>The mobo temp widget says my cpu's running at 31C, bit it's not doing
>much. Like you, I prefer to leave the machine on 24/7.
>
--
Roger Hunt
Synapse Syndrome
>>
>> Well, here's a novel thought, how about using _NO_ heatsink paste?
>> Looking at that table of thermal tests involving items a, b and c, I'm
>> rather struck by the absence of an option d in which no thermal paste is
>> used. Unless the heatsink contact surface is severely rough or bowed,
>> I'm willing to bet it won't be any worse than the arctic silver
>> alternatives used in options b and c (and possibly better, perhaps even
>> a near match for the arctic silver).
>>
> Um ... I always use heatsink paste, and very much prefer a CPU with a
> heat spreader, rather than the bare chip (like with my old XP 2800+).
>
> Without HS paste the heat has to travel through the many points of
> contact between CPU & heatsink. With a thin smear of HS paste the heat
> travels through the points of contact and through the paste as well.
> It can only be better with the HS paste surely?
Extreme overclockers only use the tiniest amount of paste, if at all. They
polish the surfaces to a flat mirror finish - it's called 'lapping'.
http://www.kustompcs.co.uk/acatalog/info_3719.html
http://www.overclockersclub.com/guides/lapping/
ss.
Roger Hunt
<syn...@NOSPAMsyndrome.me.uk> writes
it to the nth degree, to almost atomic precision.
Those lapping kits look quite neat actually - a scaled-down version of
what I have used with cylinder heads. I'm also aware of the crucial need
not to press, or allow any uneven forces while doing the job. I'm sure
brilliant results can be achieved by going slow and steady and gentle.
--
Roger Hunt
PeterC
>> Without HS paste the heat has to travel through the many points of
>> contact between CPU & heatsink. With a thin smear of HS paste the heat
>> travels through the points of contact and through the paste as well.
>> It can only be better with the HS paste surely?
>
> Extreme overclockers only use the tiniest amount of paste, if at all. They
> polish the surfaces to a flat mirror finish - it's called 'lapping'.
Even if carried out by a method that polishes the surface and ensures that
bot surfaces are 'perfectly' flat, only about 10% contact is made when
they're pressed together. The finer the surfaces the more important it is
to avoid exessive past.
I like the Zalman stuff in the bottle as it's easy to put on a small amount
well spread and it's thin enough to be easily displaced.
bill
"PeterC" <giraffe...@homecall.co.uk> wrote in message
news:181ah1eqfhc4g.q5838zid18uq$.dlg@40tude.net...
> On Sat, 13 Dec 2008 19:40:32 -0000, Synapse Syndrome wrote:
>
>>> Without HS paste the heat has to travel through the many points of
>>> contact between CPU & heatsink. With a thin smear of HS paste the heat
>>> travels through the points of contact and through the paste as well.
>>> It can only be better with the HS paste surely?
>>
>> Extreme overclockers only use the tiniest amount of paste, if at all. They
>> polish the surfaces to a flat mirror finish - it's called 'lapping'.
>
> Even if carried out by a method that polishes the surface and ensures that
> bot surfaces are 'perfectly' flat, only about 10% contact is made when
> they're pressed together.
10%? You'd need access to atomic force microscopy facilities to determine that.
Johnny B Good
from Roger Hunt <nos...@nospam.demon.co.uk> contains these words:
> In article <3130303037373...@plugzetnet.co.uk>, Johnny B Good
> <jcs.comp...@plugzetnet.co.uk> writes
> >The message <oemfiVAj...@nospam.demon.co.uk>
> >from Roger Hunt <nos...@nospam.demon.co.uk> contains these words:
> >> In article <6qhv1tF...@mid.individual.net>, Steve Walker <spam-
> >> tr...@beeb.net> writes
> >> >Roger Hunt wrote:
> >> >> In article <P5SdnTSXrt8oDN_U...@pipex.net>, TMack
> >> >> <tonyREM...@REMOVECAPS.dsl.pipex.com> writes
> >> >>> R D S wrote:
====snip====
> > Well, here's a novel thought, how about using _NO_ heatsink paste?
> >Looking at that table of thermal tests involving items a, b and c, I'm
> >rather struck by the absence of an option d in which no thermal paste is
> >used. Unless the heatsink contact surface is severely rough or bowed,
> >I'm willing to bet it won't be any worse than the arctic silver
> >alternatives used in options b and c (and possibly better, perhaps even
> >a near match for the arctic silver).
> >
> Um ... I always use heatsink paste, and very much prefer a CPU with a
> heat spreader, rather than the bare chip (like with my old XP 2800+).
> Without HS paste the heat has to travel through the many points of
> contact between CPU & heatsink. With a thin smear of HS paste the heat
> travels through the points of contact and through the paste as well.
> It can only be better with the HS paste surely?
That's the theory and it certainly made a significant improvement when
it was first used with the TO3 packaged power transistors with their
rather less than dead flat heatsink mating surfaces, especially so when
the case had to be insulated with a mica shim between it and the equally
rough heatsink surface (the microvoids were a lot less 'micro' then ;-).
Correctly applied, suitable heatsink paste _will_ reduce the thermal
resistance between the die and the heatsink by replacing the micro air
voids with something more conductive than air. However, with modern
finely finished heatsink surfaces and almost truly 'dead flat' heat
spreaders, this effect is less notable today.
The effect of 'dry mounting' the heatsink may represent only a 1 or 2
deg worsening of performance over the use of properly applied heatsink
paste. If you don't happen to have suitable heatsink paste to hand and
the heatsink assembly wasn't pre-coated with a thin layer of paste[1]
protected by a stiff celluloid cover, you'd be better off dry mounting
it rather than contaminate it with the rather unsuitable alternatives
previously suggested.
[1] Paste, mind you, not that horrible thermal gasket they used to stick
onto the heatsink base. If your brand new heatsink has been cursed with
such a 'thermal pad' gasket, you really do need to carefully scrape it
off and use proper heatsink paste instead (or dry mount if needs must).
treacle
I thought most enthusiast homebuilders had one in the cellar!
Roger Hunt
<tre...@gmail.com> writes
--
Roger Hunt
bill
Well I've got a 14 TEV hadron collider in my bedroom so there's no room for an
AFM.
Roger Hunt
<bcoo...@nospamorange.net> writes
--
Roger Hunt
bill
Roger Hunt
(I must stop this now. This minute.)
--
Roger Hunt
Clint Sharp
<syn...@NOSPAMsyndrome.me.uk> writes
>Extreme overclockers only use the tiniest amount of paste, if at all.
worse thermal conductivity than the heatsink or CPU, it's only there to
fill the voids (which lapping reduces but doesn't eliminate) where you
don't have chip to heatsink contact. In fact, too much heatsink compound
can be worse than none.
--
Clint Sharp
Clint Sharp
<nos...@nospam.demon.co.uk> writes
>>The "fundamental" choice was between thermal paste or Sudocrem
>>
>I'm buggered if I'm putting thermal paste on my fundament.
I'm guessing the thermal transfer properties might help if you had just
been?
--
Clint Sharp
Mike Tomlinson
<cl...@clintsmc.demon.co.uk> writes
>In fact, too much heatsink compound
>can be worse than none.
I'm not convinced. I think that the pressure of the heatsink on the die
would tend to force excess compound out of the joint, especially after
the "pumping" action induced by a few heat/cool cycles.
Roger Hunt
>In fact, too much heatsink compound
>can be worse than none.
>
I know - it's all over my underwear now.
--
Roger Hunt
Alex Fraser
> Athlon X2 5000+ Low power, with P95 running 24/7/365 (both cores of
> course), and Silver-based HS paste : 32ºC CPU temp. Oh no - 31º!
That is certainly pretty cool - what sort of cooling arangements and
ambient temperature?
Alex
Roger Hunt
Fraser <m...@privacy.net> writes
ambient is (lets thermometer settle) 19ºC/66ºF. CPU is still 32ºC.
(The room temp is on the cool side but I like it that way.)
The wonder is with the Low Power of the CPU - 65W I think it is, and
the (argh! caught finger in fan) heatsink is only warm.
Mind you, I know there is no build-up of dust anywhere (yet).
--
Roger Hunt
Tony Houghton
The last thing you want is a sticky white mess in your underpants,
especially at the rear.
--
TH * http://www.realh.co.uk
Roger Hunt
Mike Tomlinson
<nos...@nospam.demon.co.uk> writes
>The wonder is with the Low Power of the CPU - 65W I think it is, and
>the (argh! caught finger in fan)
Makes you jump, doesn't it? I put a grille on my fan after nearly
losing my fingertip (AMD retail heatsink with heatpipes)
PeterC
For a standard 120mm fan, say 2000rpm, it will sting a bit but will do more
damage to the fan. We had Papst fans at work (suitable for 160 - +100C) and
we were told that they wrere dangerous and need guards. I tested one with a
bit of rolled up paper then with my finger then showed people. After the
3rd. or 4th. tim it wouldn't start again. I'd guess that just once would
damage those tiny bearings.
So yes, you do need a fan guard, not a finger guard.
Rob Morley
PeterC <giraffe...@homecall.co.uk> wrote:
> On Mon, 15 Dec 2008 10:38:56 +0000, Mike Tomlinson wrote:
>
> > In article <Bn6FsWAW...@nospam.demon.co.uk>, Roger Hunt
> > <nos...@nospam.demon.co.uk> writes
> >
> >>The wonder is with the Low Power of the CPU - 65W I think it is,
> >>and the (argh! caught finger in fan)
> >
> > Makes you jump, doesn't it? I put a grille on my fan after nearly
> > losing my fingertip (AMD retail heatsink with heatpipes)
>
> For a standard 120mm fan, say 2000rpm, it will sting a bit but will
> do more damage to the fan. We had Papst fans at work (suitable for
> 160 - +100C) and we were told that they wrere dangerous and need
> guards. I tested one with a bit of rolled up paper then with my
> finger then showed people. After the 3rd. or 4th. tim it wouldn't
> start again. I'd guess that just once would damage those tiny
> bearings. So yes, you do need a fan guard, not a finger guard.
I accidentally put a finger in a high power 80mm fan. Several blades
broke off and a sharp edge cut my finger, but it would take a really
anal person to put it in the accident book.
John Jordan
>
> For a standard 120mm fan, say 2000rpm, it will sting a bit but will do more
> damage to the fan. We had Papst fans at work (suitable for 160 - +100C) and
> we were told that they wrere dangerous and need guards. I tested one with a
> bit of rolled up paper then with my finger then showed people. After the
> 3rd. or 4th. tim it wouldn't start again. I'd guess that just once would
> damage those tiny bearings.
> So yes, you do need a fan guard, not a finger guard.
Hmm. I've had substantial finger gouges from both ~2500rpm 80mm fans and
~5000rpm 60mm fans. I wouldn't push this as general advice.
The fans were fine afterwards :-)
--
John Jordan
Clint Sharp
<mi...@jasper.org.uk> writes
>In article <UGYKs9D9...@clintsmc.demon.co.uk>, Clint Sharp
><cl...@clintsmc.demon.co.uk> writes
>
>>In fact, too much heatsink compound
>>can be worse than none.
>
>I'm not convinced. I think that the pressure of the heatsink on the die
>would tend to force excess compound out of the joint, especially after
>the "pumping" action induced by a few heat/cool cycles.
that were stuffed because the silicone oil from the compound had
migrated into the CPU socket contacts.
--
Clint Sharp
PeterC
But you gave up - try 4 or 5 times. Only give up when you run out of
fingers.
Roger Hunt
<giraffe...@homecall.co.uk> writes
>On Tue, 16 Dec 2008 00:29:18 +0000, John Jordan wrote:
>> PeterC wrote:
>>>
>>> For a standard 120mm fan, say 2000rpm, it will sting a bit but will do more
>>> damage to the fan. We had Papst fans at work (suitable for 160 - +100C) and
>>> we were told that they wrere dangerous and need guards. I tested one with a
>>> bit of rolled up paper then with my finger then showed people. After the
>>> 3rd. or 4th. tim it wouldn't start again. I'd guess that just once would
>>> damage those tiny bearings.
>>> So yes, you do need a fan guard, not a finger guard.
>>
>> Hmm. I've had substantial finger gouges from both ~2500rpm 80mm fans and
>> ~5000rpm 60mm fans. I wouldn't push this as general advice.
>>
>> The fans were fine afterwards :-)
>
>But you gave up - try 4 or 5 times. Only give up when you run out of
>fingers.
I recommend the Mavis Beacon touch-typing with knuckles course.
--
Roger Hunt
Mike Tomlinson
<ju...@jaj22.org.uk> writes
>The fans were fine afterwards :-)
Your finger sacrificed itself to save the fan, just as semiconductors
sacrifice themselves to save the fuse.
John Stigter
from Mike Tomlinson <nos...@jasper.org.uk.invalid> contains these words:
Many years ago, there was a well known (at the time) saying which went:
"It is the job of the ten dollar transistor to sacrifice itself in
order to save a ten cent fuse."
The modern version (how times have changed! ;-) now goes:
"It is the job of the one cent transistor to sacrifice itself in order
to save a ten cent fuse."
The problem with this is that the "one cent transistor" is more likely
to be one of thousands to millions integrated into a 5 to 100 dollar
chip. However, what _is_ surprising is how often the rest of that "one
cent transistor"'s siblings will 'gang up' on the ten cent fuse and blow
it away. Modern cmos technology has undoubtedly improved the resilience
of integrated circuit devices against electrical abuse (within limits).
The point of fitting a 'humble ten cent fuse' is (and always has been)
not to stop the chips emitting the 'magic smoke' but to stop the office
or home from doing likewise.
--
Regards, John
Dave J.
01:04:07 GMT, in uk.comp.homebuilt, 'Johnny B Good' wrote:
>[1] Paste, mind you, not that horrible thermal gasket they used to stick
>onto the heatsink base. If your brand new heatsink has been cursed with
>such a 'thermal pad' gasket, you really do need to carefully scrape it
>off and use proper heatsink paste instead (or dry mount if needs must).
What, out of interest, was(is?) the logic behind the requirement I've seen
before now on CPU boxes to use *only* the stupid little gasket thing and
to *not* apply any thermal paste, on penalty of a void warrantee?
Dave J.
GSV Three Minds in a Can
J. <req...@freeuk.com> said
The gasket ('bubble gum') is basically idiot proof. With paste you need
to know how much, you need to spread it out, you need to use the right
stuff, etc. In olden days it was fairly easy to actually break a corner
of the die messing about with paste, or to cause a catastrophic
overheating, but these days it's pretty well idiot proof whatever you
do.
--
GSV Three Minds in a Can
13,113 Km walked. 2,500 Km PROWs surveyed. 45.1% complete.
Jaimie Vandenbergh
wrote:
>In MsgID<3130303037373...@plugzetnet.co.uk> on Sun, 14 Dec 2008
>01:04:07 GMT, in uk.comp.homebuilt, 'Johnny B Good' wrote:
>
>>[1] Paste, mind you, not that horrible thermal gasket they used to stick
>>onto the heatsink base. If your brand new heatsink has been cursed with
>>such a 'thermal pad' gasket, you really do need to carefully scrape it
>>off and use proper heatsink paste instead (or dry mount if needs must).
Dry mount? Horrors! There's no way dry mounting is going to be better
than a thermal pad, particularly with the lumpy surface left after
scraping the pad off.
>What, out of interest, was(is?) the logic behind the requirement I've seen
>before now on CPU boxes to use *only* the stupid little gasket thing and
>to *not* apply any thermal paste, on penalty of a void warrantee?
Idiot proofing. But which CPUs have you seen that on? I've never
noticed it.
Cheers - Jaimie
--
"If apathy and manipulation aren't strengths, I don't know what is" - Zadok, 1/0
Johnny B Good
from Jaimie Vandenbergh <jai...@sometimes.sessile.org> contains these words:
> On Sat, 27 Dec 2008 15:38:27 +0000, Dave J. <req...@freeuk.com>
> wrote:
> >In MsgID<3130303037373...@plugzetnet.co.uk> on Sun, 14 Dec 2008
> >01:04:07 GMT, in uk.comp.homebuilt, 'Johnny B Good' wrote:
> >
> >>[1] Paste, mind you, not that horrible thermal gasket they used to stick
> >>onto the heatsink base. If your brand new heatsink has been cursed with
> >>such a 'thermal pad' gasket, you really do need to carefully scrape it
> >>off and use proper heatsink paste instead (or dry mount if needs must).
> Dry mount? Horrors! There's no way dry mounting is going to be better
> than a thermal pad, particularly with the lumpy surface left after
> scraping the pad off.
Sorry, my bad. I forgot to add the word "_ALL_" to that reference to
'carefully scrape it off' (it, for absolute clarity, should have read
"carefully scrape it _ALL_ off"). I'm afraid I was rather taking it for
granted that such would have been implied by most readers.
The normal end result of such cleaning, ime, is that you end up with
the originally flat surface once any such gunk is _properly_ (and
carefully) removed. If the surface is _still_ lumpy afterward, it simply
means you did a less than satisfactory job of cleaning said surface(s).
> >What, out of interest, was(is?) the logic behind the requirement I've seen
> >before now on CPU boxes to use *only* the stupid little gasket thing and
> >to *not* apply any thermal paste, on penalty of a void warrantee?
> Idiot proofing. But which CPUs have you seen that on? I've never
> noticed it.
Yes, "Idiot Proofing" is exactly right! Better to have a less than
ideal thermal contact than entertain the possibilty of allowing the end
user to get an even worse one. Also, improper heatsink mounting will
become evident upon inspection of the state of the thermal pad which
acts as a 'Tell Tale' in the event of a warranty claim.
There's also the bonus that a competent user electing to make a better
job of it is automatically going to 'void' the warranty and thus relieve
the manufacturer of such an obligation. Basically, from the
manufacturer's PoV, it's a "Win Win" scenario.
--
Regards, John.
Please remove the "ohggcyht" before replying.
The address has been munged to reject Spam-bots.
Dave J.
Sat, 27 Dec 2008 16:34:01 +0000, in uk.comp.homebuilt, 'Jaimie
Vandenbergh' wrote:
>>What, out of interest, was(is?) the logic behind the requirement I've seen
>>before now on CPU boxes to use *only* the stupid little gasket thing and
>>to *not* apply any thermal paste, on penalty of a void warrantee?
>
>Idiot proofing. But which CPUs have you seen that on? I've never
>noticed it.
None recently, but then I've not built a new machine for at least three
months nor one of my 'do the basics very well indeed' entry level
favorites for more like six. However, at that time, on whichever AMD
offerings showed the best buying time instructionrate:price ratio, a fair
proportion came with a mandatory melty cushion.
I'm not a hundred percent about this but ISTR using thermal paste on the
vey last thing I built. Slightly higher up the food chain, AMD 2.2Ghz X2,
brisbane cored (I think) on an AM2 base.
Not the widest spread of samples :-)
Dave J.