Welcome to PCForumz.com!
FAQFAQ   SearchSearch      ProfileProfile    Private MessagesPrivate Messages   Log in/Register/PasswordLog in/Register/Password

"Top Of The Line" HS & F ???

 
Goto page 1, 2
   Hardware Problem Solving Community! (Home) -> Overclocking RSS
Related Topics:
Raid 0 off line - I have a problem with two hd Maxtor Plus 9 160gb ata/133 in raid 0 I have an Asus P4C800-E De luxe with the Promise PDC20378 All is gone ok for six months then, last evening at the boot I had a message..

Looking for Flow meter for 1/2 line in watercooling setup - Does anyone have an idea of where I can get a flow meter for my setup? I'd prefer the kind that connects to a fan header on the so it can be monitored (even if it only registers on or off) Any help will be Thanks

AMD Athlon 64 3200+ or Dual AMD Athlon XP 2800+ - First of all, sorry for What would you suggest between Athlon 64 3200+ and Dual Athlon XP 2800+? I mean, which will be more usable in real life (Video games, and other I'll buy it next week so any..

High-end configuration... - What would be your high-end CPU, MB, RAM, HDD, Graphic Card. Without Case, PSU, and such... I was thinking about this: AMD Athlon 64 3200+ ATI X800XT RAM, 1GB, DDR, PC-3500, 433 MHz, KINGMAX (is Kingmax best brand?) DUAL HDD MAXTOR 160..

Most processing power/Watt - Something I've been thinking about recently, of the CPUs in the 700Mhz and upwards range, which CPU, or family of CPUs gives the most for the least, I'm running a few machines for SETI and am on a budget. I have a bit of..
Author Message
sutliff

External


Since: Jan 27, 2005
Posts: 3



(Msg. 1) Posted: Sat Oct 27, 2007 5:02 am
Post subject: "Top Of The Line" HS & F ???
Archived from groups: alt>comp>hardware>overclocking (more info?)

For AMD Athlon X2 Dual-Cores? From like 3800 - 6000+'s?

I hae been perusing through Zalman & Artic Cooling's offerings. Also been
flipping through my past 2 1/2 years of Max PC mags.

Also, thank you for your answers & advice to my previous post of 10/22/2007
on/in this NG & others. =)

Checking back soon,

Sniper
sutliff DeleteThis @gmail.com
NY

..

 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Brett Kline

External


Since: Oct 27, 2007
Posts: 2



(Msg. 2) Posted: Sat Oct 27, 2007 7:56 am
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Archived from groups: per prev. post (more info?)

Sniper wrote:
> For AMD Athlon X2 Dual-Cores? From like 3800 - 6000+'s?
>
> I hae been perusing through Zalman & Artic Cooling's offerings. Also
> been flipping through my past 2 1/2 years of Max PC mags.
>
> Also, thank you for your answers & advice to my previous post of
> 10/22/2007 on/in this NG & others. =)
>


IFX-14
http://www.thermalright.com/a_page/main_product_ifx14.htm


Thermalright Ultra-120 Extreme
http://www.thermalright.com/a_page/main_product_ultra120_extreme.htm


Thermaltake offers the "Mini Typhoon" which may appeal to the more budget
conscious.
http://www.thermaltakeusa.com/product/Cooler/Retail/CL-P0268/cl-p0268....#overvi

 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Howard Goldstein

External


Since: Apr 19, 2007
Posts: 53



(Msg. 3) Posted: Sat Oct 27, 2007 4:40 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Imported from groups: per prev. post (more info?)

Back to top
Login to vote
Brett Kline

External


Since: Oct 27, 2007
Posts: 2



(Msg. 4) Posted: Sat Oct 27, 2007 4:40 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Archived from groups: per prev. post (more info?)

Howard Goldstein wrote:
> On Sat, 27 Oct 2007 07:56:05 -0500, Brett Kline <m5i4.DeleteThis@16m76ydi.ljf>
> wrote:
>> IFX-14
>> http://www.thermalright.com/a_page/main_product_ifx14.htm
>
> That really does look nice doesn't it? Like a mutant 120 extreme with
> siamese twin syndrome, in a good way.
>
> Where does that addon heat pipe doohickey go, on the solder side of
> the board?


Looks that way Howard.
http://www.thermalright.com/a_page/main_support_installation_ifx14.htm

Here's a review:
http://www.xsreviews.co.uk/reviews/cpu-coolers/thermalright-ifx-14/1
 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Phil Weldon

External


Since: Feb 10, 2007
Posts: 140



(Msg. 5) Posted: Sat Oct 27, 2007 5:41 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Archived from groups: per prev. post (more info?)

'Howard Goldstein' wrote:
| That really does look nice doesn't it? Like a mutant 120 extreme with
| siamese twin syndrome, in a good way.
|
| Where does that addon heat pipe doohickey go, on the solder side of
| the board?
_____

To bad that, according to the review, you don't get particularly good
performance because the heatsink base is not particularly flat. Also, a
cooler for the opposite side of the motherboard beneath the CPU is of
vanishingly small benefit.

High in price only, not in performance.

Phil Weldon

"Howard Goldstein" <hgoldste DeleteThis @mpcs.com> wrote in message
news:1193503246.57768@news.queue.to...
| On Sat, 27 Oct 2007 07:56:05 -0500, Brett Kline <m5i4 DeleteThis @16m76ydi.ljf> wrote:
| : IFX-14
| : http://www.thermalright.com/a_page/main_product_ifx14.htm
|
| That really does look nice doesn't it? Like a mutant 120 extreme with
| siamese twin syndrome, in a good way.
|
| Where does that addon heat pipe doohickey go, on the solder side of
| the board?
 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Howard Goldstein

External


Since: Apr 19, 2007
Posts: 53



(Msg. 6) Posted: Sun Oct 28, 2007 4:25 am
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Imported from groups: per prev. post (more info?)

Back to top
Login to vote
Howard Goldstein

External


Since: Apr 19, 2007
Posts: 53



(Msg. 7) Posted: Sun Oct 28, 2007 4:33 am
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Imported from groups: per prev. post (more info?)

Back to top
Login to vote
Phil Weldon

External


Since: Feb 10, 2007
Posts: 140



(Msg. 8) Posted: Sun Oct 28, 2007 1:12 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Archived from groups: per prev. post (more info?)

'Howard Goldstein' wrote:
| Whoa that's very bad news. Grinding the very pronounced bend -
| strangely in one direction only, the axis in line with the major axis
| of the fins - out of the 120 extreme took hours. It was a
| *nightmare*. The copper base on the 120X is made of extremely hard
| copper, much harder than the copper in solid gauge wire, or on a CPU
| heat spreader.
|
| Since it's the same problem as the 120 x this is definitely not a
| product for someone who doesn't have a brand new pack of 180 grit
| sandpaper and a couple of hours.
_____

If the copper is 'extremely hard' that is even worse news. Copper can vary
greatly in heat conductivity depending on the alloy. Some copper alloys
(with copper content of more than 90%) can have less conductivity than
aluminum.

'A brand new pack of 180 grit' will not even begin to produce a surface
suitable heatsink surface. You need to finish up with something near 1200
or finer, and there should be four or more steps between 180 and 1200.

All in all, the description of the 120X makes me suspicious of any product
from the same company, whether its venality or ignorance to blame for the
product design and execution.

Phil Weldon

"Howard Goldstein" <hgoldste RemoveThis @mpcs.com> wrote in message
news:1193546035.92727@news.queue.to...
| On Sat, 27 Oct 2007 17:41:35 -0300, Phil Weldon
<not.disclosed RemoveThis @example.com> wrote:
| : 'Howard Goldstein' wrote:
| : | That really does look nice doesn't it? Like a mutant 120 extreme with
| : | siamese twin syndrome, in a good way.
| : |
| : | Where does that addon heat pipe doohickey go, on the solder side of
| : | the board?
| : _____
| :
| : To bad that, according to the review, you don't get particularly good
| : performance because the heatsink base is not particularly flat. Also,
a
| : cooler for the opposite side of the motherboard beneath the CPU is of
| : vanishingly small benefit.
| :
| : High in price only, not in performance.
|
| Whoa that's very bad news. Grinding the very pronounced bend -
| strangely in one direction only, the axis in line with the major axis
| of the fins - out of the 120 extreme took hours. It was a
| *nightmare*. The copper base on the 120X is made of extremely hard
| copper, much harder than the copper in solid gauge wire, or on a CPU
| heat spreader.
|
| Since it's the same problem as the 120 x this is definitely not a
| product for someone who doesn't have a brand new pack of 180 grit
| sandpaper and a couple of hours.
|
|
 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Howard Goldstein

External


Since: Apr 19, 2007
Posts: 53



(Msg. 9) Posted: Sun Oct 28, 2007 7:05 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Imported from groups: per prev. post (more info?)

Back to top
Login to vote
Phil Weldon

External


Since: Feb 10, 2007
Posts: 140



(Msg. 10) Posted: Sun Oct 28, 2007 7:05 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Archived from groups: per prev. post (more info?)

'Howard Goldstein' wrote:
| Strike that; you'd want a lathe, or maybe a brand new pack of 90 grit
| sandpaper to start off. It's as best as I can describe it a
| cylindrical hump about 1/8" high and 1/4" wide running the length of
| the base in line with the fins. It's so pronounced that you're not
| going to worry about moving up a grit for at least 3-4 full sheets and
| 2 hours of solid grinding with the rough stuff. In retrospect I
| should have used a dremel with a stone bit to grind it down, but WTH
| do I know about this I'm still cutting my teeth on OCing.
|
| This was my very first lapping project and man it was a shocker.
|
| :
| : All in all, the description of the 120X makes me suspicious of any
product
| : from the same company, whether its venality or ignorance to blame for
the
| : product design and execution.
|
| I'm really concerned that it's endemic to the design or whatever
| fabrication method they've selected. I thought mine was an
| aberration. It's a pity as with the 120X lapped I never get above
| coretemps of 60C with an unlapped E6750 (2.6g) at 3.6Ghz and 1.415
| vcore with 1400RPM on the fan. And they do seem to be the only people
| who make non-permanent air coolers for the power FETs (HR-09).
|
| What's the consensus on alternatives?
_____

What are the alternatives? Almost anything. A Thermaltake i7 at less than
half the price, for one. What Thermalright SHOULD have done is to surface
grind the heatsink block (as well as losing the hinky flame fin profile);
and then it STILL would not be worth the price.

For lapping a CPU heatspreader or heatsink lapping compound with water or
waterproof sandpaper/emery paper is the best choice. Use a glass surface as
the lap, flood it with water, wash everything carefully between each grit
grade. Hold the part to be lapped loosely, rotate a random amount every
dozen strokes or so. Use very light pressure. When lapping copper the
water is especially useful for removing the copper chips which otherwise
will clog the paper or smear.

The temperature of air ENTERING the heatsink fins is just as important as
the heatsink. Reducing the temperature of the air entering the fins can
drop the heatsink temperature almost as much as the air temperature drop.
Proper air paths THROUGH the case are also important. (I use a 120 mm fan
blowing across two 500 GByte SATA drives into the heatsink fins of a
Thermaltake i7 with 120 mm fan and two 80 mm fans pulling air through the
heatsink fins an exhausting from the rear of the case. A 120 mm fan in the
case side blows down across the first PCI-e X16 slot and nVidia heatsink/fan
on the SLI SPP (Northbridge) chip.
Results for my system (E4300 @ 2.7 GHz and 1.325 v 1200 MHz FSB, EVGA 680i
SLI motherboard, 2 GByte Patriot PC-1066 SLI memory at 1200 MHz 1:2 CPU
clock : Memory clock ratio, Antec Neo HE 550 modular power supply; running
Windows XP Pro, Orthos small FFTs, CPU temperature by TAT, Thermaltake i7,
fans as above

Room ambient: 21 C

CPU temperature: 56 C

Thermaltake CPU fan: 100% 2445 rpm (the quietest fan of the bunch)

nVidia Northbridge fan: 100% 4998 rpm (undetectable noise production)
(and removable)

rear 80 mm fans: 100% 2960 rpm (noisiest fans of the bunch)

front 120 mm fan: very low noise at 1400 rpm; keeps the hard drives
cool, CPU temperature difference only a degree or two variation between 1400
rpm and highest speed

side 120mm fan: (fairly noisy at high speed, used only a low speed when
overclocking my EVGA 8800 GTS 320 MByte graphics card

Enlight 8950 rack-mountable server case: 17" X 9" X 26", 9 X 5.25 inch,
1 X 3.5 inch front bays.

For air cooling on this system the CPU tops out at 3.2 GHz and 1.425 volts -
at that point the Orthos FFTs CPU temperature reported by TAT is 70 C and an
ambient room temperature of 20 C. I have a water cooling system I'm not
using. I am not a big first-person-shooter gamer, so I built this system
for overclocking fun and because it was TIME to move from my Pentium
Prescott. If I were more disciplined (like 'Ed Medlin') I would install the
water cooling system; I'd not expect the performance increase with better
air cooling to mean much to me. What with the Q6600 now available so
cheaply I'd consider it if changing to a Nelahem involves too much expense
in a year or so.

Phil Weldon

"Howard Goldstein" <hgoldste DeleteThis @mpcs.com> wrote in message
news:1193598347.1187@news.queue.to...
| On Sun, 28 Oct 2007 13:12:20 -0300, Phil Weldon
<not.disclosed DeleteThis @example.com> wrote:
| : 'Howard Goldstein' wrote:
| : | Whoa that's very bad news. Grinding the very pronounced bend -
| : | strangely in one direction only, the axis in line with the major axis
| : | of the fins - out of the 120 extreme took hours. It was a
| : | *nightmare*. The copper base on the 120X is made of extremely hard
| : | copper, much harder than the copper in solid gauge wire, or on a CPU
| : | heat spreader.
| : |
| : | Since it's the same problem as the 120 x this is definitely not a
| : | product for someone who doesn't have a brand new pack of 180 grit
| : | sandpaper and a couple of hours.
| : _____
| :
| : If the copper is 'extremely hard' that is even worse news. Copper can
vary
| : greatly in heat conductivity depending on the alloy. Some copper
alloys
| : (with copper content of more than 90%) can have less conductivity than
| : aluminum.
|
| It was as hard as I'd imagine it'd be to try to lap a block of steel.
| "Hard" in all the senses; the high center of gravity makes it really
| awkward to hold on to for lapping.
|
| :
| : 'A brand new pack of 180 grit' will not even begin to produce a surface
| : suitable heatsink surface. You need to finish up with something near
1200
| : or finer, and there should be four or more steps between 180 and 1200.
|
| Strike that; you'd want a lathe, or maybe a brand new pack of 90 grit
| sandpaper to start off. It's as best as I can describe it a
| cylindrical hump about 1/8" high and 1/4" wide running the length of
| the base in line with the fins. It's so pronounced that you're not
| going to worry about moving up a grit for at least 3-4 full sheets and
| 2 hours of solid grinding with the rough stuff. In retrospect I
| should have used a dremel with a stone bit to grind it down, but WTH
| do I know about this I'm still cutting my teeth on OCing.
|
| This was my very first lapping project and man it was a shocker.
|
| :
| : All in all, the description of the 120X makes me suspicious of any
product
| : from the same company, whether its venality or ignorance to blame for
the
| : product design and execution.
|
| I'm really concerned that it's endemic to the design or whatever
| fabrication method they've selected. I thought mine was an
| aberration. It's a pity as with the 120X lapped I never get above
| coretemps of 60C with an unlapped E6750 (2.6g) at 3.6Ghz and 1.415
| vcore with 1400RPM on the fan. And they do seem to be the only people
| who make non-permanent air coolers for the power FETs (HR-09).
|
| What's the consensus on alternatives?
 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Howard Goldstein

External


Since: Apr 19, 2007
Posts: 53



(Msg. 11) Posted: Sun Oct 28, 2007 11:41 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Imported from groups: per prev. post (more info?)

Back to top
Login to vote
Howard Goldstein

External


Since: Apr 19, 2007
Posts: 53



(Msg. 12) Posted: Sun Oct 28, 2007 11:49 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Imported from groups: per prev. post (more info?)

Back to top
Login to vote
Ed Medlin

External


Since: Sep 28, 2004
Posts: 398



(Msg. 13) Posted: Mon Oct 29, 2007 11:36 am
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Archived from groups: per prev. post (more info?)

>
> What are the alternatives? Almost anything. A Thermaltake i7 at less
> than
> half the price, for one. What Thermalright SHOULD have done is to surface
> grind the heatsink block (as well as losing the hinky flame fin profile);
> and then it STILL would not be worth the price.
>
> For lapping a CPU heatspreader or heatsink lapping compound with water or
> waterproof sandpaper/emery paper is the best choice. Use a glass surface
> as
> the lap, flood it with water, wash everything carefully between each grit
> grade. Hold the part to be lapped loosely, rotate a random amount every
> dozen strokes or so. Use very light pressure. When lapping copper the
> water is especially useful for removing the copper chips which otherwise
> will clog the paper or smear.
>
> The temperature of air ENTERING the heatsink fins is just as important as
> the heatsink. Reducing the temperature of the air entering the fins can
> drop the heatsink temperature almost as much as the air temperature drop.
> Proper air paths THROUGH the case are also important. (I use a 120 mm fan
> blowing across two 500 GByte SATA drives into the heatsink fins of a
> Thermaltake i7 with 120 mm fan and two 80 mm fans pulling air through the
> heatsink fins an exhausting from the rear of the case. A 120 mm fan in
> the
> case side blows down across the first PCI-e X16 slot and nVidia
> heatsink/fan
> on the SLI SPP (Northbridge) chip.
> Results for my system (E4300 @ 2.7 GHz and 1.325 v 1200 MHz FSB, EVGA 680i
> SLI motherboard, 2 GByte Patriot PC-1066 SLI memory at 1200 MHz 1:2 CPU
> clock : Memory clock ratio, Antec Neo HE 550 modular power supply;
> running
> Windows XP Pro, Orthos small FFTs, CPU temperature by TAT, Thermaltake i7,
> fans as above
>
> Room ambient: 21 C
>
> CPU temperature: 56 C
>
> Thermaltake CPU fan: 100% 2445 rpm (the quietest fan of the bunch)
>
> nVidia Northbridge fan: 100% 4998 rpm (undetectable noise production)
> (and removable)
>
> rear 80 mm fans: 100% 2960 rpm (noisiest fans of the bunch)
>
> front 120 mm fan: very low noise at 1400 rpm; keeps the hard drives
> cool, CPU temperature difference only a degree or two variation between
> 1400
> rpm and highest speed
>
> side 120mm fan: (fairly noisy at high speed, used only a low speed
> when
> overclocking my EVGA 8800 GTS 320 MByte graphics card
>
> Enlight 8950 rack-mountable server case: 17" X 9" X 26", 9 X 5.25
> inch,
> 1 X 3.5 inch front bays.
>
> For air cooling on this system the CPU tops out at 3.2 GHz and 1.425
> volts -
> at that point the Orthos FFTs CPU temperature reported by TAT is 70 C and
> an
> ambient room temperature of 20 C. I have a water cooling system I'm not
> using. I am not a big first-person-shooter gamer, so I built this system
> for overclocking fun and because it was TIME to move from my Pentium
> Prescott. If I were more disciplined (like 'Ed Medlin') I would install
> the
> water cooling system; I'd not expect the performance increase with better
> air cooling to mean much to me. What with the Q6600 now available so
> cheaply I'd consider it if changing to a Nelahem involves too much expense
> in a year or so.
>
> Phil Weldon
>

Right now, the Q6600 is almost a "budget" processor for what it does.
http://service.futuremark.com/compare?3dm06=3534404 at 17743 is pretty good
and a major increase over the E6600 in 3DMark 06. Cooling it at idle is a
problem (I think) over the "E" series even with liquid. Stress temps seem
fine and where I would expect them to be at 50C or a bit over at 1.35Vcore.
I use CoreTemp and it just doesn't seem to read this processor as low as I
would expect at idle with all four cores close to 40C. Bios reads mid-upper
20sC for the CPU. I am not sure where the problem is, if there indeed is a
problem.
I still have an I630 (3Ghz EM64T) running at 3.6Ghz and I keep it cool
(relatively.......Smile with a Spire HS/Fan unit which is actually an OEM
replacement unit that is just a bit bulkier but quieter than the Intel
counterpart. The copper insert finish was really nice right out of the box
so I didn't have to mess with any lapping. I got it at my local computer
shop for around $10.00 IIRC. It isn't fancy, but just very functional and
does the job. I bet it would do well with a C2D or Quad if it cools that
furnace of a processor.



Ed
 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Phil Weldon

External


Since: Feb 10, 2007
Posts: 140



(Msg. 14) Posted: Mon Oct 29, 2007 4:16 pm
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Archived from groups: per prev. post (more info?)

'Howard Goldstein' wrote, in part:
| How would you approach a seriously out-of-flat base like this one? Or
| just ditch it. I think I still have more work to do because when I
| popped the heat sink off sure enough there was MX2 accumulating on
| either side of what's left of that cylinder protburance that I thought
| I took care of after the first 2 hours of grinding a month or so ago.
| Would you try dremeling it down?
_____

Do NOT try a Dremel tool to grind the heatsink. The heatsink (and CPU
heatspreader, which has also had a flatness problem in the past) should be
flat over a large scale (check with a straight edge placed at different
angles across the surface), on a medium scale ( 1 mm ), over a small scale
( 0.1 mm ), and at a microscopic level ( < 10 micrometers ). Grinding with
a Dremel tool could take off the high points, but is not suitable for
producing a flat surface at ANY of these levels. And if you should slip,
well, one slip and you may be worse off than when you started.

Something to put CPU heat dissipation into perspective. The Pentium III 1
GHz SECC2 CPU has NO heatspreader; the die is about 9 mm X 11 mm. I can
dissipate around 60 Watts. At that level, the Watts per square cm heat flux
is GREATER than that of the standard 2500 Watt 8" electric stove top element
{ 9 mm X 11 mm = 1 sq cm; 8"/2 X 8"/2 X 3.14 = 50 sq. in. = 120 square cm }
60 Watts / 1 square cm = 60 Watts per square cm while 2500 Watts / 120 cm =
only 21 Watts per square cm! Even with a 32 mm X 32 mm integral
heatspreader a 130 Watt tpd Q6xx Core 2 Duo quad has a heat flux of ~ 13
Watts per square cm. Perhaps before I retire my E4300 I'll remove the
heatspreader and see how much I can improve on whatever Intel has between
the die and heatspreader... or not.

At the end of this message are directions for lapping I posted here in
January 17, 2001. Your task will be more awkward than in days of yore
because of all the fin and heatpipe top-hamper.

I don't have any exact measurements for my nVidia 680i 'northbridge' and
'southbridge', but the heatsink base of the 'northbridge' and the heatpipe
from the 'southbridge' heatsink that connects to the 'northbridge' heatsink
fins are, to the touch of my finger, cozy warm - I'd estimate ~ 40 - 45 C.
The switching FETs for the + 12 vdc DC-DC downconvertor/regulator are cooled
by two chunky anodized aluminum blocks with a few fat fins. The temperature
to the touch seems about the same, ~ 40 - 45 C. The fan on the fins that
cool the 'northbridge' and 'southbridge' chips gets plenty of air
circulation. I removed the plate that fits around the rear panel ATX
connectors to provide ventilation directly adjacent to the FET heatsinks.

For filtering, I use a washable nylon mesh electrostatic filter on the 120
mm side fan (air moving over the nylon produces a static charge that
attracts dust and smaller particles.) The front 120 mm fan has a washable
foam filter. The other two chassis fans and the Antec HE 550 power supply
fan exhaust air to the outside so they don't need filters. The case overall
operates at a slight positive air pressure, so the whole thing stays pretty
clean. I lined the entire inside of the case (except underneath the
motherboard) with ~ 3/16" dense rubber foam from one of the Thermal twins.

Lapping or Grinding a Heatsink Flat and Smooth (posted by Phil Weldon to
alt.comp.hardware.overclocking January 17, 2001 - the CPU names have
changed, but the principles remain the same.
1. Use a piece of flat glass as your work surface (it must be very flat,
and supported well)
2. Use waterproof sandpaper (sometimes called wet or dry)
3. Use lots of water (several cc at a time, the water flushes the metal
particles away, keeps the sandpaper clog free, and makes it stick to the
glass)
4. Depending on how flat and smooth your heatsink is, expect to spend 30
minutes or more at this work (I do it on a glass coffee table top in front
of the television - you don't even have to watch the work [of course, if the
show is bad enough, you can watch the work, and finish faster])
5. The figure 8 pattern is correct, but I would also rotate the heatsink 90
degrees approximately (NOT exactly) from time to time
6. If your heat sink is relatively flat and smooth, then 400 grit is a good
place to start. If after ten minutes of work you do not see much progress
(there are many pits larger than the surface roughness left by the 400 grit
sandpaper), drop to a lower number grit (320) or (260).
7. Using finer grit sandpaper will help improve heat transfer: 1500 is
certainly fine enough, but since you need to get such fine grades at an
automobile parts store (it is used for finishing automobile paintwork), you
might as well get a sheet each of 600, 800, 1000, 1500, and 2000. The more
coarse grits are easy to find at a paint or hardware store 40, 60, 80, 120,
180, 220, 280, 320, 400, 600. (Grit # 40 will take the portrait off a franc
coin in a minute or so; the average heatsink with faint circular milling
marks can probably be started at 320.)
8. The work can be completed in less time if you step up through each
grade, removing all the pits left by the previous grade before moving to the
next finer grit.
9. Each time you change to a finer grit, VERY carefully wash the work
surface, your hands, the heatsink, and the sandpaper. Any larger grit left
when you start with a finer grade will leave visible scratches.
10. The waterproof (or wet/dry) sandpaper will last a long time. After you
are through with the work, wash the sandpaper and let it dry. You can reuse
it many times as long as it is not torn.
11. As an example of how fine 2000 grit is, you could use it on Mercedes
paintwork and almost not notice an effect other than the missing wax polish.
12. Do not use too much pressure when sanding the heatsink; you want a flat
as well as a smooth surface, and too much pressure can cause one edge or
corner to be ground down more quickly (that is one reason for occasionally
rotating the heatsink a random amount.) Too much pressure may also cause a
corner to tear the sandpaper. If you do not tear the sandpaper, one sheet
of each grade should last through ten or more heatsinks.
13. Emery cloth is similar to sandpaper, except that the base is cloth,
rather than paper. It is ok for the rougher grits, but I think that I will
not give as flat a surface when you move to finer grits.
14. Crocus cloth. This is a polishing material similar to emery cloth, but
it has jeweler's rouge on the surface rather than emery grit. THIS IS NOT
waterproof. I tried it to get a final polish, and added water. All the
jeweler's rouge washed off! For a final polish, a soft cloth and a solvent
that will remove dirt and grease is about all you need.
15. The method of grinding the heatsink you and I have described is the way
flat mirrors for astronomical use are made (except there are machines that
do the work now.) A spherical section mirror can be ground in exactly the
same way, except that the grinding surface is circular, and the same size as
the mirror. (The only additional action used for the flat and spherical
mirrors is that the person grinding the mirror, also, after six or eight
strokes, walks partially around the work surface to grind from a new angle.)
This is probably not necessary for a heatsink, since the method used on
astronomical mirrors makes a surface that is smooth and correct to less than
100 nanometers! At that point, thermal compound would not be necessary, and
in fact, would have a negative effect. Even just going in a careful manner
to the 1500 grit will give you a surface so flat that, when placed on an
FC-PGA CPU,
will lift the CPU when the heatsink is raised.
16. Penultimate warning: the older CPU's with a relatively large metal cap
(about 20 mm X 20 mm) can be very uneven. Sometimes the metal cap may be
concave to the extent of a millimeter or two in the center. For something
this bad, you will want to start with 280 grit.
17. Final warning: the FC-PGA CPU's, and the SECC2 CPU's have a very small
heat transfer surface, about 11 mm X 9 mm. This is the bottom of the
silicon slice the CPU is built into. It is laser cut from a single crystal
of silicon, and cannot be made any smoother or flatter. DO NOT grind or
polish this surface. It will not help, and the CPU gates and channels are
less than 1 mm below this surface.

Phil Weldon

"Howard Goldstein" <hgoldste.TakeThisOut@mpcs.com> wrote in message
news:1193615390.18183@news.queue.to...
| On Sun, 28 Oct 2007 16:34:29 -0300, Phil Weldon
<not.disclosed.TakeThisOut@example.com> wrote:
| : 'Howard Goldstein' wrote:
| : | Strike that; you'd want a lathe, or maybe a brand new pack of 90 grit
| : | sandpaper to start off. It's as best as I can describe it a
| : | cylindrical hump about 1/8" high and 1/4" wide running the length of
| : | the base in line with the fins. It's so pronounced that you're not
| : | going to worry about moving up a grit for at least 3-4 full sheets and
| : | 2 hours of solid grinding with the rough stuff. In retrospect I
| : | should have used a dremel with a stone bit to grind it down, but WTH
| : | do I know about this I'm still cutting my teeth on OCing.
| : |
| : | This was my very first lapping project and man it was a shocker.
| : |
| : | :
| : | : All in all, the description of the 120X makes me suspicious of any
| : product
| : | : from the same company, whether its venality or ignorance to blame
for
| : the
| : | : product design and execution.
| : |
| : | I'm really concerned that it's endemic to the design or whatever
| : | fabrication method they've selected. I thought mine was an
| : | aberration. It's a pity as with the 120X lapped I never get above
| : | coretemps of 60C with an unlapped E6750 (2.6g) at 3.6Ghz and 1.415
| : | vcore with 1400RPM on the fan. And they do seem to be the only people
| : | who make non-permanent air coolers for the power FETs (HR-09).
| : |
| : | What's the consensus on alternatives?
| : _____
| :
| : What are the alternatives? Almost anything. A Thermaltake i7 at less
than
| : half the price, for one. What Thermalright SHOULD have done is to
surface
| : grind the heatsink block (as well as losing the hinky flame fin
profile);
| : and then it STILL would not be worth the price.
|
| Thanks I'll look into the i7. FWIW, the 120X doesn't have the flames
| or the funky little addon back-side cooler. Now I wouldn't mind
| having a backside cooler but not for my computer...
|
| : For lapping a CPU heatspreader or heatsink lapping compound with water
or
| : waterproof sandpaper/emery paper is the best choice. Use a glass
surface as
| : the lap, flood it with water, wash everything carefully between each
grit
| : grade. Hold the part to be lapped loosely, rotate a random amount
every
| : dozen strokes or so. Use very light pressure. When lapping copper the
| : water is especially useful for removing the copper chips which
otherwise
| : will clog the paper or smear.
|
| How would you approach a seriously out-of-flat base like this one? Or
| just ditch it. I think I still have more work to do because when I
| popped the heat sink off sure enough there was MX2 accumulating on
| either side of what's left of that cylinder protburance that I thought
| I took care of after the first 2 hours of grinding a month or so ago.
| Would you try dremeling it down?
|
| : The temperature of air ENTERING the heatsink fins is just as important
as
| : the heatsink. Reducing the temperature of the air entering the fins
can
| : drop the heatsink temperature almost as much as the air temperature
drop.
| : Proper air paths THROUGH the case are also important. (I use a 120 mm
fan
| : blowing across two 500 GByte SATA drives into the heatsink fins of a
| : Thermaltake i7 with 120 mm fan and two 80 mm fans pulling air through
the
| : heatsink fins an exhausting from the rear of the case. A 120 mm fan in
the
| : case side blows down across the first PCI-e X16 slot and nVidia
heatsink/fan
| : on the SLI SPP (Northbridge) chip.
| : Results for my system (E4300 @ 2.7 GHz and 1.325 v 1200 MHz FSB, EVGA
680i
| : SLI motherboard, 2 GByte Patriot PC-1066 SLI memory at 1200 MHz 1:2 CPU
| : clock : Memory clock ratio, Antec Neo HE 550 modular power supply;
running
| : Windows XP Pro, Orthos small FFTs, CPU temperature by TAT, Thermaltake
i7,
| : fans as above
| :
| : Room ambient: 21 C
| :
| : CPU temperature: 56 C
| :
| : Thermaltake CPU fan: 100% 2445 rpm (the quietest fan of the bunch)
| :
| : nVidia Northbridge fan: 100% 4998 rpm (undetectable noise
production)
| : (and removable)
| :
| : rear 80 mm fans: 100% 2960 rpm (noisiest fans of the bunch)
| :
| : front 120 mm fan: very low noise at 1400 rpm; keeps the hard
drives
| : cool, CPU temperature difference only a degree or two variation between
1400
| : rpm and highest speed
| :
| : side 120mm fan: (fairly noisy at high speed, used only a low speed
when
| : overclocking my EVGA 8800 GTS 320 MByte graphics card
| :
| : Enlight 8950 rack-mountable server case: 17" X 9" X 26", 9 X 5.25
inch,
| : 1 X 3.5 inch front bays.
| :
| : For air cooling on this system the CPU tops out at 3.2 GHz and 1.425
volts -
| : at that point the Orthos FFTs CPU temperature reported by TAT is 70 C
and an
| : ambient room temperature of 20 C. I have a water cooling system I'm
not
| : using. I am not a big first-person-shooter gamer, so I built this
system
| : for overclocking fun and because it was TIME to move from my Pentium
| : Prescott. If I were more disciplined (like 'Ed Medlin') I would
install the
| : water cooling system; I'd not expect the performance increase with
better
| : air cooling to mean much to me. What with the Q6600 now available so
| : cheaply I'd consider it if changing to a Nelahem involves too much
expense
| : in a year or so.
|
|
| Now that 680i starts out hot to begin with. Ed and I have Strikers
| with the dodgy (my opinion) heat sinks on the toaster I mean the
| northbridge. Are you staying nice and cool with your northbridge?
|
| Are you doing anything for filtering air going into the case? I have
| an antec 900 which seems to be very breezy on the low settings and I
| keep them there but it's collecting dust like a vacuum cleaner. I'll
| compare your temps against mine when my Striker gets back here on
| Tuesday if only as a test of my case cooling. It's going to use an
| Artic Freezer 7 which is the only other cooler I have here besides the
| stock.
 >> Stay informed about: "Top Of The Line" HS & F ??? 
Back to top
Login to vote
Howard Goldstein

External


Since: Apr 19, 2007
Posts: 53



(Msg. 15) Posted: Wed Oct 31, 2007 1:45 am
Post subject: Re: "Top Of The Line" HS & F ??? [Login to view extended thread Info.]
Imported from groups: per prev. post (more info?)

Back to top
Login to vote
Display posts from previous:   
   Hardware Problem Solving Community! (Home) -> Overclocking All times are: Pacific Time (US & Canada) (change)
Goto page 1, 2
Page 1 of 2

 
You can post new topics in this forum
You can reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum



[ Contact us | Terms of Service/Privacy Policy ]