The anatomy of a 'bow'.

Seems many are confused as to the whole 'bowing' sensation going round the town. Back in the day, before IHS's (or when they were easy to remove) we wanted bases flat as could be so to have the best contact with the silicon. A few months ago it is discovered that making a block convex can (and does) have a good impact on performance with IHS based CPU's. I won't go into the why's of it all, it should be pretty obvious though. Anyhow, Swiftech has started shipping fatter o-rings with their Apogee blocks, not installed but as an addition so the end user can install it and bow his base. I have verified that on a 6600 the bowed base is ~ 4.5C better than the non bowed base. Not to confuse here, the bases themselves are the SAME, it is the O-RING which will bow the base. Here is a pic of the bowed base, note the GAP between the base and the top, especially in the center between the screws that hold the block together:
http://www.anonforums.com/builds/tes.../GTX_bowed.jpg

One thing worth mentioning here is that once a base is bowed, it must remain this way for good - you can NOT revert it back to 'flat' as it will (the base) look sort of like a fun house mirror :D .

I must confess I tinker around with stuff, this is why I STILL haven't posted my testing results yet, I'm still NOT done, sigh. Anyhow, in my tinkering I discovered that I bowed the Fuzion. Yes that is right, the FUZION. I was working (well testing really) with the nozzles and the performance was NOT there, not like when I made my CRUDE nozzle. So I thought maybe water was not flowing all through the nozzle, but perhaps out between the top and mid plate and bypassing the base altogether. I know, sounds crazy so in an attempt to PREVENT this, I took an o-ring from an EK barb and placed it between the top and middle plate like so: (sorry for the crap focus, I need to pwn my camera) :mad:
http://www.anonforums.com/builds/tes...zion_oring.jpg

Well doing this had unexpected results! The base was not bowed like the apogee base, it is more bowed from the center, not the outside. You see, the o-ring forces the mid plate center to push the center of the base out. The cool thing is that now I'm experimenting with different size o-rings. For those of you that will rush out and try this, the EK o-ring is as BIG as I would go, a tiny bit smaller in thickness would be better IMO. The EK oring makes the base a bit 'dome-ish'. Does it perform better than stock? Yes it does, not as dramatically as the Apogee, but it does. It is a bit trickier to mount too as it has more of a point than a bowed apogee, at least with the EK barb o-ring.

Perhaps these photos will show what I mean, again sorry for the blur.

Here we have a FLAT base just removed (MP-05 LE):
http://www.anonforums.com/builds/tes..._cpu_paste.jpg

Even with the blur you will note how the paste is pretty even (don't mind there corners where there is no paste).

Next up is the bowed Fuzion look at the paste, note the center where there is very LIGHT paste, this is where most contact is being made (most pressure I should say)
http://www.anonforums.com/builds/tes...owed_paste.jpg

Next up are 3 bowed Apogee mounts, note the STRIP you see, this is the point of most pressure:
http://www.anonforums.com/builds/tes...ogee_bowed.jpg

http://www.anonforums.com/builds/tes...gee_bowed2.jpg

http://www.anonforums.com/builds/tes...gee_bowed3.jpg

A very important aspect of using a bowed based block are the mounts. YOu will have a larger deviation from mount to mount as you are not mating 2 flat surfaces... the Apogee's bow is a bit more forgiving in this regard - at least so far as I haven't had much time to play with different thicknesses of o-rings and the fuzion yet. Bottom line is if you are after best possible temps - use a bowed base and mount several times so you can get a baseline of where your mounts are, you will get a feel for good and bad.

I hope at least this clears up some confusion with bowed vs. flat. AS you know I have NOT released any test data yet, much to the dismay of many but I will say this, the Apogee's and the Fuzion are the 3 best blocks there are - the Storm and the MP-05 LE SP can NOT compete on IHS cpu's, it is just a fact. I am not going to recommend which block to people, right NOW there are only 3 to consider, the Apogee GT (bowed), the Apogee GTX (bowed) and the Fuzion. You can't go wrong with any of these and the performance difference is quite small anyway. I soon hope to have official numbers, thoughts, concerns and ramblings. :woot:

Very interesting.

Regarding the Fuzion :

What was the temp difference between the stock base and bowed base (with the EK O-ring)?

Quote:

---

Originally Posted by Jupiler

What was the temp difference between the stock base and bowed base (with the EK O-ring)?

---

I'll not go into that here... I don't have all the data anyways yet.

great post :toast:

this should be in the FAQ somewhere :)

I ordered the fuzion and am running a naked opteron 165. Because it is naked I'm assuming there would be no point in bowing it right?

Sticky! please, I am really glad you posted this, because now i can use a bowed fuzion when i get it. I'll also order a bunch of different brbs to experiment with the bow amount. it looks like the one nik has is actually a bit too bowed.

question, how does bowing work with a lapped IHS?

Quote:

---

Originally Posted by SparkyJJO

I ordered the fuzion and am running a naked opteron 165. Because it is naked I'm assuming there would be no point in bowing it right?

---

So, no bow for any of the apogees makes the fuzion a better performer ?

Does the Fuzion not have an already convex base?

Quote:

---

Originally Posted by serialk11r

Does the Fuzion not have an already convex base?

---

It does... the o-ring-turned-spacer causes the midplate to push the base out further, though.

Quote:

---

Originally Posted by Petra

It does... the o-ring-turned-spacer causes the midplate to push the base out further, though.

---

Sorry that was in response to Sparky's post.

Ok, but what if the CPU was lapped 100% flat, and the Base of the waterblock (fuZion) is lapped 100% flat? Since we now have 2 100% flat surfaces, would you want to introduce a "BOW"???

Quote:

---

Originally Posted by newls1

Ok, but what if the CPU was lapped 100% flat, and the Base of the waterblock (fuZion) is lapped 100% flat? Since we now have 2 100% flat surfaces, would you want to introduce a "BOW"???

---

There are no 100% flat surfaces, and when you mount the block onto the board...who knows what could happen to the base.

Quote:

---

Originally Posted by serialk11r

Does the Fuzion not have an already convex base?

---

That's true, I forgot about that. IIRC it is a slight convex shape but convex nonetheless.

Quote:

---

Originally Posted by serialk11r

There are no 100% flat surfaces, and when you mount the block onto the board...who knows what could happen to the base.

---

So then why do we lap our CPU's, Heatsinks, waterblocks? So even when I lap my CPU on a piece of thick solid glass (very flat surface) down to 1200grit, my CPU still isn't flat?

How much of a difference is there exactly between a fuzion and a bowed GTX ?

Quote:

---

Originally Posted by newls1

So then why do we lap our CPU's, Heatsinks, waterblocks? So even when I lap my CPU on a piece of thick solid glass (very flat surface) down to 1200grit, my CPU still isn't flat?

---

It is never going to be perfectly flat. But the fact is, after you've lapped your CPU, it is much flatter than it used to be.

Quote:

---

Originally Posted by SparkyJJO

It is never going to be perfectly flat. But the fact is, after you've lapped your CPU, it is much flatter than it used to be.

---

diamond bladed mill lapping !!! (maybe those could get it flat :P)

Quote:

---

Originally Posted by Darkatom

How much of a difference is there exactly between a fuzion and a bowed GTX ?

---

I want to know too.

Quote:

---

Originally Posted by [XC] M411b

I want to know too.

---

Depends on flow...and heatload ;)

its a pretty simple concept and i have known about it for a while, but where do i buy the o-ring?

Alex, me thinks you might have a run on EK barb o-rings :D

Quote:

---

Originally Posted by Darkatom

How much of a difference is there exactly between a fuzion and a bowed GTX ?

---

This question can NOT be answered as it depends.

nik, you said the EK was a little too thick with lot of bow, do you know an o-ring that is a little thinner that will reduce the bow a bit?

Quote:

---

Originally Posted by SNiiPE_DoGG

nik, you said the EK was a little too thick with lot of bow, do you know an o-ring that is a little thinner that will reduce the bow a bit?

---

No I don't have calipers to properly measure the thickness of the EK o-ring... perhaps Alex could help us out here? I'm thinking that a FLAT rubber gasket may be better than a round o-ring - perhaps it will spread the bow a bit better I just don't know yet - I've only used the EK oring and it is as large as I would go but it is useable. Once you do it though, it is a one way street.

Quote:

---

Originally Posted by nikhsub1

No I don't have calipers to properly measure the thickness of the EK o-ring... perhaps Alex could help us out here? I'm thinking that a FLAT rubber gasket may be better than a round o-ring - perhaps it will spread the bow a bit better I just don't know yet - I've only used the EK oring and it is as large as I would go but it is useable. Once you do it though, it is a one way street.

---

So, no matter the block, if you bow it, you cannot go back?

Quote:

---

Originally Posted by nikhsub1

No I don't have calipers to properly measure the thickness of the EK o-ring... perhaps Alex could help us out here? I'm thinking that a FLAT rubber gasket may be better than a round o-ring - perhaps it will spread the bow a bit better I just don't know yet - I've only used the EK oring and it is as large as I would go but it is useable. Once you do it though, it is a one way street.

---

The fact once bowed always bowed is a bit of a concern for future uses

On a GTX clealry the only way to go is a bowed block

On Fuzion...it seems to me you maybe best to leave it alone...as performance is close to bowed GTX

I will wait for niksub1 final results...dont forget if you bow the blocks to have a good backplate to protect mobo

BTW..excellent observations and creative thinking niksub1!!!!!!!!!!!

well im going to bow my regular apogee cus i really have nothing to loose since i'm about to mill my own block that will pwn all since it will be out of 1/2" thick copper instead of 1/4":p:

Quote:

---

Originally Posted by soccerswim8

wait, so with a lapped CPU, is the bow as effective?

and how is the bowed induced just from a thicker O-ring?

---

Remember when you mount it, the mounting plate comes down on FOUR CORNERS of the block. I suspect that this can slightly bend the base upward...I guess that's a big reason why the bow mount works so well, gives wonderful contact in the center where you need it. Sure with a lapped CPU it should work...:P

Quote:

---

Originally Posted by p8ntslinger676

well im going to bow my regular apogee cus i really have nothing to loose since i'm about to mill my own block that will pwn all since it will be out of 1/2" thick copper instead of 1/4":p:

---

? Who said 1/4" sucked? :p: heh the base of blocks these days are 1/4" so I'd like to see your block beat the Fuzion just with 1/2" copper ;) Remember that the water running through makes super tall pins not so useful if you don't have ginormous amounts of heat going through...

lol no, look at the dtek and how that would work to bow,

thicker middle + thinner edges = () <---- a convex shape

Quote:

---

Originally Posted by soccerswim8

oh ok. so the thicker O-ring just makes it stronger and less resistant to bend upward?

---

Note that the PRESSURE is now focused in the center of the IHS...Back in the old Apogee GT thread (I think it was that one) or was it the CES one? we were discussing the issue...that was one of the theories that came up, that the slight bending upwards made contact very poor. Anyways we can clearly see here (courtesy of nikhsub's camera...which is better than my webcam :P Hey I'll have that camera for free if you don't want it :D) that the bowed base is concentrating all the force near the center of the IHS where the core (s) are=more pressure=better contact :banana:

Quote:

---

Originally Posted by soccerswim8

but HOW does a thicker O-ring induce the bow?? i get the application, but i dont get how the thicker o-ring creates the bow

---

Uh oh, this is the second time today that I'm gonna show everyone my super leet MSP skillz...brb

http://img49.imageshack.us/img49/903...ountjn8.th.jpg

Quote:

---

Originally Posted by soccerswim8

but HOW does a thicker O-ring induce the bow?? i get the application, but i dont get how the thicker o-ring creates the bow

---

The thicker O-ring creates a bigger gap between the actual copper plate and the plastic block reservoir ( if you will ) and allows the copper to bend slightly in the center to create a "BOW" to allow more pressure to be directly placed on the IHS directly over the core (s) to allow for better cooling.
:D I learnt that in college.. :D

Quote:

---

Originally Posted by R3APER

Why do you have those big bolts sticking out of the motherboard? I have never understood that...

---

Those big "Bolts" are there to "Bolt" the Water Bock down to the Mother Board.
:D I Learnt That in College too :D

Quote:

---

Originally Posted by R3APER

Why do you have those big bolts sticking out of the motherboard? I have never understood that...

---

I suppose you use magic to affix your WB to the cpu? Perhaps you have little elves hold the damn things on for you? Or maybe, Ivanka Trump comes and begs you to let her apply 60 lbs of pressure to your WB? No, I bet that Heather Mills lends you her fake leg to wedge the WB into place :shocked:

Quote:

---

Originally Posted by nikhsub1

No I don't have calipers to properly measure the thickness of the EK o-ring... perhaps Alex could help us out here?

---

Hmm... Sounds like a job for my spiffy new Mitutoyo digital calipers. :)

Little did I realize how hard it would be to measure all of these damn o-rings (too squishy/tacky). The first measurement given is, well, what I measured and the second is what I'm guessing they're specced. as (assuming that their thickness spec. is metric).

EK 1/4" BSPP fitting (older version with larger collar/flange): 2.50mm (2.5mm)

EK 1/4" BSPP fitting (current version with smaller collar/flange): 2.43mm (2.4mm)

Swiftech nickel plated brass 1/4" BSPP fitting: 2.62mm (2.6mm)

Danger Den 1/4" BSPP "high-flow" fitting (older version w/thick o-ring): 2.55mm (2.5mm)

Danger Den 1/4" BSPP "high-flow" fitting (current version w/thin o-ring): 1.80mm (1.8mm)

D-Tek 1/4" BSPP fitting: 2.02mm (2mm)

Swiftech Nylon 1/4" NPSM fitting (Eldon James): 2.59mm (2.6mm)

EK 1/4" BSPP plug: 2.01mm (2mm)

EK thread spacer: 1.53mm (1.5mm)


So, how's that Scott? ;)

BTW, Brian offered to lend me his Sper manometer for pressure drop testing and I think that I'm either gonna take him up on that or buy my own, as I'm realizing that I don't have the precision needed for the FuZion in the lower flowrate range (talking <= 0.1psig @ 0.5GPM).

Quote:

---

Originally Posted by nikhsub1

I suppose you use magic to affix your WB to the cpu? Perhaps you have little elves hold the damn things on for you? Or maybe, Ivanka Trump comes and begs you to let her apply 60 lbs of pressure to your WB? No, I bet that Heather Mills lends you her fake leg to wedge the WB into place :shocked:

---

LOL,LOL To :banana::banana::banana::banana:in funny man!
You just made my day..:D

Quote:

---

Originally Posted by serialk11r

? Who said 1/4" sucked? :p: heh the base of blocks these days are 1/4" so I'd like to see your block beat the Fuzion just with 1/2" copper ;) Remember that the water running through makes super tall pins not so useful if you don't have ginormous amounts of heat going through...

---

i need it for my e6600 cus my apogee cant keep up with the heat output, it is currently idling at 47degrees C which is not very good:( i NEED that extra pin height.

Quote:

---

Originally Posted by p8ntslinger676

i need it for my e6600 cus my apogee cant keep up with the heat output, it is currently idling at 47degrees C which is not very good:( i NEED that extra pin height.

---

:P surface area closer to the bottom is more important ;) Just remember keep the base somewhat thin...and make the pins or whatever small, if you have the milling bits to do so :D good luck.
Oh back on topic, guys you can get O rings at home depot or orchard supply hardware...

looks like dd high flow (current) might be nice, either that or EK thread spacer, just a little bow

also i hope that the guys designing these waterblocks are using calculus to find max surface area dimensions... i know its the best way if you factor in the rate that heat from the cpu reaches different heights on the copper pins, double limit is bound to have an intersection that is ideal. maybe ill look into water block design for a science intensive study/project for the summer :)

Quote:

---

Originally Posted by SNiiPE_DoGG

looks like dd high flow (current) might be nice, either that or EK thread spacer, just a little bow

also i hope that the guys designing these waterblocks are using calculus to find max surface area dimensions... i know its the best way if you factor in the rate that heat from the cpu reaches different heights on the copper pins, double limit is bound to have an intersection that is ideal. maybe ill look into water block design for a science intensive study/project for the summer :)

---

There's a problem, you can't calculate exactly how the turbulence in the block is going to affect the performance...now there's patterns of pins that are effective, but I can imagine that its not possible to calculate exactly what pin shape and exactly what pattern will get you the best results...sure you can calculate the ideal height for the pins on a block like the fuzion but getting a pattern/design is the main issue...

Quote:

---

Originally Posted by nikhsub1

This question can NOT be answered as it depends.

---

Ok, what about stock fuzion with stock mounting hardware against a stock gtx also with stock mounting hardware ? I ask olny because I already have a GTX, it looks great but I really dont want to bow it. I have remounted several times but on average my idle temps were 41c, load is 52c with an ambient of 22c. My e6700 is running at 3.6ghz with 1.49 vCore. I have seen may people post temps lower with the fuzion while using MORE vCore. I'm currently getting the same temps with the GTX as I was with my orginal Apogee :confused: .

Quote:

---

Originally Posted by Darkatom

Ok, what about stock fuzion with stock mounting hardware against a stock gtx also with stock mounting hardware ? I ask olny because I already have a GTX, it looks great but I really dont want to bow it. I have remounted several times but on average my idle temps were 41c, load is 52c with an ambient of 22c. My e6700 is running at 3.6ghz with 1.49 vCore. I have seen may people post temps lower with the fuzion while using MORE vCore. I'm currently getting the same temps with the GTX as I was with my orginal Apogee :confused: .

---

I will say this again, unbowed Apogee's including the GTX are at best mediocre performers. Without the induced bow, the Fuzion KILLS the GTX and GT in my testing so far.

Thanks for the response, thats all I wanted to know and that is that the fuzion doesnt need any fancy bowing or mounting to better than a GTX. I guess i'll be ordering a Fuzion tomorrow.

On a side note, dont my temps seem a little off though ?

Quote:

---

Originally Posted by nikhsub1

I will say this again, unbowed Apogee's including the GTX are at best mediocre performers. Without the induced bow, the Fuzion KILLS the GTX and GT in my testing so far.

---

damn it stop saying that

you're forcing me to box my apogee :D

Gabe said dual cores ~2C difference which is not that great....i don't have quads so it is worth messing with it

Quote:

---

Originally Posted by serialk11r

There's a problem, you can't calculate exactly how the turbulence in the block is going to affect the performance...now there's patterns of pins that are effective, but I can imagine that its not possible to calculate exactly what pin shape and exactly what pattern will get you the best results...sure you can calculate the ideal height for the pins on a block like the fuzion but getting a pattern/design is the main issue...

---

this is true, there are many variables to consider, but it seems it could be done, i definitely know most of the calculus and the physics (calc. 2 and Harvard extension physics exp.)

thermodynamics, calculus, fluid dynamics, and a little computer modeling ;)

once you know the ideal surface area, pin height, and density you would have to consider the arrangement of pins very carefully. From my initial thoughts it seems that the pin design on the apogee is most apt for surface area. The d-tek has cylindrical pins which is not ideal, it does however compliment the high flow nature of the Fuzion.

It will definitely be interesting to explore the in-depth physics behind these blocks

Quote:

---

Originally Posted by dinos22

damn it stop saying that

you're forcing me to box my apogee :D

Gabe said dual cores ~2C difference which is not that great....i don't have quads so it is worth messing with it

---

Well the Apogee are great blocks, but testing proves the Fuzion is better if the apogees arent bowed. I for sure would not feel comfortable doing it.

Its sorta like, back then when they designed the internal combustion engine. Its not exactly possible to calculate the best shape for any parts of an engine...there's so many different possible designs. Infinitely many if you count small details, if I may add :) Just look at Mazda's rotary engine...does it look very much like your standard engine with pistons?
Anyways, pin shape and arrangement are a BIG thing...

Quote:

---

Originally Posted by serialk11r

Its sorta like, back then when they designed the internal combustion engine. Its not exactly possible to calculate the best shape for any parts of an engine...there's so many different possible designs. Infinitely many if you count small details, if I may add :) Just look at Mazda's rotary engine...does it look very much like your standard engine with pistons?
Anyways, pin shape and arrangement are a BIG thing...

---

ahhhh but did THEY have computers, i think not. ;)

sometimes it can be hard for us younger people to consider that at one point there were no computers to do anything

Does 7/16 ID, 5/8 OD tubing work on the stock barbs that come with the fuzion ?

Quote:

---

Originally Posted by Darkatom

Does 7/16 ID, 5/8 OD tubing work on the stock barbs that come with the fuzion ?

---

Sure, the tubing is plastic and can stretch over.

Well i have flow works so it calculates the thermodynamics and fluid dynamics of the blocks for me, its a cool program :D

Quote:

---

Originally Posted by p8ntslinger676

Well i have flow works so it calculates the thermodynamics and fluid dynamics of the blocks for me, its a cool program :D

---

It'll calculate it but you gotta give it a block design first ;) You can't tell it "I want water to flow like THIS" and let it draw a block for you lol.

Quote:

---

Originally Posted by Darkatom

Well the Apogee are great blocks, but testing proves the Fuzion is better if the apogees arent bowed. I for sure would not feel comfortable doing it.

---

YOu wouldn't feel comfortable doing what? Swapping the o-ring on an apogee? WTF not? Hmm, let's see, swap to fat o-ring which bows base = 4-5C better than using standard shipping o-ring. Naw, not worth it :rolleyes:

Quote:

---

Originally Posted by nikhsub1

YOu wouldn't feel comfortable doing what? Swapping the o-ring on an apogee? WTF not? Hmm, let's see, swap to fat o-ring which bows base = 4-5C better than using standard shipping o-ring. Naw, not worth it :rolleyes:

---

Let me see, adding a backplate to my 680i and screwing down the block so tight just to get the full effect of the bow which could potentially crush some important stuff on the back = NOT WORTH IT.

Backplates are over rated.

isn't the bowing part kinda, errm, irrelevant, and not the best method here.

hey i've been out of watercooling for ages but surely the increased performance is just from increased pressure. pressure = good, better contact. but people seem to be focusing on the wrong way to get that contact. the 4 points you introduce pressure from the hold down, only a certain amount of pressure really the board can safely be used with. so the same mounting pressure but its applied over a smaller area in contact with the cpu, thats all fine and dandy. but with the curve of the base you have just that a curve, only part of the base will be in perfect contact. that should be shown by the review i just read, the CES testing where 2 cores obviously obtained great contact and 2 cores didn't, because the curve was causing great pressure at one point.

if someone knows the actual location of the cores under the IHS, would the best way to get best pressure on ONLY the parts with the cores on, be to lap a tiny amount of the base away in the area's no in contact. hows best to explain it, like with a ram covering gpu stock sink, the memory/gpu contact area's are lower than the rest of the base of the sink. if you can find the exact area you want on the base of the waterblock, and lap away the rest a tiny amount, so that the whole cpu area is in contact, completely flat contact, and the rest of the base is ever so slightly higher, not in contact and none of the pressure is wasted on those areas?

bowing is giving half the effect, but also causing a curve which has to be giving the opposite effect. If someone came up with a good design, and you could buy just a new baseplate designed specifically for a certain cpu, $15 copper baseplate to give perfect contact area for your cpu. also, when you bend a metal you essentially disturb the lattice pattern of the metal so its not perfect anymore. a flat piece of metal curved to induce certain pressure points wouldn't be anywhere near as effective as a flat piece of metal with basically the edges up to where there needs to be contact, being slightly higher. i know the slight damage in the lattice of the metal would be tiny, and probably in reality not have any real effect on end results. but you add them all together and i would think that would be the best way to design it.

idea patented &#172;_&#172;

Quote:

---

Originally Posted by Darkatom

Let me see, adding a backplate to my 680i and screwing down the block so tight just to get the full effect of the bow which could potentially crush some important stuff on the back = NOT WORTH IT.

---

Uh, if you're not using a backplate as of now without a bow, that's a problem in the first place. The bow won't make your mounting mechanisms clamp the board any tighter.

Quote:

---

Originally Posted by drunkenmaster

isn't the bowing part kinda, errm, irrelevant, and not the best method here.

hey i've been out of watercooling for ages but surely the increased performance is just from increased pressure. pressure = good, better contact. but people seem to be focusing on the wrong way to get that contact. the 4 points you introduce pressure from the hold down, only a certain amount of pressure really the board can safely be used with. so the same mounting pressure but its applied over a smaller area in contact with the cpu, thats all fine and dandy. but with the curve of the base you have just that a curve, only part of the base will be in perfect contact. that should be shown by the review i just read, the CES testing where 2 cores obviously obtained great contact and 2 cores didn't, because the curve was causing great pressure at one point.

if someone knows the actual location of the cores under the IHS, would the best way to get best pressure on ONLY the parts with the cores on, be to lap a tiny amount of the base away in the area's no in contact. hows best to explain it, like with a ram covering gpu stock sink, the memory/gpu contact area's are lower than the rest of the base of the sink. if you can find the exact area you want on the base of the waterblock, and lap away the rest a tiny amount, so that the whole cpu area is in contact, completely flat contact, and the rest of the base is ever so slightly higher, not in contact and none of the pressure is wasted on those areas?

bowing is giving half the effect, but also causing a curve which has to be giving the opposite effect. If someone came up with a good design, and you could buy just a new baseplate designed specifically for a certain cpu, $15 copper baseplate to give perfect contact area for your cpu. also, when you bend a metal you essentially disturb the lattice pattern of the metal so its not perfect anymore. a flat piece of metal curved to induce certain pressure points wouldn't be anywhere near as effective as a flat piece of metal with basically the edges up to where there needs to be contact, being slightly higher. i know the slight damage in the lattice of the metal would be tiny, and probably in reality not have any real effect on end results. but you add them all together and i would think that would be the best way to design it.

idea patented &#172;_&#172;

---

you kinda don't get it do you.... the cores are fairly small and in the center of the chip. the bowing provides pressure only in a portion of the IHS, this is true, but the IHS is not what your trying to cool. you are cooing the CPU which is right under the center of the IHS.

the only thing the pressure achieves is to flatten the IHS to provide more fluid and complete contact with the actual cpu.

this noted, just because it is bowed doesn't mean there isn't still contact with the whole IHS. the contact is just more complete at the cpu area.

also the flat metal actually becomes concave if it is not bowed... ponder that

well, as seems to be shown the contact from the bowing doesn't seem to increase the surface pressure at "all" the right points for a kentsfield as the CES show test shows as two cores were significantly hotter than the other two. the bowing will always cause the central area to get the best contact. i was talking about increasing the pressure on the area's of the baseplate that correspond to where the cpu cores are under the IHS. a single core cpu is centered in the middle of the package, but a dual core and quad core aren't centered the same.

how about creating a C clip that attaches to the middle of the water block and on the same position at the rear of the board??? that would give you all the pressure you need exactly on the middle.

Its sort of like the athlon xp cpus that had no IHS. If you put the "spring" mount the wrong position it would not excerpt the force on the middle of the die but somewhere else and you would loose 3-4C. Maybe that is one of the reasons that the TT kit being so bad in its parts had decent cooling. That H type of affixing itself to the board is very very good.

Quote:

---

Originally Posted by the highlander

how about creating a C clip that attaches to the middle of the water block and on the same position at the rear of the board??? that would give you all the pressure you need exactly on the middle.

Its sort of like the athlon xp cpus that had no IHS. If you put the "spring" mount the wrong position it would not excerpt the force on the middle of the die but somewhere else and you would loose 3-4C. Maybe that is one of the reasons that the TT kit being so bad in its parts had decent cooling. That H type of affixing itself to the board is very very good.

---

well, the athlon xp mounting without IHS, thats basically the point. if you had a complete, and much larger flat surface than the size of the die, the die is still the only bit in contact and the whole die area has the full force of pressure applied over it(assuming good mount)> with the IHS this pressure is spread out over the IHS because, assuming cpu/ihs lapped, both are flat and the IHS does provide contact with the packaging and not cores so lets pressure be applied elsewhere. if you can remove the contact between baseplate and IHS in area's where the core is not under the IHS, you are wasting less pressure and getting more pressure onto the core area's. this IS whats happening with bowing, its essentially aiming more of the pressure at the centre point, so i would assume the method would(and i think it has) shown best results on a single core where the single core is in the very centre of the packaging, but dual and quad cores aren't quite like that. testing at CES showed that the bowing, while it helped as it no doubt concentrated the pressure in the centre, that worked out great for two cores, and less well for the others. i think my idea would work well for all the cores.

i don't know, also something strikes me as , not unsafe as such, but i don't really like the idea of having a base thats able to be bent.

for a quad you wouldn't use a bowed GTX, fuzion has the correct pre-bow for that chip

Quote:

---

Originally Posted by SNiiPE_DoGG

for a quad you wouldn't use a bowed GTX, fuzion has the correct pre-bow for that chip

---

Actually IHSes come in all shapes and sizes (sizes just to make the phrase more complete :D) (heat from soldering on the IHS warps them, or not...) :) They didn't make the kentsfield concave on purpose, and I'm betting there ARE some good intel IHSes out there (my CPU has one for instance...) and there most likely are conVEX kentsfield IHSes as well...If they made IHSes concave on purpose, then there would be a LOT of problems...

Seems a bodgy fix for slack intell c2d IHS and not much more to me.

Interestingly, almost all the internal combustion engine design ideas had been had by 1918. Very little new ideas after that. Heaps of development, metalurgy improvement etc but you can count the new ideas after that one one hand.

nikhsub1
anymore results to pass along

Quote:

---

Originally Posted by SNiiPE_DoGG

for a quad you wouldn't use a bowed GTX, fuzion has the correct pre-bow for that chip

---

You got that absolutely wrong there, buddy. I have both and have tried. The differential is big enough to be material.

Quote:

---

Originally Posted by IanY

You got that absolutely wrong there, buddy. I have both and have tried. The differential is big enough to be material.

---

well i was saying that an extreme bow is not useful for the quad because the cores are so spread out, idk the actual results, just logic on my part.

Quote:

---

Originally Posted by SNiiPE_DoGG

well i was saying that an extreme bow is not useful for the quad because the cores are so spread out, idk the actual results, just logic on my part.

---

They are really not very spread out, dude.

DUDE! lol idk this is just a conceptual thing for me at this point, and i do know that the quad distributes its heat load over much more of the IHS than a core2 does. i was just saying that the fuzion probably does better on the quad, but seeing as we are splitting 2 C hairs here just lay off

Take a look at this as to how far apart the dies are from each other.

Quote:

---

Originally Posted by CrazyJoe

Here's a couple images I sketched up at work today for reference use so anybody else can join in on all the fun. :thumb:

http://www.stealth-and-mutilation.com/ocnet/core.jpg
The dotted circle in the center is a 1/2" inlet barb for reference. The outside 4 black dots are the mounting holes on the motherboard.

http://www.stealth-and-mutilation.co.../core_spec.jpg
This is a representation of the
* 62mm square block
* 29mm square IHS
* Both core 2 Duo die's within the QX6700/QX6800

Measurements are approximations but should be within .5mm at the most.

Hope this helps some of you ! :D

---

http://www.overclock.net/2104912-post78.html

And an actual image:

Quote:

---

Originally Posted by Knitelife

Actual image of a Quad Core without its IHS

http://www.thetechrepository.com/att...1&d=1160341004

---

http://www.overclock.net/2101989-post50.html

I'm wondering if the better center contact patterns shown using the bowed blocks could be duplicated using a flat block and lapping that CPU, along with a corresponding decrease in the amount of paste used. I see a pattern in the paste shown in the photos that suggests the outer edge of the IHS is lifting the water block off of it.

Alot of guys that bow the base already lap their IHSes. One of the reasons that the bowed base helps is because it puts most of the pressure in the direct center of the IHS.

Why not have a raised section machined into the baseplate, like on the GPU waterblocks...aka MCW60's ??

Quote:

---

Originally Posted by NEvolution

Take a look at this as to how far apart the dies are from each other.


http://www.overclock.net/2104912-post78.html

And an actual image:


http://www.overclock.net/2101989-post50.html

---

You may have spotted this already, but the image posted by Knitelife claiming to be a quad core without it's IHS is actually a pentium D presler. The orientaion of the dies is wrong for a quad core Kentsfield. Another image is avalialable @ http://tweakers.net/reviews/614/18 which is a weblog of the Intel developers forum spring 2006.

PCTwin
FYI that presentation was made by Sampsa and all those cpus are some not working extremely old first silicon ; )

Quote:

---

Originally Posted by \Karting_freak

PCTwin
FYI that presentation was made by Sampsa and all those cpus are some not working extremely old first silicon ; )

---

Even so, all the reviews and images I've seen show the orientaion of the cores to be opposite to the ones in knitelife's image. If I am going to bow my apogee gtx i want to get the orientation of the bow right. This review http://www.techpowerup.com/reviews/Intel/QX6700/2 credited to D.O.S also has an image of a qx6700 without ihs, so which one is right.

As both Nikhsub1 and IanY have said (having both the GT and GTX blocks) the orientation of the GT does matter, while the orientation of the GTX does not.

EDIT: The Tech Power Up review has the correct picture of the Kentsfield cores.

would a bowed fuzion help reduce temp on a de-ihsed conroe or penryn?

^^ I think it would just crush the core. Lol.

But the worrying part about Bowing is that once Bowed, you cant go back to Flat.
I still dont understand why though? Wouldnt it just be that the Copper is somewhat "Flexing" outwards, and non bowed, it would just flatten back. I guess im just a noob though. Lol..

I still would prefer Lapped IHS + Lapped HS.

Quote:

---

Originally Posted by Ad1tya

But the worrying part about Bowing is that once Bowed, you cant go back to Flat.
I still dont understand why though? Wouldnt it just be that the Copper is somewhat "Flexing" outwards, and non bowed, it would just flatten back. I guess im just a noob though. Lol.

---

You are bending the copper in one direction. when you bend it back you are not doing it in a press so it doesnt bend back exactly the way it was bent in the first place. This causes the surface of the block to be wavy.

If you were to lay the flat side of the block on a piece of granite and then put another piece of granite on top if it and then stand on it (with the top of the waterblock off) you could probably get it flat again, baring you crushing the pins. But even if you did that you would want to lap the base to ensure that the block was indeed flat.

can someone please be kind enough and let me know what is the exact size and measurements of the o-ring i will need to bow my original Apogee guys?

i'll head down to the hardware store to try and find one with measurements provided by you guys :) :up:

just cough up the money and get teh gtx/fuzion ;)

Quote:

---

Originally Posted by dinos22

can someone please be kind enough and let me know what is the exact size and measurements of the o-ring i will need to bow my original Apogee guys?

i'll head down to the hardware store to try and find one with measurements provided by you guys :) :up:

---

i don't think you can because the o-ring is not in the middle of the block ...

i pulled it apart and took the o-ring out

will head out and find a 3mm thick one and that's it i think :)

Quote:

---

Originally Posted by dinos22

i pulled it apart and took the o-ring out

will head out and find a 3mm thick one and that's it i think :)

---

dinos,

The 3mm O-ring is what you need. :up:

Quote:

---

Originally Posted by Hard H20

dinos,

The 3mm O-ring is what you need. :up:

---

yeah did some searching and found that stock one is 2.5mm and the bigger one is 3mm

but no one actually states what size the damn ring is lol so i have to take one physically to the store and see if they can find the same :(

farking hell can't find it in a couple of major hardware stores and even a dedicated rubber store

anyone able to source one for me

Quote:

---

Originally Posted by dinos22

farking hell can't find it in a couple of major hardware stores and even a dedicated rubber store

anyone able to source one for me

---

PM gabe, he will hook you up!

Woah lots of good infos down here! I'm having some probs in here with my FuZion not wanting to make good contact on my new mobo with my lapped E6600 ... Great infos! Thanks!

Went through the whole thread BTW ... And as always, i'm late on stuff.

Quote:

---

Originally Posted by Hard H20

PM gabe, he will hook you up!

---

i did a few days ago
no response

@nikhsub1

Concerning the fuzion o-ring your using (ek ring)

I would go to Home Depot and pick up a set of washers from the plumbing section. They have an assorted box with varying heights and 1/2 id on them. They are all flat so it should work well for your tests.

Thanks for the info btw. I will stick to my polarflo for now and probably my next block will be a fuzion. But thats a ways off.

So, how do I get this bow thing?

Which way do I tighten down the screws? The way I see it, you can have your bow in two directions, which one is the best?

Quote:

---

Originally Posted by D_o_S

So, how do I get this bow thing?

Which way do I tighten down the screws? The way I see it, you can have your bow in two directions, which one is the best?

---

pay close attention the the pics in this link

as there is only one proper way to install the top :)

http://www.cooling-masters.com/articles-45-1.html

translation, sort of
http://translate.google.com/translat...3Doff%26sa%3DG

Quote:

---

Originally Posted by dinos22

pay close attention the the pics in this link

as there is only one proper way to install the top :)

http://www.cooling-masters.com/articles-45-1.html

---

Ya, I realize that there is only one way to install the top...

However, when I tighten the screws, I can get a bow in two directions, can't I?

Screw pattern:
TL-----------TR
BL-----------BR

Scenario A:
I tighten down TL-BR-TR-BL.... I get a bow going from where to where?

Scenario B:
I tighten TL, BL, TR, BR - bow from left to right

Scenario C:
I tighten TL, TR, BL, BR - bow from top to bottom....

So, which one? Or have I got something wrong?

Quote:

---

Originally Posted by D_o_S

Ya, I realize that there is only one way to install the top...

However, when I tighten the screws, I can get a bow in two directions, can't I?

Screw pattern:
TL-----------TR
BL-----------BR

Scenario A:
I tighten down TL-BR-TR-BL.... I get a bow going from where to where?

Scenario B:
I tighten TL, BL, TR, BR - bow from left to right

Scenario C:
I tighten TL, TR, BL, BR - bow from top to bottom....

So, which one? Or have I got something wrong?

---

I would like to know this answer as well

Moin !

@Nikhsub1:
Any idea of when the final results will be finished ? I have to change my Waterblock and for this need the results. Any chance of getting them in the next 2 weeks or so ?

Don't let yourself get hasted but do it quickly ! :D

http://www.cooling-masters.com/image...ges/fleche.png

this sealed the deal for me! all i needed to see right here in this one pic