For all you copper vs. aluminum techies...

[pHaestus]

There have been several tests of the effects of baseplate thickness. Check out Aesik's posts on o/cers forums for some modeling of aluminum vs. copper, BillA has an article on coldplate thickness, and Cathar on oc-au forums has some test data for diffterent baseplate thickness on the same block (Cyclone5 I think). Bill has a lot of different bps for the old MCW462 blocks but I am not sure if he ever posted the test results for them.

Maskedgeek:

weres your proof besides the numbers and theory, pleze show me

I hope you can reread this comment and realize how silly it sounds. Theory in this case is basic thermodynamics. You would be the one needed to provide proof that principles of thermo doesn't work, not the other way around.

[morphling1]

Maskedgeek, let me be the one to show you a test I made long time ago with very basic equipment, and if I think back not realy quite accurate, but it can still give you a ballpark of the difference between two exactly the same blocks 6mm AL spiral and 6mm Cu spiral , it is around 5°C with that testing Duron, and if I would test with something hotter like overvolted Axp the difference would be even greater here: http://forums.procooling.com/vbb/sho...&threadid=2183

[Can O' Beans]

In response to the original idea:

I think that it would work, for initial design purposes, as AL is a LOT easier to work with & a hell of a lot cheaper. Once you get your best designs in AL, you can transfer that design to CU and THEN start tweaking the design to work better with the properties of CU.

So basically, you use the cheaper/easier AL initially to test & work torward a better design, then transfer that to CU and continue on until you got it right.

[jaydee]

Originally posted by morphling1
Maskedgeek, let me be the one to show you a test I made long time ago with very basic equipment, and if I think back not realy quite accurate, but it can still give you a ballpark of the difference between two exactly the same blocks 6mm AL spiral and 6mm Cu spiral , it is around 5°C with that testing Duron, and if I would test with something hotter like overvolted Axp the difference would be even greater here: http://forums.procooling.com/vbb/sho...&threadid=2183

It depends on the design and the eqipment used in the testing. I made 2 identical blocks in AL and Cu (thank god for CNC!) and came up with a .5C difference between the two. It was the basic e design but whatever. AL works just fine for computer water blocks. Sure Copper performs better but who cares, .5-5C makes no real world difference. That is like overclocking a 2000mhz CPU to 2025mhz. Wooopteeeeedoooooo!!!!!!

[maskedgeek]

Originally posted by jaydee116
It depends on the design and the eqipment used in the testing. I made 2 identical blocks in AL and Cu (thank god for CNC!) and came up with a .5C difference between the two. It was the basic e design but whatever. AL works just fine for computer water blocks. Sure Copper performs better but who cares, .5-5C makes no real world difference. That is like overclocking a 2000mhz CPU to 2025mhz. Wooopteeeeedoooooo!!!!!!

EXACTLY WHAT IM TRYING TO SAY!!!!!

[Jessfm]

Identical Neptune blocks .
with a Low heatload, the ALu was not too bad, 2-3°c difference between the Cu/Alu version at 50W.
The Alu sucked at 80W load, 8-10°c differnceial, and things got nasty after 100W so i stpped testing.
Both CNC'd
I have hacked the Block now, but Ill get some pics later.
Yes It was part of the R&D on the block - I wanted to check viablility to make a budget block , it was that bad I scrapped the whole idea.

[BillA]

now things are indeed becoming clear

if low-tech is good, is no-tech not even better ?

why not just piss on a brick ?
(with suitable low-tech plastic sheeting of course)

why would an OCer even consider buying stuff from you guys ?

(not you Jess, those other wannabes)

[maskedgeek]

*edit* derogatory statement removed *edit*

[maskedgeek]

*derogatory statement removed*

[pHaestus]

This thread has some merit, but the personal attacks do not. I am going to clean it up and if it happens again I will Lock it Down.

[BillA]

why get upset over the term 'wannabe' ?

we are all wannabes by one measure or another,
I, for example, wannabe a better tester; and I work diligently at it

when someone posts here wannabeing a wb mfgr/vendor, then an evaluation WILL be made of their efforts;
in design, materials, proof-of-design testing, manufacturing methods, inspection/QC, and - oh yes - incidentally the price
-> but most here (and elsewhere) are interested in performance
(else the Dell box is just fine, eh ?)

when someone is observed who:
doesn’t know, won't listen, can't learn, complains about criticism, and persistently chooses the 'low performance' path -- what is one to think ?

come on ‘geek, get a grip
if you wish to make cheap stuff that's fine by me, just don't peddle your ignorance as technical 'knowledge'

it ain't 'all the same'

EDIT: removed my comment about personal invective

[maskedgeek]

i hope you saw my posts bill, but as of now im out of this craphole

[Joe]

wow someone’s taking his toys and going home it looks like!

Sorry man, one of the reasons the forums are useful is the mods/staff keep stuff above a 8 year old mentality around here. There are plenty of forums where you are expected to be immature... this isn’t one of them.

If you don't like that the mod removed your inflammatory and overall immature remarks, then I think it is best that you don’t bother posting here.

EDIT:
Hey Everyone!
Stop the personal attacks. Everyone. Bill, Geek, and others in this thread seem to see it as a suitable form of debate, but its not.

[jaydee]

Originally posted by unregistered
w persistently chooses the 'low performance' path --

The point that you performance freeks are missing is the difference between the low performance and the high performance blocks are only a few C difference. Which makes no difference in real world computing or overclocking. What are you all tring to gain by going a couple C cooler????????

[gone_fishin]

As Jessfm pointed out, it may be a couple C diff at stock speed but the problem is magnified when the wattage is cranked up.

[bigben2k]

Jaydee, I see your point, and I agree, but I think that the new Thoroughbred core is going to make us throw out a lot of ideas.

I would think that the fact that AMD specifications now require all HSF bases (not the fins) to be of copper, would be a big clue.

But hey, that's just my .02$ ...

[morphling1]

What's the point if you can't be convinced otherwise.... but even 5°C means a lot of difference in watercooling and therefore overclocking, and the difference only grows with larger heat output not even counting smaller die ....

[Fixittt]

HEY IM ONE OF THE WANNABES

I have tried to follow this tread and keep my mouth sut on alot of things. And I think I have done a good job. Its no secret that I design the blocks. And my experiance is that copper is just a small amount better then aluminum. Sure all the techy babble is there. but I dont reley on post, or published words. I have to see for my self. Im a hands on kind of guy. Personally I did not get any speed increase or higher overclock with my existing hardware. Sure it ran a few degrees hotter, but was still stable at my highest clock speed thure all tests. But one thing people have to look at who make and sell these things. Aluminum works great. I have proven to myself that fact. Whos gonna buy aluminum blocks? Not many people. So if you want to sell something, yolu have to sell what people are going to buy. Hence the copper based product. I would much reather work with aluminum all day long, then mess with one peice of copper. But oh well.

I to this day am running an aluminum based water block. And will not change. Unless future changes warrant it. Also anyone who is wanting to design something for todays hardware is going to be re-designing when tomarrow comes. So think ahead a bit. plan, so you can still be having fun tomarrow.

My little pennies worth.

[jaydee]

Originally posted by bigben2k
Jaydee, I see your point, and I agree, but I think that the new Thoroughbred core is going to make us throw out a lot of ideas.

I would think that the fact that AMD specifications now require all HSF bases (not the fins) to be of copper, would be a big clue.

But hey, that's just my .02$ ...

I am not just talking about AL vrs. Copper I mean even the lower performance Copper vrs. Higher performance Copper aswell. A few C will not help anything.

The only thing the T-Bred core will change in design is the mounting, other than the the same design that are out now will suffice as they seem to work just fine with 220watt pelts and I highly doubt T-Breds will put out that much heat. Waterblocks can't get much better without active cooling period.

Originally posted by morphling1

What's the point if you can't be convinced otherwise.... but even 5°C means a lot of difference in watercooling and therefore overclocking, and the difference only grows with larger heat output not even counting smaller die ....

5C does not mean a lot of difference. I been OCing for 7 years and 5C means jack anymore. It may get you 50mhz but what good is 50mhz on a 2000mhz stock CPU??? If you want to get REAL results overclocking you will have to use active cooling otherwise it is a waste of time. 5C used to gain you 200mhz sometimes but not anymore. You right there is no convincing me as I been there done that and all for not much in the end. All that water cooling is good for is midrange overclocking and quiter computers.

Also the T-Bred is supposed to have the head spreader so die size is not much of an issue.....

Originally posted by fixittt
HEY IM ONE OF THE WANNABES

I have tried to follow this tread and keep my mouth sut on alot of things. And I think I have done a good job. Its no secret that I design the blocks. And my experiance is that copper is just a small amount better then aluminum. Sure all the techy babble is there. but I dont reley on post, or published words. I have to see for my self. Im a hands on kind of guy. Personally I did not get any speed increase or higher overclock with my existing hardware. Sure it ran a few degrees hotter, but was still stable at my highest clock speed thure all tests. But one thing people have to look at who make and sell these things. Aluminum works great. I have proven to myself that fact. Whos gonna buy aluminum blocks? Not many people. So if you want to sell something, yolu have to sell what people are going to buy. Hence the copper based product. I would much reather work with aluminum all day long, then mess with one peice of copper. But oh well.

I to this day am running an aluminum based water block. And will not change. Unless future changes warrant it. Also anyone who is wanting to design something for todays hardware is going to be re-designing when tomarrow comes. So think ahead a bit. plan, so you can still be having fun tomarrow.

My little pennies worth.

Same here....... Well said.

[morphling1]

To get to essence of this thread, could the very best desing in Al be transfered to Copper and get the same best performance, NO never, the differneces in thermal properties are just to big and the block would act differently, even optimal thickenss for those two block is different, so I don't think prototyping in AL help you in Cu. The only thing that help is machine testing your cnc program if you need to but I would use wood here.

And I can understand you two, but shoot me I'm a performanse freak and I know that those few degrees will get me higer and faster. Just one info Fixxit or Jaydee, have you two ever run athlon xp close to 2GHz at 2.16V on Aluminium block, NO WAY in the world could that be done, if I wouldn't have good designed copper block, and running at those speed stable at 37°C socket thermistor, 48°C in die diode full load, and those temps. are making me calm, no way I could be confortable running close to those speed with aluminium 24/7, 10 or more degrees higher.
And that my 0,02$

[jaydee]

Originally posted by morphling1
Just one info Fixxit or Jaydee, have you two ever run athlon xp close to 2GHz at 2.16V on Aluminium block, NO WAY in the world could that be done, if I wouldn't have good designed copper block, and running at those speed stable at 37°C socket thermistor, 48°C in die diode full load, and those temps. are making me calm, no way I could be confortable running close to those speed with aluminium 24/7, 10 or more degrees higher.
And that my 0,02$

Well yeah actually. Not a XP but a T-Bird 1400@1705 2.20Vcore on the Epox 8K7A. That is roughly 120watts and the AL e design block did just peachy at keeping it 45C load 24/7 crunching SETI. People are using 220watt pelts with AL chillers and doing rather well to. And remember industrial applications use AL with great success in more extream situations than computers. I think you seriuously underestimate AL.....

I find larger channels work better with AL and smaller channles (much like your sweet multi channel spiral) work better with copper. Thats probably why you see such extream temp differences in your AL vrs. Copper blocks.

[Jessfm]

Interestng thread for the most.

Cheers Mr.Adams btw -although you know Im still a wanabe-wanabe realy

I would be more then happy to manufacture in Alu, but rem Copper comes in lots of grades, and some have allot of crap in there making the difference between Alu and Cu fairly close.

jaydee116

-> I make your Tbird @ closer to 140W TBH, but anyway. Seems pretty good to have a Alu waterblock keep your CPU temp @ 45°c.
That would put your block under 0.2°c/w with water temp of around 20°c. Got any more info to help us out on this pls ?
Neptune Alu was more like 0.4c/w for me.
Cheers

[pHaestus]

Jessfm:

Try moving the compunurse probe a little further from the core or else you could bend the in-socket thermistor down a bit. Temps should improve then.

[morphling1]

Ok, I won't argue with you, but how much flow did took to do that, (that was probably that center inlet, two outlet block right), but still did you test at the same condition with the same copper block and see the difference of just few degrees?
And that large chanell, small chanell labeling for block is just not good. It's all in the basic equation for heat transport it consist of heat conductivity (material properties) and heat transfer (preasure,temperature, velocity properties of the tested fluid, and shape and quality of the surface properties). So with large chanell you get at the same pump output better flow-> higher velocity ->better heat transfer so that part of the equation become more important for total heat transport then the part of heat conductivity , that's what makes you think that aluminium is better, but never the less cooper in same condition would still be better but the difference would be smaller, then in second case with small chanells where water flow would drop, and material properties (heat conductivity) begins to be more important in that equation, because it comes to heat spreading effect through the fins and whole block and here is copper supperior to AL.
I can't put it any simpler I have enough trouble explaining this from my language to english.

But to the utabintarbo main point, do you also argue my take on that, do you think that designing top AL block would result that the same Cu block would be top? This certanly won't be the case

[Jessfm]

Originally posted by pHaestus
Jessfm:

Try moving the compunurse probe a little further from the core or else you could bend the in-socket thermistor down a bit. Temps should improve then.

Ah thats where I go wrong !!

Cheers

Whats that old saying, Ignorance is bliss, or something like that.
I shall squint next time at my Fluke/Thermocouple readouts