View Full Version : Another Full Cover WB (GF6800)
Eddy_EK
05-08-05, 03:30 PM
The idea for GF6800 WB came itself.
I like GeForces because of one-side placed rams and therefore I find it easyer to construct and make WB to cover the rams.
I got the inovatek WB few days ago in my hands and I took the distances between holes for mounting the WB.
The concept of WB is that where the high preformance coling isn't needed (rams) I will use aluminium. And for GPU processor I will use copper. The combination of material is mostly for the weight reduction and easier to work with.
The aluminium part of the WB is (will be) eloxed or nickel plated, therefore there will bo no conflicts with copper. The silicon gasket is used for sealing between copper and aluminium. It is resistant on temperatures till 250°C.
The block weights 405 g.
Any comments :)
ludeboy12
05-08-05, 03:34 PM
looks good.....looking forward to seeing how well it performs....
also how big is it?? cant really tell from the pix if its small enough for sli....
jamesavery22
05-08-05, 04:47 PM
looks good.....looking forward to seeing how well it performs....
also how big is it?? cant really tell from the pix if its small enough for sli....
If you cant mount barbs so they are facing either way (either screwed into the top or screwed into the base) then chances are it wont work for SLI.
ludeboy12
05-08-05, 04:59 PM
couldnt you use an L shapped barb though.....hmm....well not sure if anyone makes one now....
well if those exist it could easily be implemented but how big is the acual wb still....
by the looks of it im sure it would at least fit on the asus sli board...
jamesavery22
05-08-05, 05:14 PM
couldnt you use an L shapped barb though.....hmm....well not sure if anyone makes one now....
well if those exist it could easily be implemented but how big is the acual wb still....
by the looks of it im sure it would at least fit on the asus sli board...
Yeah there are 90 degree barbed x threading fittings. Im sure Mcmaster has some. But dont forget one barb will have to be short enough to alow it to thread/turn in when there is already one barb installed. PLus 90 degree fittings are no-no's b/c of the pressure drop.
tyrant222
05-09-05, 12:55 AM
seriously cool
outhouse
05-09-05, 01:13 AM
where and how much
Eddy_EK
05-09-05, 03:37 AM
Hi! I'm glad you like it :)
The WB is 19 mm thick (9mm Al, 10mm acr. top)
The 90° barbs would do.
If anyone has concrete question please contact me by PM oe e-mail. :welcome:
looks great :) one question though. wouldn't the water be able to flow over the central island part since it is not in contact with the top because of the rubber seals around the edge but not the middle? maybe i'm just imagining it or maybe it just doesn't matter
clocker2
05-09-05, 05:08 AM
Beautifully made, Eddie.
Your flowpath through the copper section looks more efficient than the NV-68 that I have.
What does the retaining mechanism look like?
Eddy_EK
05-09-05, 05:08 AM
looks great :) one question though. wouldn't the water be able to flow over the central island part since it is not in contact with the top because of the rubber seals around the edge but not the middle? maybe i'm just imagining it or maybe it just doesn't matter
No, there is nothing to be afraid. It is close enought.. Almost perfect fit :)
gigabit
05-09-05, 01:03 PM
.Thats Freaking AWESOME dude:drool:
Silversinksam
05-09-05, 01:11 PM
Hi! I'm glad you like it :)
The WB is 19 mm thick (9mm Al, 10mm acr. top)
The 90° barbs would do.
If anyone has concrete question please contact me by PM oe e-mail. :welcome:
I have one question, if they go on sale will they be cheaper than DangerDen's NV-68 (http://www.dangerdenstore.com/product.php?productid=64&cat=48&page=1) which has an msrp of $124.95
PS, The waterblock looks really well thought out.
Voodoo Rufus
05-09-05, 01:13 PM
The sealing trick sounds interesting. I like the "diverging-converging" area where the water goes to the copper, similar to what I've been learning about aircraft radiators. Reduces drag.
Nicely made.
Eddy_EK
05-09-05, 03:34 PM
I have one question, if they go on sale will they be cheaper than DangerDen's NV-68 (http://www.dangerdenstore.com/product.php?productid=64&cat=48&page=1) which has an msrp of $124.95
Definitely, but the series will be limited.
sunrunner20
05-09-05, 04:22 PM
have you considered making the flow conections straight instead of having a 90 digree turn? (which would also make it easyer for SLI.)
Hi! I'm glad you like it :)
The WB is 19 mm thick (9mm Al, 10mm acr. top)
The 90° barbs would do.
If anyone has concrete question please contact me by PM oe e-mail. :welcome:
does this mean ur a rep for whoever is making that thing? looks cool
EDIT: heh just saw ur sig...COOL! we got a new rep! :) welcome dude.
clocker2
05-09-05, 04:28 PM
have you considered making the flow conections straight instead of having a 90 digree turn? (which would also make it easyer for SLI.)
That would require one of the layers (either the aluminum or the acrylic) be thick enough to accept the drilled/tapped holes.
Maybe the solution would be to just drill/tap both sides and use plugs to seal the unused pair.
lithker
05-09-05, 04:45 PM
do u still have to use the power regulaters heatsink? i see there is a narrow strip on the one side with holes
Eddy_EK
05-09-05, 05:13 PM
do u still have to use the power regulaters heatsink? i see there is a narrow strip on the one side with holes
At GT and Ultra there is not needed. It must be demontaged in order to place the WB.
After trying for awhile to convert Slovenian money to USD, I found their prices, on their website, to be comparable to DD. I'd expect that the above would be cheaper than DD's offering just because it isn't all copper. The shipping seemed pretty reasonable too considering. About what DD charges (slightly more, like $1).
They have some pretty Bling Bling blocks on their site, including a lucite block with a color LED inside. I'm sure somebody has been dying for something like that.
Here the site (http://www.cerkovnik.si/indexh2o_e.htm) if you're interested.
^^wow yeah pretty nice stuff on that site, great looking CPU blocks. I like the RAM sinks alot too lol.
Eddy_EK
05-10-05, 03:04 AM
After trying for awhile to convert Slovenian money to USD, I found their prices, on their website, to be comparable to DD. I'd expect that the above would be cheaper than DD's offering just because it isn't all copper. The shipping seemed pretty reasonable too considering. About what DD charges (slightly more, like $1).
They have some pretty Bling Bling blocks on their site, including a lucite block with a color LED inside. I'm sure somebody has been dying for something like that.
Here the site (http://www.cerkovnik.si/indexh2o_e.htm) if you're interested.
I must say that my VGA WB are better than DD, try to look at the WB cooling area. The bottom base is copper. The asrylic top is shaped for gpu ram sinks can be used. Try to put ram sinks with DD VGA VB :cool: That was the main aim while planning the WB. And I like to show the inside of WB, so if you use UV active coolant the image is much better :)
Cerberus2k7
05-10-05, 03:32 AM
OMFG...
That block just looks simply amazing! But are you going to have any spots over where the RAM is to make the water slow down a bit a cool better?
Eddy_EK
05-10-05, 03:59 AM
OMFG...
That block just looks simply amazing! But are you going to have any spots over where the RAM is to make the water slow down a bit a cool better?
The fact that water is going over ram is good enought because it don't need extreme cooling. Why would I make channels to slow down the water and make WB more resistant... :)
Cerberus2k7
05-11-05, 01:59 AM
point taken. SLI compatable? And just *how* restrictive is this block?
Eddy_EK
05-11-05, 06:36 AM
point taken. SLI compatable? And just *how* restrictive is this block?
The block is going to be sli compatable. I will make thread in acrylic top tor barbs, and the unused thread will be pluged with a plug like on this picture:
http://www.dangerdenstore.com/product_image.php?imageid=290
If you look at the width of the channels you can tell that block won't be much restrictive.
Flip-Mode
05-11-05, 07:01 AM
Dibs! LOL
DAmn!
now im interested in WC'ing :D
this thing is simply amazing esp since itll workin sli!
Cerberus2k7
05-11-05, 02:14 PM
DAmn!
now im interested in WC'ing :D
this thing is simply amazing esp since itll workin sli!
Youre telling me. This and Cathars block are the only ones i'm actually checking in on daily because i'm loving the designs.
Eddy_EK
05-12-05, 09:57 AM
OK!
I got aluminium part with nickel coating (nickel plated).
I am sorry but at the moment I can't give you any picture, but will tomorrow.
The aluminium block loocks well coated so I think I/we have solved that problem.
Right now I am looking for small springs for mounting the block.
clocker2
05-12-05, 10:30 AM
Right now I am looking for small springs for mounting the block.
Man o man, I hate spring mounting for blocks.
Have you looked into some form of cam-lock or any other option?
Spring mounting is so uncertain and difficult to remount with consistency.
As beautifully thought out as the block itself is why revert back to basic hardware store cobbling to mount it?
Eddy_EK
05-12-05, 11:51 AM
Man o man, I hate spring mounting for blocks.
Have you looked into some form of cam-lock or any other option?
What do you mean with cam-lock. Don't know the expression. :shrug:
What would you recomend. Just screw? Any pictures?
clocker2
05-12-05, 01:25 PM
Many of the new stock AMD heatsinks use a cam lock to apply pressure to the mounting clamp.
I like the concept since the clamping pressure is either right or wrong...no grey area.
There is also the advantage of reproducable mounting pressure, no guesswork involved.
I can see why the screw/spring/wingnut assembly is so popular...it's so easy to provide, but the level of sophistication falls short of the waterblock itself.
Whenever there is thread involving bad temps, the invariable first response is that the block is not properly seated, which probably happens due to the hit or miss nature of the mounting.
It would be a pity if all your work went for naught due to poor customer installation.
Eddy_EK
05-12-05, 02:13 PM
It would be a pity if all your work went for naught due to poor customer installation.
I can agree to that. In fact to all. But the mounting clamp can not be done with that block.
I might combine spring for GPU and for other mountings just screws with silicon distancer. (I hope you know what I mean)
Eddy_EK
05-13-05, 11:23 AM
Hi!
As I promissed, here are the pictures of nickel plated aluminium part.
I better like the aluminium look, but that can't be made...
Anyway, it looks very shiney. :D
clocker2
05-13-05, 01:12 PM
It looks beautiful.
Do you have any pictures of water flowing through it?
I'm curious about the flowpath over the GPU...my NV-68 seems to have issues in this area.
Also, does it matter which direction the water flows through?
The NV-68 seems to like one direction better than the other, but your design appears to be pretty symetrical.
When might these be available?
Eddy_EK
05-20-05, 11:11 AM
Hi!
Here are the pictures of FC water block of leadtek GF 6800 Ultra.
This was just for mountnig test.
http://slike.slo-tech.com/13963.jpg
http://slike.slo-tech.com/13964.jpg
http://slike.slo-tech.com/13965.jpg
I tried to mount with springs, but I think I will use just screw with silicon distance ring, which will compensate the force.
http://slike.slo-tech.com/13966.jpg
http://slike.slo-tech.com/13967.jpg
Silversinksam
05-20-05, 01:01 PM
In a word.... Awesome!!
clocker2
05-20-05, 01:31 PM
Now all you need to make it perfect would be to drill/tap the upper plate also.
This would be necessary for SLI anyway and would offer options to the single card user, too.
Skeptical.
Aluminum plus copper, meh.
I personally run all distilled water in my system, and the mix of aluminum (or nickel) would definately require some anti-corrosive. Plus, why nickel plate it :shrug:. Sure it looks pretty, but, nickel's thermal properties aren't great.
Eddy_EK
05-20-05, 02:48 PM
I personally run all distilled water in my system, and the mix of aluminum (or nickel) would definately require some anti-corrosive. Plus, why nickel plate it :shrug:. Sure it looks pretty, but, nickel's thermal properties aren't great.
The nickel is not the problem nither on the corosion nor on the thermal properties. Because the GPU, who needs the best cooling has copper above, and the ram's don't need that much cooling. The fact that water is near is good enought for them. The nickel plating on aluminium is 12 microns thick.
Anyway Knivez, I respect your point of view.
Flip-Mode
05-22-05, 12:39 AM
Very nice, when are they going into production?
Ad Rock
05-22-05, 12:45 AM
Skeptical.
Aluminum plus copper, meh.
I personally run all distilled water in my system, and the mix of aluminum (or nickel) would definately require some anti-corrosive. Plus, why nickel plate it :shrug:. Sure it looks pretty, but, nickel's thermal properties aren't great.
90% of us who use watercooling have a mixture of Alu and Copper in our systems. The cost of making the whole block out of copper would be far to great to justify the benefits I bet.
This is an amazing peice of art, if only I had a Nvidia card I would be drooling even more.
90% of us who use watercooling have a mixture of Alu and Copper in our systems. The cost of making the whole block out of copper would be far to great to justify the benefits I bet.
This is an amazing peice of art, if only I had a Nvidia card I would be drooling even more.
90%?
Well I for one run no alum in my line, expcept for my T-line fillport. But water isnt constantly being moved through it.
90%?
Yeah I thought that was the major no-no in regards to watercooling.
where and how much
Very nice, when are they going into production?
When might these be available?
sicloan
05-22-05, 03:34 PM
Looks like a very nice lightweight solution. Cant wait to pick one up!!
Eddy_EK
05-22-05, 04:25 PM
Yes, yes! Im still here...
The developement is in progress, I am deciding on right mounting mechanicm wich will preform good eaven with noob :cool:
And finishing some details...
Eddy_EK
07-01-05, 12:06 PM
Well, I must tell you that I had serious problems with nickel coated aluminium part of WB. :(
The aluminium part was nickel plated once, as the company forgot my instructions, therefore they had to denickel it and nickel plate it again.
But the nickel coating was no good at all. :shrug:
I put ordinary water in block and the problems were seen in few days. I am sorry to have listened to some guys who said that eloxation will not be good enought. I think, that as many companies uses aluminium parts, it should be good for me too.
So I will try to denickel it again and than try to eloxate it. And then test it again.
Cerberus2k7
07-05-05, 09:58 PM
Gonna be making these for the 7800s also? :D
Eddy_EK
09-05-05, 02:48 PM
I had problems, and I sloved them.
I denickeled the aluminium parts and then I eloxate them.
Here are the results.
PS: the pictures corrupted a little when i resized them and compressed. Actually the WB looks eaven nicer.
Best! Eddy
http://www.cerkovnik.si/images/FCe2.jpg
http://www.cerkovnik.si/images/FCe3.jpg
four4875
09-05-05, 05:57 PM
that is beautifully done. I like the way the finish looks as opposed to the bare machined aluminum. I just dont know how it will affect heat transfer tho.
as a precision machining student, i am curious as to how you measured the thickness of the card to account for the differences in the height of different components, such as ram chips and gpu. would a dial gauge be sufficient? my micrometer isnt large enough to get to all of the different chips, and would love it if you could share techniques.
Do you have a published accuracy that the block is machined to? how do you plan to fill the gaps and ensure tight fits on all of the surfaces? I would love to design and build such a block, but i'm afraid that i wont have a good fit on either the GPU or the ram. the best solution i could think of would be putting in thick layers of AS5 or thermal pads, which isnt desirable.
Thank you for your time :-)
Perseus
09-05-05, 10:45 PM
VERY nice, Eddy. Are you selling these?
I love the nickel plated body.
Eddy_EK
09-06-05, 12:59 AM
I measured one block for those cards, and then I had to correct it a little bit while testing it.
On the surfaces touching together should be used thermal paste.
I think i got obout 10 of that blocks left.
And I think that WB could fit on 7800 too, becouse the fitting holes are the same (I think tho).
Slammin
09-06-05, 07:02 AM
I think what four4875 is getting at is the fact that there seems to be no consistency with the components used on the NV cards, specifically, the ram chips. Some sit higher off the pcb with some cards.
UBER COOL! I'll have to look in to those blocks more.
four4875
09-06-05, 03:35 PM
I think what four4875 is getting at is the fact that there seems to be no consistency with the components used on the NV cards, specifically, the ram chips. Some sit higher off the pcb with some cards.
thats sorta it, but hte chip packages are usually made to a standard, and are poretty consistant in heigth. but im concerned with keeping preasure on all of the chips, and if the surface for the gpu or ram chips are a couple thousandths off it could cause less preasuure or a gap between one of the components, whether its ram or the gpu, which wouldnt sound too appealing to me.
thegreek
09-06-05, 08:05 PM
that looks beautiful
thorilan
09-07-05, 03:56 AM
nickle plating aluminum costs more than just making the entire block from copper doesnt it?
Eddy_EK
09-07-05, 05:56 AM
nickle plating aluminum costs more than just making the entire block from copper doesnt it?
Nickel plating for aluminium costs about 2 USD.
The eloxating of aluminium about 3 USD.
It is cheap...
ctrl_alt_del_
09-07-05, 04:08 PM
so, by nickel plating the base, it eliminates the chance of galvanic corrosion correct? nevermind, you removed the plating - stupid me did not read the whole post, sorry!
Eddy_EK
09-08-05, 12:20 AM
so, by nickel plating the base, it eliminates the chance of galvanic corrosion correct? nevermind, you removed the plating - stupid me did not read the whole post, sorry!
The nickel plating wasn't applied well or I don't know why, so there become in a three days time some marks where aluminium that started corroded in most of tested blocks. I used ordnary water to test it to accelerate the effect.
Therefore I decided to try to "save" them and denickeled them and eloxate them. It all went well, although 3 aluminium parts were not good to use them anymore.
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.