Need feedback on these baseplate ideas

aBo · 10-15-2003, 09:11 PM

So... I finally got around to making my first post here so I might make a new thread too =), here goes...

Ihave been working on changing to water cooling because I want lower noise levels from my computer (currently running a Vapochill classic). I already bought a 310mm*50mm*10mm piece of copper and I'm hopefully going to use it as baseplates for CPU, GPU and Northbridge...

I'm propably going to do #Rotor on all of them with polycarb top with 13mm/16mm id barbs... If I don't screw up somewhere along the way I should have some extra copper left and here's a couple of ideas I might try, so please tell why these wont work :D.

Here's a top view (cut half in the middle) of

this baseplate, so it's basically direct die (correct lining up would be a bitch)... If I have understood correctly, this kind of setup shouldn't corrode the die too fast (distilled water and 5-10% glycol in circulation)?

Next one looks like this:

It's basically a classic #Rotor but I have been toying with an idea of threading the holes above the core (I would use 4mm drill everywhere else and then drill these holes to 3mm and thread them to M4 (4mm))... Would that give any kind of performance gain a) because of bigger surface area and/or b) maybe increased turbulence?

And the last one is more of a late-night-persistent-idea -kind of combination of previous two:

...where the threaded holes above core would be actual holes reaching the core :) (or maybe very very thin baseplate, like 0,5mm or something like that)...

So, please flame me back to stone age =).

siavash_s_s · 10-15-2003, 09:17 PM

the direct die method seems like a waste of time. the normal approach made by (cant remember his name, his avatar is an amd chip in a glass of water) is the easist method and should work better than this. i have no real experience in making blocks so ill leave it as that.. (btw rotors block is the best performing easy make block atm so there isnt much room for improvement in his design, i think.......)

winewood · 10-16-2003, 01:36 AM

#4 has self defeating fins on either side of the core. Heat really wouldn't have a way of reaching them.
#3 looks nice, but that finning on the sides of the middle fins.. not necessary. The thermal barrier will radiate beyond them. They are more for looks. And frankly, looks don't count for much on a homeade block under a copper top.

is that an HK USP compact? I can't tell due to the dark pic. Nice weapon.

aBo · 10-16-2003, 08:19 AM

Thanks for comments, more would be welcome =).

siavash_s_s

Direct die because of experimentation and possible performance gains... I would be interested in finding out the flaws in this particular desingn rather than the generic "direct die sucks, forget it", that is all-too often spread around =).
About the #Rotor design, I agree there's propably not much room for improvement but let's not let that limit our creativity?-)

winewood

#4 - propably so, but they wouldn't hurt... but this is more of an attempt to learn from faulty ideas than a serious baseplate

#3 - The heat will radiate beyond the central holes, that is a good thing, but wouldn't incrased surface area increase their capability to release heat into the water, though? They might be more for the looks but I'd like to get comments on their possible impacts on performance, no matter how small =) (I have M2-M12 (2-12mm threads) taps available at no cost so it wouldn't be an extra cost). (BTW, I´ll propably do these blocks with polycarb top, as I mentioned).

The gun is Glock 17, should be pretty evident (maybe you have too dark settings on your monitor?), but it's not real though but an airsoft replica =).

---

BTW, how deep holes you think I should make to my #Rotor -baseplates? I know 4mm holes and 100-110% depth is recommended by the guru himself but that would leave me like ~6mm of undrilled copper between the core and the water channels, isn't that a little bit too much for optimal heat dissipation? So, what would you suggest for the size of the drill bit and the depth for 10mm thick copper with polycarb top?

aBo · 10-16-2003, 12:20 PM

Come on, think about swiftech heatsinks =)...

pelle76 · 10-16-2003, 01:48 PM

2: No chans you can make that without getting leakage.... Why use copper at all if youre not going to use it for cooling but mearly as casing for the water... 2 could be made out of plastic...

4: I wonder how good this block would hold together when you preasure it against the core...

Wat will you use as sealant on the DD approaches?

aBo · 10-16-2003, 03:50 PM

pelle76

2 - Yeah, the material would be pretty trivial, so let's not get stuck into that =)... maybe some kind of core-shaped&sized "setting" (term?) would help correct placement and thus reduce chances of leakage. Water wouldn't have to be in touch of all of the core's area, one could leave small "lip" above the core (as I tried to illustrate above)

4 - Good point... although the top plate would support the structure...

I haven't decided on any particular brand as sealant, perhaps sikaflex or something similar good quality&heavy duty sealant...

*edit*

Wanted to add this link about direct die cooling... So, people, please check it out, they seem to be pretty valid designs and that kind of approach is rarely used...

I'm not planning to do direct die all around the core, just the topside, because of the reports about problems on the core attachment substance.

jaydee · 10-16-2003, 06:43 PM

Quote:

Originally posted by aBo
pelle76

Wanted to add this link about direct die cooling... So, people, please check it out, they seem to be pretty valid designs and that kind of approach is rarely used...

I'm not planning to do direct die all around the core, just the topside, because of the reports about problems on the core attachment substance.

Unfortuantly this guy proved jack shit. Where is the results against any other name brand block like a TC-4, Maze 3-4, Cascade, White Water, etc.... All I see was a reference to the Hedeghog waterblock which means nothing. If you look down a few posts below this one you will see my Direct Die cooling project. Came out about as good as a Maze 4. it has jet impingment, high flow rates, etc.... Just isn't worth it....

aBo · 10-16-2003, 07:32 PM

jaydee116

I hear you... as I mentioned, I'm not planning to do direct die as my first option, just a few thoughts I have been toying around with... So basically, the processor core material isn't thick enough to dissipate efficiently heat into cooling liquid?

I followed that direct die thread of yours but I thought it kinda died away, I can't remember any results being published, but I must have skipped some crucial part =/... Anyways, nice that some people have the guts to go down the unusual roads too, sometimes they even lead somewhere (like whitewater=).

About that guys blocks, I'm not claiming he has proved anything, I don't think he even tries =)... although he has set up a site where he is going to market his stuff it hasn't been updated for ages, so basically he is a hobbyist.

This was one of the sites of interest for me, It's in german so use altavista's babel fish when in doubt. His testing methods/documentation look pretty good but guess that what principles might work for Celeron might not work for bigger heat sources such as Athlons...

jaydee · 10-16-2003, 11:25 PM

Quote:

Originally posted by aBo
jaydee116

I hear you... as I mentioned, I'm not planning to do direct die as my first option, just a few thoughts I have been toying around with... So basically, the processor core material isn't thick enough to dissipate efficiently heat into cooling liquid?

I followed that direct die thread of yours but I thought it kinda died away, I can't remember any results being published, but I must have skipped some crucial part =/...

Here is the results I got. They were buried in that thread.

Quote:

Originally posted by jaydee116
I got the HydroThruster setup and running. Pretty disappointed in this direct die cooling so far. I got plenty of flow now and still the temps are pretty close the Maze 4. Right now 1C cooler than the Maze 4 and 3C warmer than the Lemon Block. I see no point in doing a long term test as you would be pretty stupid not to use a good regular water block over this. However Direct Die #2 is in the works. I am going ot use the T-Bird bretb sent me. This time I am going to remove the jets and try and make one big cone style jet right over the die and see what happens.

I got some pics here but my host is still having issues so I am not sure if they will show up or not. Sooner or later they will.

msv · 10-17-2003, 01:28 AM

aBo, check Your tap and You will see that design #3 will be hard to thread. #4 will be easier, since You can run the tap all through the block.
I´ll humbly leave the direct cie cooling discussion to the more experienced.
regards
Mikael S.

10-15-2003, 09:11 PM	#1
aBo Cooling Neophyte Join Date: Aug 2003 Location: Finland Posts: 18	Need feedback on these baseplate ideas So... I finally got around to making my first post here so I might make a new thread too =), here goes... Ihave been working on changing to water cooling because I want lower noise levels from my computer (currently running a Vapochill classic). I already bought a 310mm50mm10mm piece of copper and I'm hopefully going to use it as baseplates for CPU, GPU and Northbridge... I'm propably going to do #Rotor on all of them with polycarb top with 13mm/16mm id barbs... If I don't screw up somewhere along the way I should have some extra copper left and here's a couple of ideas I might try, so please tell why these wont work :D. Here's a top view (cut half in the middle) of this baseplate, so it's basically direct die (correct lining up would be a bitch)... If I have understood correctly, this kind of setup shouldn't corrode the die too fast (distilled water and 5-10% glycol in circulation)? Next one looks like this: It's basically a classic #Rotor but I have been toying with an idea of threading the holes above the core (I would use 4mm drill everywhere else and then drill these holes to 3mm and thread them to M4 (4mm))... Would that give any kind of performance gain a) because of bigger surface area and/or b) maybe increased turbulence? And the last one is more of a late-night-persistent-idea -kind of combination of previous two: ...where the threaded holes above core would be actual holes reaching the core :) (or maybe very very thin baseplate, like 0,5mm or something like that)... So, please flame me back to stone age =).

10-16-2003, 03:50 PM	#7
aBo Cooling Neophyte Join Date: Aug 2003 Location: Finland Posts: 18	pelle76 2 - Yeah, the material would be pretty trivial, so let's not get stuck into that =)... maybe some kind of core-shaped&sized "setting" (term?) would help correct placement and thus reduce chances of leakage. Water wouldn't have to be in touch of all of the core's area, one could leave small "lip" above the core (as I tried to illustrate above) 4 - Good point... although the top plate would support the structure... I haven't decided on any particular brand as sealant, perhaps sikaflex or something similar good quality&heavy duty sealant... edit Wanted to add this link about direct die cooling... So, people, please check it out, they seem to be pretty valid designs and that kind of approach is rarely used... I'm not planning to do direct die all around the core, just the topside, because of the reports about problems on the core attachment substance. Last edited by aBo; 10-16-2003 at 03:55 PM.

10-17-2003, 01:28 AM	#11
msv Cooling Savant Join Date: Dec 2002 Location: Sweden Posts: 336	aBo, check Your tap and You will see that design #3 will be hard to thread. #4 will be easier, since You can run the tap all through the block. I´ll humbly leave the direct cie cooling discussion to the more experienced. regards Mikael S. __________________ The only constant factor in all Your failures is You. Catapultam habeo. Nisi pecuniam omnen mihi habis, ad caput tuum saxum immane mittam.

10-15-2003, 09:17 PM	#2
siavash_s_s Cooling Savant Join Date: Aug 2003 Location: here n there Posts: 194	the direct die method seems like a waste of time. the normal approach made by (cant remember his name, his avatar is an amd chip in a glass of water) is the easist method and should work better than this. i have no real experience in making blocks so ill leave it as that.. (btw rotors block is the best performing easy make block atm so there isnt much room for improvement in his design, i think.......) __________________ buy me a mill

10-16-2003, 01:36 AM	#3
winewood Cooling Savant Join Date: Nov 2002 Location: in my chair Posts: 574	#4 has self defeating fins on either side of the core. Heat really wouldn't have a way of reaching them. #3 looks nice, but that finning on the sides of the middle fins.. not necessary. The thermal barrier will radiate beyond them. They are more for looks. And frankly, looks don't count for much on a homeade block under a copper top. is that an HK USP compact? I can't tell due to the dark pic. Nice weapon. __________________ -winewood-

10-16-2003, 08:19 AM	#4
aBo Cooling Neophyte Join Date: Aug 2003 Location: Finland Posts: 18	Thanks for comments, more would be welcome =). siavash_s_s Direct die because of experimentation and possible performance gains... I would be interested in finding out the flaws in this particular desingn rather than the generic "direct die sucks, forget it", that is all-too often spread around =). About the #Rotor design, I agree there's propably not much room for improvement but let's not let that limit our creativity?-) winewood #4 - propably so, but they wouldn't hurt... but this is more of an attempt to learn from faulty ideas than a serious baseplate #3 - The heat will radiate beyond the central holes, that is a good thing, but wouldn't incrased surface area increase their capability to release heat into the water, though? They might be more for the looks but I'd like to get comments on their possible impacts on performance, no matter how small =) (I have M2-M12 (2-12mm threads) taps available at no cost so it wouldn't be an extra cost). (BTW, I´ll propably do these blocks with polycarb top, as I mentioned). The gun is Glock 17, should be pretty evident (maybe you have too dark settings on your monitor?), but it's not real though but an airsoft replica =). --- BTW, how deep holes you think I should make to my #Rotor -baseplates? I know 4mm holes and 100-110% depth is recommended by the guru himself but that would leave me like ~6mm of undrilled copper between the core and the water channels, isn't that a little bit too much for optimal heat dissipation? So, what would you suggest for the size of the drill bit and the depth for 10mm thick copper with polycarb top?

10-16-2003, 12:20 PM	#5
aBo Cooling Neophyte Join Date: Aug 2003 Location: Finland Posts: 18	Come on, think about swiftech heatsinks =)...

10-16-2003, 01:48 PM	#6
pelle76 Cooling Savant Join Date: Apr 2003 Location: Sweden, Skovde Posts: 101	2: No chans you can make that without getting leakage.... Why use copper at all if youre not going to use it for cooling but mearly as casing for the water... 2 could be made out of plastic... 4: I wonder how good this block would hold together when you preasure it against the core... Wat will you use as sealant on the DD approaches?

10-16-2003, 07:32 PM	#9
aBo Cooling Neophyte Join Date: Aug 2003 Location: Finland Posts: 18	jaydee116 I hear you... as I mentioned, I'm not planning to do direct die as my first option, just a few thoughts I have been toying around with... So basically, the processor core material isn't thick enough to dissipate efficiently heat into cooling liquid? I followed that direct die thread of yours but I thought it kinda died away, I can't remember any results being published, but I must have skipped some crucial part =/... Anyways, nice that some people have the guts to go down the unusual roads too, sometimes they even lead somewhere (like whitewater=). About that guys blocks, I'm not claiming he has proved anything, I don't think he even tries =)... although he has set up a site where he is going to market his stuff it hasn't been updated for ages, so basically he is a hobbyist. This was one of the sites of interest for me, It's in german so use altavista's babel fish when in doubt. His testing methods/documentation look pretty good but guess that what principles might work for Celeron might not work for bigger heat sources such as Athlons...

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)