#Rotor-style, another take on the design...yes folks, thats another one.. :)

*klonk* · 11-08-2003, 07:08 PM

Hey, im pondering a new WB for my CPU, so this evening ive spent 3-4 hours going through the archives reading up on the #Rotor-style WB. I didn't see this design so bear with me if it's somewhere there in the archives (YOU try entering #Rotor in the search function

)

I'd really like to only use two barbs and skip the Y connector so heres what im thinking....

Im a noob as i proved here yesterday ( doh :shrug: ) so i'd really appreiatiate your input on this idea.

The idea is to increase VELOCITY (heh) above the die area.

And yeah you can stop laughing now, the drawing is made with Paint.

Caeberos · 11-08-2003, 09:10 PM

Well in case you missed it #rotor's site:
http://3rotor.homelinux.com/

link to his block http://3rotor.homelinux.com/images/F...templates.html

Hope this helps

#Rotor · 11-08-2003, 10:01 PM

You are very much on the right track...

Velocity is indeed the one factor that will almost always get you
superior performance.

some get confused between velocity and volume. I always say, it's not how much water you move through your block, it's how fast you can move it through.

The faster your volumetric replacement rate is, the more heat you are going to move... be warned, there is more tips hidden in that single line than most people would want or care to comprehend.

sevisehda · 11-09-2003, 01:00 AM

Quote:

There are only 10 kinds of programmers in this world! There's those that understand binary and then there's those that just don't get it at all...

Thats the best sig I've seen in a while.

Edit(In response to the next post): Yes it did.

#Rotor · 11-09-2003, 02:06 AM

It got ya the first time.... didn't it....

*klonk* · 11-09-2003, 08:29 AM

another question:

The copperbars i have are 10mm thick and then i have a piece of 5mm. I wonder if i should use 2x10mm bars to make the WB or use a 10mm as base and a 5mm as lid.

If i use the 2x10mm i could drill holes in the top piece as well allowing more water to get thru, but if i use the 5mm plate as lid, only the 10mm bar would act as coolant.

The latter one would increase pressure ie. velocity, but the first one would be less restrictive...

I notice some of your WB on your site (#Rotor) stand 19mm tall so i guess you use 2 x 10mm bars..?

Puzzdre · 11-09-2003, 09:32 AM

Here's mine #Rotor block, 5 mm top, no holes, 10 mm base. Works great

So, no problem to do it this way, not to mention that you loose considerable amount of total weight of the block with 5 mm top.

*klonk* · 11-09-2003, 09:47 AM

Quote:

Originally posted by Puzzdre
Here's mine #Rotor block, 5 mm top, no holes, 10 mm base. Works great

So, no problem to do it this way, not to mention that you loose considerable amount of total weight of the block with 5 mm top.

Okay - nice finish btw

- are there any scientific consiterations i should have about the narrowing in the middle of my soon to be WB?

Like having 1/2" clearence in the middle or wont that have that big of a difference with performance?

artsite · 11-13-2003, 06:35 AM

What type of glue you place between the plates of 5mm and 10mm?

The screws are of how many mm?

WAJ_UK · 11-13-2003, 06:49 AM

Klonk
If you increase velocity the pressure will decrease. The higher the velocity of the water the better the heat transfer coeficient will be but the increase in turbulance will result in a higher pressure loss through the block. So the balance would be between getting a good enough pressure from the block outlet so your radiator can work properly and getting the heat transfer in the waterblock. I dunno how to work this out, it'll probably take a couple of months with good cfd software.
The channel should probably be at least a couple of mm wider than the CPU die. Don't know for sure though it'll probably need some testing

Puzzdre · 11-13-2003, 07:28 AM

Thx *klonk*!

Artsite, the glue between the plates is ordinary good quality transparent bathroom silicone, works fine. Rotor uses goop, but that kind of stuff is not available in Europe AFAIK).

The screws are M4 Allen bolts 15 mm in length. I don't use teflon tape no more, rather Loctite liquid sealant or even metal epoxy kit for the barbs.

artsite · 11-13-2003, 01:48 PM

Tanks Puzzdre

I will use slicone that i by last month. I also by some glue fthat glues cooper and acrilic. I will test it whit my waterblock tester.

Tanks

*klonk* · 11-13-2003, 03:32 PM

heh blast it - i don't have the measuring equipment...

It'll be pretty unscientific this one... i'll save all measurements of the WB (i.e. width, height, size of channles and so on)

Too bad i don't have a digi-cam before i seal the sucker up... I don't have much time these days so development is going slowly.

in a min or two i'll post a pic with measurements...

Btw, i lock em together with epoxy, it says that no chemicals will dissolve it and mechanical force is needed so im pretty confident it will hold the sucker together

*klonk* · 11-13-2003, 03:59 PM

Heres the edited pic, the blurred parts are cut away with a Dremel (which died during this - POS) so #rotor only applies to the channels in the middle:

Base is roughly 2mm - 3mm

The channels should be something like: 2*7*4 = 56mm2

which i believe is (pretty) roughly equal to an 2"...?

erhm - anyways heres the pic..

Puzzdre · 11-13-2003, 05:06 PM

G'luck to both of you guys, and don't forget to post the pics!

WAJ_UK · 11-14-2003, 06:34 AM

sorry I think I was wrong about having a good enough pressure after the water block for the radiator to work well. If I had a brain I might be dangerous

. Reading what I wrote and having a think about it made me change my mind. If the pressure drop in the block is too high then it will slow down the flow rate in the whole circuit which will effect how all the other components perform, including the radiator, sorry for being dim

Socko · 11-15-2003, 09:22 PM

Quote:

Originally posted by #Rotor
You are very much on the right track...

Velocity is indeed the one factor that will almost always get you
superior performance.

some get confused between velocity and volume. I always say, it's not how much water you move through your block, it's how fast you can move it through.

The faster your volumetric replacement rate is, the more heat you are going to move... be warned, there is more tips hidden in that single line than most people would want or care to comprehend.

Doesn't the velocity of the water directly affect the total volume of water going through a block????????????????????????????????????

Pears0 · 11-15-2003, 09:30 PM

Not if u increase the velocity by push the same ammount of water through a smaller hole!

Socko · 11-15-2003, 10:16 PM

I was thinking as in velocity of water throughout the loop, not just the die.

#Rotor · 11-18-2003, 03:03 PM

Indeed and that is the reason why thicker tubing though-out the loop works better, the more volume your tubing has, the slower the flow-speed will be for the same amount of water passing through any point in the tube, and this is good, because it limits the formation of turbulence in the tubes.

This is exactly what you do not want in your block, for obvious reasons. And also why so many commercial block-Designers have missed to boat (by a mile)

Nowa days they are starting to get the bigger picture. you will be seeing a lot more restrictive block designs popping up everywhere, as already is the case......

The block need to be the only part that should offer resistance to the flow.... because it is the block that need to convert the pressure it sees from the pump, into violent turbulent flow through it, which is what makes a block perform well. Smooth open flow paths are not what you want in a block.

obviously a good design should then give you the smallest possible liquid-volume, for the biggest possible exposed surface-area. The best thing of it all, there are some designs out there that, once you get to that point, it actually also gives you the much needed turbulent flowpath as well...

*klonk* · 11-28-2003, 07:22 PM

Eureka! finished the block and idle temps went from 38c to 32c

Load temps from 45c to 35c thanks to the modified rotor block i made

*Happy happy joy joy dance*

No pictures sadly since i haden't had the chance to borrow a digicam, but the inside and the solderingjob is beautiful if i may say so myself

#Rotor · 11-29-2003, 08:34 AM

That is a very nice improvement. what where you running with, before you made the new block?

*klonk* · 11-29-2003, 10:55 AM

Quote:

Originally posted by #Rotor
That is a very nice improvement. what where you running with, before you made the new block?

Basically a hollow Cu block with very rough/rugged baseplate for turbulence, made it before really reading up on contruction of WBs - a rough estimate suggests a delta of 6c but i need better equipment to be positively sure on that.
Mind you that its is completely passive cooling.

11-08-2003, 10:01 PM	#3
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	You are very much on the right track... Velocity is indeed the one factor that will almost always get you superior performance. some get confused between velocity and volume. I always say, it's not how much water you move through your block, it's how fast you can move it through. The faster your volumetric replacement rate is, the more heat you are going to move... be warned, there is more tips hidden in that single line than most people would want or care to comprehend. __________________ There is no Spoon....

11-09-2003, 02:06 AM	#5
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	It got ya the first time.... didn't it.... __________________ There is no Spoon....

11-13-2003, 06:35 AM	#9
artsite Cooling Neophyte Join Date: Jul 2003 Location: Porto - Portugal Posts: 14	What type of glue you place between the plates of 5mm and 10mm? The screws are of how many mm? __________________ Asus A7N8X non-deluxe - AMD Tbred-B 1800+ @ 2.2 Ghz (11*200) - Termaltake Volcano9 - Smart Fan2 Coolmod - 256 Mb OCZ 3500EL - Chaintheck Geforce4 MX440 @324/513 - Seagate Barracuda IV 80Gb 7200rpm

11-13-2003, 07:28 AM	#11
Puzzdre Cooling Savant Join Date: Aug 2002 Location: Croatia Posts: 969	Thx klonk! Artsite, the glue between the plates is ordinary good quality transparent bathroom silicone, works fine. Rotor uses goop, but that kind of stuff is not available in Europe AFAIK). The screws are M4 Allen bolts 15 mm in length. I don't use teflon tape no more, rather Loctite liquid sealant or even metal epoxy kit for the barbs. __________________ 'Out of cheese error... ...please reboot the universe (press the GBL to continue)'

11-13-2003, 01:48 PM	#12
artsite Cooling Neophyte Join Date: Jul 2003 Location: Porto - Portugal Posts: 14	Tanks Puzzdre I will use slicone that i by last month. I also by some glue fthat glues cooper and acrilic. I will test it whit my waterblock tester. Tanks __________________ Asus A7N8X non-deluxe - AMD Tbred-B 1800+ @ 2.2 Ghz (11*200) - Termaltake Volcano9 - Smart Fan2 Coolmod - 256 Mb OCZ 3500EL - Chaintheck Geforce4 MX440 @324/513 - Seagate Barracuda IV 80Gb 7200rpm

11-08-2003, 09:10 PM	#2
Caeberos Cooling Neophyte Join Date: Mar 2003 Location: New York Posts: 32	Well in case you missed it #rotor's site: http://3rotor.homelinux.com/ link to his block http://3rotor.homelinux.com/images/F...templates.html Hope this helps

11-09-2003, 08:29 AM	#6
klonk Cooling Neophyte Join Date: Jun 2003 Location: denmark Posts: 73	another question: The copperbars i have are 10mm thick and then i have a piece of 5mm. I wonder if i should use 2x10mm bars to make the WB or use a 10mm as base and a 5mm as lid. If i use the 2x10mm i could drill holes in the top piece as well allowing more water to get thru, but if i use the 5mm plate as lid, only the 10mm bar would act as coolant. The latter one would increase pressure ie. velocity, but the first one would be less restrictive... I notice some of your WB on your site (#Rotor) stand 19mm tall so i guess you use 2 x 10mm bars..?

11-13-2003, 06:49 AM	#10
WAJ_UK Cooling Savant Join Date: Oct 2003 Location: Sussex Posts: 109	Klonk If you increase velocity the pressure will decrease. The higher the velocity of the water the better the heat transfer coeficient will be but the increase in turbulance will result in a higher pressure loss through the block. So the balance would be between getting a good enough pressure from the block outlet so your radiator can work properly and getting the heat transfer in the waterblock. I dunno how to work this out, it'll probably take a couple of months with good cfd software. The channel should probably be at least a couple of mm wider than the CPU die. Don't know for sure though it'll probably need some testing

11-13-2003, 03:32 PM	#13
klonk Cooling Neophyte Join Date: Jun 2003 Location: denmark Posts: 73	heh blast it - i don't have the measuring equipment... It'll be pretty unscientific this one... i'll save all measurements of the WB (i.e. width, height, size of channles and so on) Too bad i don't have a digi-cam before i seal the sucker up... I don't have much time these days so development is going slowly. in a min or two i'll post a pic with measurements... Btw, i lock em together with epoxy, it says that no chemicals will dissolve it and mechanical force is needed so im pretty confident it will hold the sucker together

11-13-2003, 05:06 PM	#15
Puzzdre Cooling Savant Join Date: Aug 2002 Location: Croatia Posts: 969	G'luck to both of you guys, and don't forget to post the pics! __________________ 'Out of cheese error... ...please reboot the universe (press the GBL to continue)'

11-14-2003, 06:34 AM	#16
WAJ_UK Cooling Savant Join Date: Oct 2003 Location: Sussex Posts: 109	sorry I think I was wrong about having a good enough pressure after the water block for the radiator to work well. If I had a brain I might be dangerous . Reading what I wrote and having a think about it made me change my mind. If the pressure drop in the block is too high then it will slow down the flow rate in the whole circuit which will effect how all the other components perform, including the radiator, sorry for being dim

11-15-2003, 09:30 PM	#18
Pears0 Cooling Neophyte Join Date: Nov 2003 Location: Birmingham, england Posts: 61	Not if u increase the velocity by push the same ammount of water through a smaller hole! __________________ England rule @ rugby. Lookin to build custom water coolin rig in next few months with compleat new hardware!

11-15-2003, 10:16 PM	#19
Socko Cooling Neophyte Join Date: Nov 2003 Location: Vineland, NJ Posts: 19	I was thinking as in velocity of water throughout the loop, not just the die.

11-18-2003, 03:03 PM	#20
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	Indeed and that is the reason why thicker tubing though-out the loop works better, the more volume your tubing has, the slower the flow-speed will be for the same amount of water passing through any point in the tube, and this is good, because it limits the formation of turbulence in the tubes. This is exactly what you do not want in your block, for obvious reasons. And also why so many commercial block-Designers have missed to boat (by a mile) Nowa days they are starting to get the bigger picture. you will be seeing a lot more restrictive block designs popping up everywhere, as already is the case...... The block need to be the only part that should offer resistance to the flow.... because it is the block that need to convert the pressure it sees from the pump, into violent turbulent flow through it, which is what makes a block perform well. Smooth open flow paths are not what you want in a block. obviously a good design should then give you the smallest possible liquid-volume, for the biggest possible exposed surface-area. The best thing of it all, there are some designs out there that, once you get to that point, it actually also gives you the much needed turbulent flowpath as well... __________________ There is no Spoon....

11-28-2003, 07:22 PM	#21
klonk Cooling Neophyte Join Date: Jun 2003 Location: denmark Posts: 73	Eureka! finished the block and idle temps went from 38c to 32c Load temps from 45c to 35c thanks to the modified rotor block i made Happy happy joy joy dance No pictures sadly since i haden't had the chance to borrow a digicam, but the inside and the solderingjob is beautiful if i may say so myself

11-29-2003, 08:34 AM	#22
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	That is a very nice improvement. what where you running with, before you made the new block? __________________ There is no Spoon....

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