Cyclone: Clear Drill Press CPU Block

airspirit · 02-24-2003, 02:22 PM

Here is the concept:

The left most diagram is the top of the block. This will fit a 1/2" barb.

The right most diagram is the middle of the barb and also the jet impingement nozzle. This piece is seperate so that it can be modified to fit the mounting holes on your board and the nozzle can be drilled to taste.

The center diagram is the base of the block with two 3/8" barbs coming out the sides.

On the very bottom will be a copper plate of user-defined thickness.

There will be three O rings used in the construction of this block to ensure no leakage. The top portions can be made out of any material you desire, though I personally would use a clear plastic for ease of construction. When stacked and the four screw holes are aligned, screws will be threaded from the bottom to the top, and will be bolted down tightly with nuts (not of the deez variety) on the top.

The system pressure will ensure a lack of hot spots inside the block. How?

There will be positive pressure at the top inlet and negative pressure at the two outlets due to the pump. Due to the offset outlets, it will cause a swirling pattern to appear in the coolant as it leaves, pushing the old hot water out and leaving only nice and fresh cold water. This also will use the jet impingement to focus on the "eye" of the swirl: the place right over the die itself.

I have been thinking on this for months, ever since an old waterblock design thread. What are your thoughts? If made out of lexan or acrylic, it would be quite a looker, and I personally think that it would perform extremely well while (depending on the size of your jethole) limiting flow loss. It might not be on par with the White Water or such, but I think it may be a very simple block for a newbie to make on a weak drill press ... and one that will perform nicely as well as be a centerpiece for a window mod.

Here is the diagram:

airspirit · 02-24-2003, 02:27 PM

This is the poor and lazy man's Radius (by BB2K), btw. I estimate this thing would take only about twenty minutes to drill out once the plastic blocks were cut, even if you are totally paranoid about measurements ... which would be pretty forgiving in a simple and open block like this (though all except two holes are just punching straight through the blocks with no finesse or skill required).

airspirit · 02-24-2003, 02:29 PM

The original plastic block would be a 2.25" cube cut into two 2.25x2.25x1.125 semicubes and the center plate would just be made out of 1/4" lexan or something. These materials could be acquired for around $5 USD.

jaydee · 02-24-2003, 02:44 PM

How do you do the O ring grooves with a drill press?

airspirit · 02-24-2003, 04:03 PM

You don't, unless you have a good and steady hand (lock the drill and guide it through the lexan). I'd use a dremel with the router stand (the tripod thing that mounts to the end) to make a even groove all the way around. Since we're using acrylic/lexan, it is extremely easy to do this kind of thing.

bigben2k · 02-24-2003, 07:34 PM

Why do I get the feeling that we've been over your idea before?

(Go ahead, make my day!)

Hey, as long as the baseplate is replaceable, it's all good!

The o-ring is top notch. I'll be using a 1/16 silicone o-ring, and if I keep the groove smooth, everything should stay sealed very nicely. Just remember to let the o-ring sit in the middle of the groove (or equivalent, if not round), and everything will be fine.

jaydee · 02-24-2003, 08:03 PM

Quote:

Originally posted by airspirit
You don't, unless you have a good and steady hand (lock the drill and guide it through the lexan). I'd use a dremel with the router stand (the tripod thing that mounts to the end) to make a even groove all the way around. Since we're using acrylic/lexan, it is extremely easy to do this kind of thing.

Ok, wasn't sure if you had a better way to do it or not. For anyone else wanting to do this but doesn't have the tooling to make the O ring you can just use clear RTV silicone. My REV. 2.0 block has 2 layers of 1/4" acrylic and I used that silicone. Been running for over a month now 24/7 without issue. I used 400 grit sandpaper and scrated up the surfaces that needed siliconing for a better bond. I thought it would make the acrylic look bad with all those scratches I made but I will be damned the clear silicone made them invisable!!!

Anyway. My $.02.

airspirit · 02-24-2003, 09:12 PM

I was thinking O rings for ease of maintenance and portability of the block.

With the clamp integrated as it is, it would allow you to build the clamp to the platform you'd be cooling.

Take a Maze 3 ... if you want to use a socket A model on a P4, you're going to have a chore to make it work because the holes are socket A specific. In this model, you can easily remove the center portion, drill out new holes for your P4, Hammer, etc., and be on your way. Further, it makes it easier for people to experiment easily with the size of the nozzle they want to use.

Though it would save a ton of detail in the end to just glue it together. Sometimes simplicity is best, eh?

jaydee · 02-24-2003, 11:55 PM

Siliconed parts are easy to take apart. Especially acrylic peices. You pretty much HAVE to sand it though because the acrylic is TO smooth. Peices pop apart pretty easy though. Simplicity is sometimes overlooked.

Spamz0r · 02-25-2003, 04:51 AM

Quote:

Originally posted by jaydee116
Siliconed parts are easy to take apart. Especially acrylic peices. You pretty much HAVE to sand it though because the acrylic is TO smooth. Peices pop apart pretty easy though. Simplicity is sometimes overlooked.

thats what i was planning to do with my block

airspirit · 02-25-2003, 08:46 AM

What do you guys think of the design concept itself? I would hazard a guess that it would out-perform a Swiftech due to the lack of hot spots ... what do you think?

jaydee · 02-25-2003, 09:40 AM

Quote:

Originally posted by airspirit
What do you guys think of the design concept itself? I would hazard a guess that it would out-perform a Swiftech due to the lack of hot spots ... what do you think?

I have no idea how it would perform, should work though. Havn't quite figured out the outlets though?

airspirit · 02-25-2003, 10:48 AM

The outlets are the two side holes that are offset off of the main central chamber. You'd probably want to thread two nylon 45 degree 3/8" barbs into those. The reason they are offset is to create swirling in the chamber of the block, "scraping" the hot water out of it and ensuring a lack of hot spots.

Dr. Strangelove · 02-25-2003, 11:27 AM

Hmm interesting idea! I have a few questions though. Are you sure that the nozzle idea will not create to much load on your pump, simply slowing down the water flow too much? (I guess it depends on the nozzle size...) Is the nozzle needed? Compared to just placing the inlet over the die, what do you gain by using the nozzle?

Hope you can clarify these questions.

jaydee · 02-25-2003, 11:59 AM

Quote:

Originally posted by Dr. Strangelove
Hmm interesting idea! I have a few questions though. Are you sure that the nozzle idea will not create to much load on your pump, simply slowing down the water flow too much? (I guess it depends on the nozzle size...) Is the nozzle needed? Compared to just placing the inlet over the die, what do you gain by using the nozzle?

Hope you can clarify these questions.

The nozzel forces water pressure on the base (which should be placed right over the CPU core), which in turn creates a smaller boundry layer and more "friction" (probably not the right word, but can't think of a better term) in which the heat can be transfered. Depending on the design it can be very helpfull or not help at all. Of course the rest of the system will play a small role in its usefullness aswell.

airspirit · 02-25-2003, 01:04 PM

This was the primary reason for the "center plate" in the design, and the reason that the mounting holes were in the center of the block when viewed in profile. The nozzle does create flow resistance, but it also accelerates the flow similar to what you get when you use a nozzle on a hose. When you're spraying that hose farther, you're getting less flow than when it is just pouring out the end, but on the other hand, it can pound a surface harder when sprayed than if the water is just dribbled across it normally.

This helps ensure that the coldest water possible is located on top of the hottest part of the block. The faster flow also helps push the newly warmed water out of the way for fresh, cold water from the jet. By lowering the location of the nozzle toward the baseplate of the block, it ensures more direct pressure against the die area.

airspirit · 02-25-2003, 11:24 PM

You know, with this type of configuration, is the nozzle even needed? Couldn't the extra flow that is generated make up for the lack of pressure since the bottom plate is just going to be flat copper?

airspirit · 02-27-2003, 02:46 PM

To Blackeagle (sorry for the public shout-out, but your box is full ... I hope you get this!):

I don't have the drill press or cutting tool to cut the plastic even if I could get it. I may, however, build one if I can get the supplies cheap enough and that coincides with a trip to the WA coast where my in-laws have the necessary equipment.

A dimpled base (roughed, bumpy, whatever), as long as it doesn't interfere with the swirling flow: that is the key of this design, I think. I went with a flat sheet of copper for the bottom out of simplicity. If you can cut it square, you don't need to mill it or nothing. A roughed up bottom should give better results, though, I suspect.

My goal with this is a simple to make block for the masses that gives good performance. Nice looks are an added bonus, hehe. Most people can figure a way to do this if they have a drill, a dremel, and a tablesaw (though the saw is only needed for a couple of minutes, so you may be able to use someone else's).

Quote:

Blackeagle wrote on 02-27-2003 12:19 PM:
Hi airspirit,

I liked the block idea you posted up. But I don't quite get why a dimpled base would be a bad thing. More surface.

Is this just a idea you're throwing out for comment, or do you intend to build it if you can get the thick plexi?

BE

airspirit · 02-28-2003, 10:17 AM

After looking at some cast acrylic that Blackeagle pointed me to, I'm thinking that these can be manufactured at around $5 per piece, and if mass produced, you can make about 15 of them in the time/effort it would take to make one #rotor block. Since these are easily adaptable per chipset, this may be a solution for someone building a custom dually rig.

Blackeagle · 02-28-2003, 08:25 PM

Sorry about the PM box, glad the link was of help.

BE

airspirit · 03-03-2003, 03:03 PM

I'm thinking I may crank out a stack of these next time I go to western WA ... just jump on the in-laws' tools. If I do, I'll have a bunch available for people to screw around with ... I just NOW need to find a supplier for O rings that has dimensions and such for them. I really like the O ring idea and I don't want to get away from it.

utabintarbo · 03-04-2003, 08:04 AM

Quote:

Originally posted by airspirit
... I just NOW need to find a supplier for O rings that has dimensions and such for them. I really like the O ring idea and I don't want to get away from it.

Get O-ring stock from McMaster's and don't worry about the dim's.

Bob

airspirit · 03-04-2003, 10:10 AM

Cool, I'll check that out.

02-24-2003, 02:27 PM	#2
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	Confessions ... hehehehe This is the poor and lazy man's Radius (by BB2K), btw. I estimate this thing would take only about twenty minutes to drill out once the plastic blocks were cut, even if you are totally paranoid about measurements ... which would be pretty forgiving in a simple and open block like this (though all except two holes are just punching straight through the blocks with no finesse or skill required). __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-24-2003, 02:29 PM	#3
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	The original plastic block would be a 2.25" cube cut into two 2.25x2.25x1.125 semicubes and the center plate would just be made out of 1/4" lexan or something. These materials could be acquired for around $5 USD. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-24-2003, 04:03 PM	#5
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	You don't, unless you have a good and steady hand (lock the drill and guide it through the lexan). I'd use a dremel with the router stand (the tripod thing that mounts to the end) to make a even groove all the way around. Since we're using acrylic/lexan, it is extremely easy to do this kind of thing. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-24-2003, 07:34 PM	#6
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Why do I get the feeling that we've been over your idea before? (Go ahead, make my day!) Hey, as long as the baseplate is replaceable, it's all good! The o-ring is top notch. I'll be using a 1/16 silicone o-ring, and if I keep the groove smooth, everything should stay sealed very nicely. Just remember to let the o-ring sit in the middle of the groove (or equivalent, if not round), and everything will be fine. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

02-24-2003, 09:12 PM	#8
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	I was thinking O rings for ease of maintenance and portability of the block. With the clamp integrated as it is, it would allow you to build the clamp to the platform you'd be cooling. Take a Maze 3 ... if you want to use a socket A model on a P4, you're going to have a chore to make it work because the holes are socket A specific. In this model, you can easily remove the center portion, drill out new holes for your P4, Hammer, etc., and be on your way. Further, it makes it easier for people to experiment easily with the size of the nozzle they want to use. Though it would save a ton of detail in the end to just glue it together. Sometimes simplicity is best, eh? __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-24-2003, 02:44 PM	#4
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	How do you do the O ring grooves with a drill press?

02-24-2003, 11:55 PM	#9
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	Siliconed parts are easy to take apart. Especially acrylic peices. You pretty much HAVE to sand it though because the acrylic is TO smooth. Peices pop apart pretty easy though. Simplicity is sometimes overlooked.

02-25-2003, 08:46 AM	#11
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	What do you guys think of the design concept itself? I would hazard a guess that it would out-perform a Swiftech due to the lack of hot spots ... what do you think? __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-25-2003, 10:48 AM	#13
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	The outlets are the two side holes that are offset off of the main central chamber. You'd probably want to thread two nylon 45 degree 3/8" barbs into those. The reason they are offset is to create swirling in the chamber of the block, "scraping" the hot water out of it and ensuring a lack of hot spots. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-25-2003, 11:27 AM	#14
Dr. Strangelove Cooling Neophyte Join Date: Oct 2002 Location: Denmark Posts: 14	Hmm interesting idea! I have a few questions though. Are you sure that the nozzle idea will not create to much load on your pump, simply slowing down the water flow too much? (I guess it depends on the nozzle size...) Is the nozzle needed? Compared to just placing the inlet over the die, what do you gain by using the nozzle? Hope you can clarify these questions.

02-25-2003, 01:04 PM	#16
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	This was the primary reason for the "center plate" in the design, and the reason that the mounting holes were in the center of the block when viewed in profile. The nozzle does create flow resistance, but it also accelerates the flow similar to what you get when you use a nozzle on a hose. When you're spraying that hose farther, you're getting less flow than when it is just pouring out the end, but on the other hand, it can pound a surface harder when sprayed than if the water is just dribbled across it normally. This helps ensure that the coldest water possible is located on top of the hottest part of the block. The faster flow also helps push the newly warmed water out of the way for fresh, cold water from the jet. By lowering the location of the nozzle toward the baseplate of the block, it ensures more direct pressure against the die area. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-25-2003, 11:24 PM	#17
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	You know, with this type of configuration, is the nozzle even needed? Couldn't the extra flow that is generated make up for the lack of pressure since the bottom plate is just going to be flat copper? __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-28-2003, 10:17 AM	#19
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	After looking at some cast acrylic that Blackeagle pointed me to, I'm thinking that these can be manufactured at around $5 per piece, and if mass produced, you can make about 15 of them in the time/effort it would take to make one #rotor block. Since these are easily adaptable per chipset, this may be a solution for someone building a custom dually rig. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

02-28-2003, 08:25 PM	#20
Blackeagle Thermophile Join Date: Oct 2002 Location: U.S.A = Michigan Posts: 1,243	Sorry about the PM box, glad the link was of help. BE

03-03-2003, 03:03 PM	#21
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	I'm thinking I may crank out a stack of these next time I go to western WA ... just jump on the in-laws' tools. If I do, I'll have a bunch available for people to screw around with ... I just NOW need to find a supplier for O rings that has dimensions and such for them. I really like the O ring idea and I don't want to get away from it. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

03-04-2003, 10:10 AM	#23
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	Cool, I'll check that out. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

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