water cooling your thoughts theorys and suggestions

Aardil · 10-26-2002, 01:39 AM

I am in the process of building a water cooling system for my PC's
some of this may be over kill but it is my thoughts and plan (both are subject to change)

I think this is a rather unique Idea, it may have already been done but being fairly new to the world of WC I have not seen anything like it. My thoughts and experiances are these:

Why use a WC that can fit in a Mid size PC case when at best it will only hold a quart of water ( Iceburg system) ?
Tried it and other than the 1/2 hour of wet sanding to get the gouges out of the water block it went together pretty well. entire system fit inside a mid case but no noticeable difference in chip temp, but was quiet compared to the cooler master H-002 cooler. after about a week I noticed a lake forming under the PC. the Injected molded res cracked from all 4 screw points that hold the lid on down to about 1 inch from the bottom of res. And before anyone says I over tightened the screws I barely snugged the top because I was afraid of the very thin material. Well it only took the rest of the day to make a new Res out of 1/4 inch plexiglass, I did make it as big as I could and still fit in the case, increased the water res to close to 1/2 gallon. further testing revieled a slight temp drop with added water about 1C, but at least it doesnt leak.

which leads to project #2

PC's to be cooled AMD 2000+, AMD 800
Planned colling system
I figure I can put a system together and put the cooling components into a gutted 3rd case:
make a res as large as possible:
run a single Eheim 1250 pump 317 GPH:
Hose is 1/2 tygon tubing and brass 1/2 inch fittings through out:
Special note had to use adapters at water block Strange fitting some thing called bspp thead:
route water out of pump through a black Ice extreme rad with 120mm fan (kinda like pre cooling before it goes to water blocks) :
splits into 2 separate lines one to the cooling block on the 800 the other to the 2000+:
Cooling blocks are both Innovatek innovaCOOL rev3.0:
Return lines come back to cooling tower and are routed through a adjustable valve ( this is in case the difference in hose length tries to send all or most of the cooling to one block or the other I.E. path of least resistance):
reutrn lines then merge back into one line and water is past through a second black Ice extreme rad with 120mm fan ( removing heat before returning to res.
there are 3 fans in the system, 2 = 120 mm radio shack fans and one 80 mm radio shack fan. the reasoning for the Radio Shack fans is : they are 110 volt AC fans. Since the case is gutted there is no power supply for 12 volts only 110 curent .

This is what I have planned tell me if you see any basic flaws in the system, suggestions about blocks or Radiators, any other thoughts you may have. I will have about 600 bucks tied up in this single point dual system cooler.
I have as they say " It looked good on paper"

Aardil

g.l.amour · 10-26-2002, 04:06 AM

try to get a hold of dax, he is juggling with the exact same idea as you. maybe u can exchange some ideas.

me personally wouldn't think of it. it would become a nightmare moving the pc around the house for , let's say maintenance work, draining, whatever.

but, that is my opinion... i u got a challenge, go for it, and surprise us

airspirit · 10-26-2002, 02:13 PM

If you put butterfly valves mounted in the in/out lines of each case and provide for a good drainage solution, you'll have good results. This is being done in rackmount systems (#rotor did it, I believe), and it works great. When you need to move you equipment, just drain, close the valves, undo the hoses OUTSIDE the cases, move, plug her back together, and refill. This has been done many times with success, though I haven't seen anyone in this forum do it.

godsdice · 10-26-2002, 02:27 PM

Quote:

Originally posted by Aardil
Return lines come back to cooling tower and are routed through a adjustable valve ( this is in case the difference in hose length tries to send all or most of the cooling to one block or the other I.E. path of least resistance):

This will never happen. Think about it.

Aardil · 10-26-2002, 02:58 PM

ok two further comments
butter fly valves:
good thought but was thinking of going a step farther, I know they make quick connects for fluid that seal off when taken apart. this would do the same thing but no valve to turn just connect or disconnect and go. If they work as I think they do there would be little to no fluid loss. Also yes I am using Brass connections outside the case, to get everything up and running I will be using just Barbed 1/2 inch fittings to targon tubes. when everything works I simply separate the male/female barb ends and replace them with the quick connect set up.

second thought:
Not gonna happen?
while it is true water will go through all lines I like the idea of being to regulate the flow at least to some extint. by using the flow meters after the valves and before they tee back together and enter second radiator I should be able to equalize them.
Or am I wrong?

Aardil

airspirit · 10-26-2002, 03:16 PM

Godsdice is wrong. Think about it.

Each system has its own flow resistance that is unique to it. By using ball valves (best regulation but not failsafe due to the usually organic packing and the additives you will be using) or butterfly valves (harder to regulate flow, but less likely to blow a seal) you can equalize your flow. Rather, what you can do is watch the temps on your machines and fiddle until they even out.

Godsdice: to avoid being seen as a troll, please back up statements like what you said. Most people here will appreciate the input, however wrong it may be, if it as least reasoned out. Just stating "You are wrong" will get you flamed more often than not.

Quick disconnects of the type you're talking about (most often used in high pressure and hydraulic applications) are heavy, bulky, and expensive. Plus, unless you use an oversize type, they will create restriction, and you will still need a regulation valve. It would, however, eliminate the need to drain your system. I don't know why I hadn't thought of it before. By using oversized disconnects you could do an exterior mount of the disconnect that would make for extremely easy reconfiguring/moving without hurting flow. You will definitely want to equip an efficient airtrap if you do this, because these types of disconnects will allow small amounts of air into the system, and they will definitely leak a SMALL amount of fluid when you pop them apart (I've only been hosed down with hydraulic fluid once using them, and it was because the connect failed ... normally you just get the equivalent of a quarter inch or less worth of fluid lost in the process.

Aardil · 10-26-2002, 05:11 PM

A further note on the quick disconnects:
I understand the restriction problem and am going to use oversize fittings. Also all cases are using bulk head fittings with 1/4 NPT thead size. I understand I may have been better off to use 1/2 inch NPT but case room wont allow it unless I attach them into the side panel ( not what I wanted to do).
Thank you one and all for your input, it at least makes me think I am headed in the right direction.
I hope to have the system completed in about 3 weeks, if all goes well I will post a summary and some pics.
one last thought:
do ya think braided steel line might be a bit overkill?? :P

Aardil

Aardil · 10-30-2002, 09:39 PM

As promised a few pics of my work in progress probably nothing new to you guys but id you see something you dont understand post I will try to explain.

Aardil

http://members.cox.net/aardil/front.jpg

http://members.cox.net/aardil/insidefrontfan.jpg

http://members.cox.net/aardil/insidetop.jpg

http://members.cox.net/aardil/rear.jpg

http://members.cox.net/aardil/side_resivior.jpg

Waiting on 1250 pump to come in ( should be here Monday) then it wqill start taking shape.
Rads and fans will be in I hope the week after that.

Aardil · 11-02-2002, 10:36 PM

As I wait on my pump to arrive I thought perhaps a drwaing of the proposed cooler case system might clarify some of my thoughts.

http://members.cox.net/aardil/drawing.jpg

Aardil · 11-10-2002, 07:54 PM

Well i week later ( and 3 failed attempts) the Res is finally complete.
No leaks and holds 1.7 U.S. Gallons
Fits into the case nicely and Pump fit in perfect as well.
I ended up going with 1/4 inch Plexiglass and used a plexiglass bender to make the main body, came out pretty good considering its my first time making one.
Used a 1/2 NPT to 5/8 barb hose to connect pump inlet, all other fittings are 1/4 NPT to 1/2 barb
I hope to post some actual flow rates later but waiting on rads to come in.
So far I am happy with the results
take a look and tell me what you think.
Aardil

http://members.cox.net/aardil/resrev3.jpg

http://members.cox.net/aardil/res3close.jpg

10-26-2002, 01:39 AM	#1
Aardil Cooling Savant Join Date: Oct 2002 Location: Louisiana Posts: 174	water cooling your thoughts theorys and suggestions I am in the process of building a water cooling system for my PC's some of this may be over kill but it is my thoughts and plan (both are subject to change) I think this is a rather unique Idea, it may have already been done but being fairly new to the world of WC I have not seen anything like it. My thoughts and experiances are these: Why use a WC that can fit in a Mid size PC case when at best it will only hold a quart of water ( Iceburg system) ? Tried it and other than the 1/2 hour of wet sanding to get the gouges out of the water block it went together pretty well. entire system fit inside a mid case but no noticeable difference in chip temp, but was quiet compared to the cooler master H-002 cooler. after about a week I noticed a lake forming under the PC. the Injected molded res cracked from all 4 screw points that hold the lid on down to about 1 inch from the bottom of res. And before anyone says I over tightened the screws I barely snugged the top because I was afraid of the very thin material. Well it only took the rest of the day to make a new Res out of 1/4 inch plexiglass, I did make it as big as I could and still fit in the case, increased the water res to close to 1/2 gallon. further testing revieled a slight temp drop with added water about 1C, but at least it doesnt leak. which leads to project #2 PC's to be cooled AMD 2000+, AMD 800 Planned colling system I figure I can put a system together and put the cooling components into a gutted 3rd case: make a res as large as possible: run a single Eheim 1250 pump 317 GPH: Hose is 1/2 tygon tubing and brass 1/2 inch fittings through out: Special note had to use adapters at water block Strange fitting some thing called bspp thead: route water out of pump through a black Ice extreme rad with 120mm fan (kinda like pre cooling before it goes to water blocks) : splits into 2 separate lines one to the cooling block on the 800 the other to the 2000+: Cooling blocks are both Innovatek innovaCOOL rev3.0: Return lines come back to cooling tower and are routed through a adjustable valve ( this is in case the difference in hose length tries to send all or most of the cooling to one block or the other I.E. path of least resistance): reutrn lines then merge back into one line and water is past through a second black Ice extreme rad with 120mm fan ( removing heat before returning to res. there are 3 fans in the system, 2 = 120 mm radio shack fans and one 80 mm radio shack fan. the reasoning for the Radio Shack fans is : they are 110 volt AC fans. Since the case is gutted there is no power supply for 12 volts only 110 curent . This is what I have planned tell me if you see any basic flaws in the system, suggestions about blocks or Radiators, any other thoughts you may have. I will have about 600 bucks tied up in this single point dual system cooler. I have as they say " It looked good on paper" Aardil

10-26-2002, 02:13 PM	#3
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	If you put butterfly valves mounted in the in/out lines of each case and provide for a good drainage solution, you'll have good results. This is being done in rackmount systems (#rotor did it, I believe), and it works great. When you need to move you equipment, just drain, close the valves, undo the hoses OUTSIDE the cases, move, plug her back together, and refill. This has been done many times with success, though I haven't seen anyone in this forum do 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

10-26-2002, 03:16 PM	#6
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	Godsdice is wrong. Think about it. Each system has its own flow resistance that is unique to it. By using ball valves (best regulation but not failsafe due to the usually organic packing and the additives you will be using) or butterfly valves (harder to regulate flow, but less likely to blow a seal) you can equalize your flow. Rather, what you can do is watch the temps on your machines and fiddle until they even out. Godsdice: to avoid being seen as a troll, please back up statements like what you said. Most people here will appreciate the input, however wrong it may be, if it as least reasoned out. Just stating "You are wrong" will get you flamed more often than not. Quick disconnects of the type you're talking about (most often used in high pressure and hydraulic applications) are heavy, bulky, and expensive. Plus, unless you use an oversize type, they will create restriction, and you will still need a regulation valve. It would, however, eliminate the need to drain your system. I don't know why I hadn't thought of it before. By using oversized disconnects you could do an exterior mount of the disconnect that would make for extremely easy reconfiguring/moving without hurting flow. You will definitely want to equip an efficient airtrap if you do this, because these types of disconnects will allow small amounts of air into the system, and they will definitely leak a SMALL amount of fluid when you pop them apart (I've only been hosed down with hydraulic fluid once using them, and it was because the connect failed ... normally you just get the equivalent of a quarter inch or less worth of fluid lost in the process. __________________ #!/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

10-30-2002, 09:39 PM	#8
Aardil Cooling Savant Join Date: Oct 2002 Location: Louisiana Posts: 174	Phase one underway As promised a few pics of my work in progress probably nothing new to you guys but id you see something you dont understand post I will try to explain. Aardil http://members.cox.net/aardil/front.jpg http://members.cox.net/aardil/insidefrontfan.jpg http://members.cox.net/aardil/insidetop.jpg http://members.cox.net/aardil/rear.jpg http://members.cox.net/aardil/side_resivior.jpg Waiting on 1250 pump to come in ( should be here Monday) then it wqill start taking shape. Rads and fans will be in I hope the week after that.

11-10-2002, 07:54 PM	#10
Aardil Cooling Savant Join Date: Oct 2002 Location: Louisiana Posts: 174	A few changes Well i week later ( and 3 failed attempts) the Res is finally complete. No leaks and holds 1.7 U.S. Gallons Fits into the case nicely and Pump fit in perfect as well. I ended up going with 1/4 inch Plexiglass and used a plexiglass bender to make the main body, came out pretty good considering its my first time making one. Used a 1/2 NPT to 5/8 barb hose to connect pump inlet, all other fittings are 1/4 NPT to 1/2 barb I hope to post some actual flow rates later but waiting on rads to come in. So far I am happy with the results take a look and tell me what you think. Aardil http://members.cox.net/aardil/resrev3.jpg http://members.cox.net/aardil/res3close.jpg

10-26-2002, 04:06 AM	#2
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	try to get a hold of dax, he is juggling with the exact same idea as you. maybe u can exchange some ideas. me personally wouldn't think of it. it would become a nightmare moving the pc around the house for , let's say maintenance work, draining, whatever. but, that is my opinion... i u got a challenge, go for it, and surprise us

10-26-2002, 02:58 PM	#5
Aardil Cooling Savant Join Date: Oct 2002 Location: Louisiana Posts: 174	ok two further comments butter fly valves: good thought but was thinking of going a step farther, I know they make quick connects for fluid that seal off when taken apart. this would do the same thing but no valve to turn just connect or disconnect and go. If they work as I think they do there would be little to no fluid loss. Also yes I am using Brass connections outside the case, to get everything up and running I will be using just Barbed 1/2 inch fittings to targon tubes. when everything works I simply separate the male/female barb ends and replace them with the quick connect set up. second thought: Not gonna happen? while it is true water will go through all lines I like the idea of being to regulate the flow at least to some extint. by using the flow meters after the valves and before they tee back together and enter second radiator I should be able to equalize them. Or am I wrong? Aardil

10-26-2002, 05:11 PM	#7
Aardil Cooling Savant Join Date: Oct 2002 Location: Louisiana Posts: 174	A further note on the quick disconnects: I understand the restriction problem and am going to use oversize fittings. Also all cases are using bulk head fittings with 1/4 NPT thead size. I understand I may have been better off to use 1/2 inch NPT but case room wont allow it unless I attach them into the side panel ( not what I wanted to do). Thank you one and all for your input, it at least makes me think I am headed in the right direction. I hope to have the system completed in about 3 weeks, if all goes well I will post a summary and some pics. one last thought: do ya think braided steel line might be a bit overkill?? :P Aardil

11-02-2002, 10:36 PM	#9
Aardil Cooling Savant Join Date: Oct 2002 Location: Louisiana Posts: 174	As I wait on my pump to arrive I thought perhaps a drwaing of the proposed cooler case system might clarify some of my thoughts. http://members.cox.net/aardil/drawing.jpg

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