Go Back   Pro/Forums > ProCooling Technical Discussions > General Liquid/Water Cooling Discussion
Password
Register FAQ Members List Calendar JavaChat Mark Forums Read

General Liquid/Water Cooling Discussion For discussion about Full Cooling System kits, or general cooling topics. Keep specific cooling items like pumps, radiators, etc... in their specific forums.

Reply
Thread Tools
Unread 09-13-2002, 01:42 PM   #26
redleader
Thermophile
 
Join Date: Jun 2001
Location: The deserts of Tucson, Az
Posts: 1,264
Default

Common sense, two fans distirb twice the air and thus make twice the sound.

---------

All good blocks perform nearly the same. I'd look cost and reputation. You don't want something that probably won't leak or probably won't damage you CPU. You want something that other people have used extensively and won't damage anything.

Thats why I'd stick with DangerDen, Dtek, Swiftek, etc.
redleader is offline   Reply With Quote
Unread 09-13-2002, 01:56 PM   #27
gmat
Thermophile
 
gmat's Avatar
 
Join Date: Mar 2001
Location: France
Posts: 1,221
Default

Quote:
Originally posted by Heavy_Equipment
My apologies gmat.

I took it as 28 db + 28 db, for example.
Hehe no problem Yes indeed 3dB more is about twice the noise.

About the rad (to get back to this thread...): The bigger, the better. 5/8" connectors are good with 1/2" tubes. Many ppl are going this way, one huge heater core with one huge 172mm (or bigger !) fan...
Between 2 heater cores, choose the one with more surface area. Thickness does not contribute as well.
gmat is offline   Reply With Quote
Unread 09-15-2002, 06:28 PM   #28
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

I was glad to purchase my 10 feet of PVC "tygon" tubing from Reno Depot here when I discovered that its sold online for about 2.50 dollars US a foot as opposed to the .50 cents I paid in the store. Did I buy the wrong kind? Difference in price is very suspicious.:P
mo is offline   Reply With Quote
Unread 09-16-2002, 08:31 AM   #29
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

Correct me if Im wrong but wouldnt a fill line be better between the pump and the block as opposed to between the reservoir and the pump, for the quickest bleeding/priming?
Or does it all not matter as long as it is on the highest point in the loop?
mo is offline   Reply With Quote
Unread 09-16-2002, 08:38 AM   #30
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

Quote:
Originally posted by mo
I was glad to purchase my 10 feet of PVC "tygon" tubing from Reno Depot here when I discovered that its sold online for about 2.50 dollars US a foot as opposed to the .50 cents I paid in the store. Did I buy the wrong kind? Difference in price is very suspicious.:P
The popular Tygon is of the 3603 formulation. It is designed specifically for lab environments, as it is resistant to the effect of miscellaneous chemicals.

Your tube should be marked with a number. What is it?

You can look up the properties at www.tygon.com . I hear that CL60 is available in Canada, is just as good as Tygon 3603 and is cheaper.

The refill point is best at the top. The bleeding point is best at the bottom. I'm putting a T and a valve at the bottom of my loop, and cutting a hole at the bottom of my case, so that if I need to bleed the thing, I can just prop it up, slip a container under it, then open the valve.
bigben2k is offline   Reply With Quote
Unread 09-17-2002, 02:05 PM   #31
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

ok! its looking good. I've had to mod my case a bit to get everything laid out properly but its starting to come together.

Im trying to figure out the ideal balance between cfm and noise:P My heatercore's size requires two 120 mm fans to cover all its surface area. Got a choise between the Panaflo L1As or M1As. Considering the fact that Im using two fans, would I be able to get away with using the L1As wish are substantially quieter? The shroud looks similar to(but twice the length):
http://www.phlux.co.uk/articles.php?aid=7&page=6
mo is offline   Reply With Quote
Unread 09-17-2002, 04:40 PM   #32
myv65
Cooling Savant
 
Join Date: May 2002
Location: home
Posts: 365
Default

FWIW, three dBA is twice the sound power, but is not perceived as twice as loud by the human ear. Our ears are also logarithmic devices, meaning a 10 dBA increase sounds twice as loud even though it is ten times the sound power level.

dB = 10 * LOG (sound power / 10^(-12)) where sound power is in watts.

dBA merely adds or subtracts a dB or few based on the primary frequency to correspond to the ear's non-linear (vs frequency) sensitivity.
myv65 is offline   Reply With Quote
Unread 09-17-2002, 09:54 PM   #33
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default radiator chiller?

you know...Ive been wracking my brains about how much air I need to get through my rad to get best performance and how to situate it to get the coolest possible air when I thought about chilling the radiator.

Instead of putting a 226W pelt on the cpu (for extreme overclocking) and having to go through insulated tubing ruining the entire look of your cooling rig and worry about condensation etc.. couldnt you just use a couple 80W on the Rad instead? Use stainless steel/anodized aluminum for the shroud and who cares if you have condensation on the rad... Only problem is, how to dissipate all that heat ?:P Perhaps a second Rad loop?
mo is offline   Reply With Quote
Unread 09-18-2002, 06:59 AM   #34
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

Are you talking about converting a rad into a water chiller? It won't work. The cold side of the pelt needs to have good exposure to the coolant. Check out Kevin's article on this.
bigben2k is offline   Reply With Quote
Unread 09-18-2002, 07:35 AM   #35
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

If you use a coldplace that covers the surface area of the rads fins and put the pelt on that then why wouldnt it cool the fluid going through it? the coldplate/pelt would substitute the fan blowing air into the rad. then use some sort of heatsink with the fan on the otherside of the pelt to dissipate it... Due to the large surface area of the rad compared to the processor, we can use low speed 120 mm (or even 175mm) fans at low speed to dissipate the heat and not have as much noise.
mo is offline   Reply With Quote
Unread 09-18-2002, 07:48 AM   #36
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

It's a really strange idea...

I see where you're headed, and I suppose it has some potential, but you're going to loose a lot of contact between the coldplate and the rad fins. You'd be better off trying to mill channels (in a pattern similar to the heatercore) into a big block of copper.
bigben2k is offline   Reply With Quote
Unread 09-18-2002, 08:25 AM   #37
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

eeek! Dunno anyone with a mill, unless someone on here would be willing to help. Any canadians with a Mill feel like trying this out?
Ill finish my original setup first I guess and if it is a good idea afterall I might do this as an upgrade once all works properly.
mo is offline   Reply With Quote
Unread 09-18-2002, 11:26 AM   #38
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

Is then what you mean ?
http://dtekcustoms.safeshopper.com/18/49.htm?778
mo is offline   Reply With Quote
Unread 09-18-2002, 11:36 AM   #39
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

That's half of it...

Read Kevin's experiment!
bigben2k is offline   Reply With Quote
Unread 09-18-2002, 02:21 PM   #40
Brians256
Pro/Staff
 
Brians256's Avatar
 
Join Date: Oct 2001
Location: Klamath Falls, OR
Posts: 1,439
Default

I hope I've read your posts correctly and understand what you proposed.

Using peltiers is best for directly cooling the device. The reason is twofold. First, peltiers are really inefficient so you end up using kilowatts of electricity trying to appreciably cool the water to any significant measure. Second is that when you use a chiller, you end up fighting thermal "leakage" from your hoses. If you want your coolant to be really cold, you might end up moving as much or more heat from leakage as you do from the CPU core itself!

Many experiments have been tried, and unless you have a phase change setup, don't try and chill the coolant. It just doesn't work very well. I've seen about 4 attempts to make it work, and the results aren't nearly as good as if the person just put the peltier units onto the CPU and then used a normal water cooling system for the hot side of the peltier(s).

The best setup, of course, would be a phase change chiller for the coolant combined with a peltier sitting on the chip.

Don't try phase change without a significant investment in time and money though. #rotor has to be one of the most knowledgeable forum members for phase change. He has vast experience with doing it, as opposed to merely theoretical knowledge. But, don't discount many other people too. I can think of bowman and several others who actively work on phase change setups too. There are a lot of smart and experienced people here!
Brians256 is offline   Reply With Quote
Unread 09-18-2002, 03:15 PM   #41
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

Thanks for the input. I am actually avoiding phase change like the plague. A compressor would be way too noisy and the insulated tubing would ruin the who "snazziness" of the pc(I love the ambiance of a watercooled pc). Im simply trying to setup a "quiet" watercooling system and push it to the edge.

The idea originally was to use a coldplate on a radiator and cool the fins of the rad using pelts. While I realise the contact between the two would be less that ideal, it SHOULD cool much better than air blowing through it. This would allow the radiator to be placed inside the case without having to worry about getting fresh outside air for maximum watercooling. It would also remove the variable of ambiant temperature(ambient would actually bring the temperature up a bit due to the tubing), and give people living in a hot tropical climate a great improvement.

As Ben mentioned however, the contact betwen the coldplate and fins isnt going to be good so its going to be very inefficient. And Kevin's experiement wasnt much of a success either and that used a copper block.

The whole thing bothers me like an itch I cant scratch though. Theoretically, if you split the water into multiple small channels inside one big copper block (say 2 or 3 channels per row for 10 rows for example) very much like a regular heatercore works, then it puts the coolant in a very controlled environment that you can cool much easier than if the water was just rushing through a tube 4 or 5 loops and then out. Optimising the block would really just be about using channels that are as thin as possible without ruining the flow rate of the whole circuit. I really have to dig up my thermodynamics book from back in college:P its been 10 years and dont remember much. I just need to start with a fixed sized block, start with a well sized width for say two channels per row, and figure out how long a X Watt peltier would need to cool the volume of liquid in each channel. Based on flow rate I will be able to see if the liquid has enough time to cool down. At the end of the day Ill be able to figure a decent combination between peltier power, and channel size/number or Ill see the folly of my thoughts:P

In anycase, Im getting ahead of myself. Its a theory and I am probably completely off. Especially if others have tried. Once my new system is up, Ill consider that to be the next focus for what my fiance considers computer obsession:P
mo is offline   Reply With Quote
Unread 09-18-2002, 03:18 PM   #42
jtroutma
Cooling Savant
 
jtroutma's Avatar
 
Join Date: Aug 2001
Location: SLO, CA
Posts: 837
Default

One idea I just came up with thanks to mo is to use a submerged radiator in a containter of water that is cooled by pelts on say a HeatSink.

Imagin you have an aluminum box filled with water. One side you have a radiator sitting in the water. In the middle you have say two or three old PII heatsinks with pelts underneath. Pelts are cooling the PII heatsinks and the heat is being removed via separate loop. Other side, you have a pump, wheel, what-ever sturring the water up a bit so that water is passing over the heatsinks then flowing throught the radiator and circulating back around the outside of the box. INSULATE THE BOX!!!

I know that this is not the most effecient way of using pelts BUT its an idea.

Sorry in advance mo if this is what you were talking about earilier.
__________________
Athlon64 X2 4200+ @ 2.5Ghz (250FSB x 10)
OCZ VX 1GB 4000 @ 250FSB (6-2-2-2 timmings)
DFI LANParty nForce4 Ultra-D
SCSI Raid 5 x (3) Cheetah 15K HDDs
LSI Express 500 (128MB cache)
OCZ PowerStream 520W PSU
ATI X850XT PE (Stock)
DTEK WhiteWater + DTEK Custom Radiator
Eheim 1250
jtroutma is offline   Reply With Quote
Unread 09-19-2002, 05:41 AM   #43
gmat
Thermophile
 
gmat's Avatar
 
Join Date: Mar 2001
Location: France
Posts: 1,221
Default

Avoid extra stages.

Heat transfer efficiency drops at each interface.
Putting a rad in *another* res will only further the tranfer path, add extra layers for heat to go through and thus reduce overall efficiency.
To get short here's a little diagram (heat goes from left to right)
Conventional watercooling, or heatpipe:
Code:
CPU -> WB -> Water -> Rad -> Air
Kryotech / Vapochill / phase change look-a-like:
Code:
CPU -> Evaporator -> Freon or other toxic stuff -> Condenser -> Air
Your proposal:
Code:
CPU -> WB -> Water -> Rad -> Water -> Chiller -> Water -> Rad -> Air
or
Code:
CPU -> WB -> Water -> Rad -> Water -> Chiller -> Heatsink -> Air
I hope you now see why it would not be quite efficient...
Put the 'chiller' on the CPU directly (like in every pelt setup).

Actually there *are* industrial-strength water chillers, but costs are quite high, and these are bulky and noisy units.
Instead of putting the rad in a bucket (which is useless), they work like this:
Code:
CPU -> WB -> Water -> Chiller -> Heatsink -> Air
This solution exists, and is used in the industry. But be aware that you must insulate your *whole* tubing lines...
gmat is offline   Reply With Quote
Unread 09-19-2002, 07:04 AM   #44
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

The latter is exactly what Im talking about gmat:
CPU -> WB -> Water -> chiller -> Heatsink -> Air


Extreme performance is far from the object either. Dont need an industrial noisy chiller:P And while efficiency is a goal, it is not the primary goal. Think of it this way.. What do you benefit from cooling your pc to --20 degrees? can you push it all the way to its normal +20 degrees temperatures for a crazy overclock? I have yet to see that. Even with phase change units , from what Ive seen stability seems to start to waver while you are still below zero.
mo is offline   Reply With Quote
Unread 09-19-2002, 10:01 AM   #45
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

You know, if you really, really want to do this, here's what would work best (but still very inneficient): attach the coldplate to the heatercore's tanks.

The problem with the pelts is that they have a very small surface area, and the water would need to pass by it very slowly in order to be cooled. Phase change however can be used over a much bigger surface area, and that's why it makes such a great water chiller.

You need to check out #rotor's site, if you haven't done so already. He converts a dehumidifier into a chiller. Nice, simple and elegant, but most of all, very efficient.
bigben2k is offline   Reply With Quote
Unread 09-24-2002, 04:31 PM   #46
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default UV sterlisation

Has anyone ever considered substituting the use of a biocide in the coolant to using something like this?:

http://www.customsealife.com/NewFiles/u2.html

These are becoming extremely popular in aquarium scenarios, and wondering if they would do the job in a watercooling setup.
mo is offline   Reply With Quote
Unread 09-24-2002, 04:58 PM   #47
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

Yes, but it would have to be sealed, which takes away from the oooh ahhh effect. This type of UV light (UV-C) is not healthy, so the system needs to be contained in that black tube.

They're also very expensive...
bigben2k is offline   Reply With Quote
Unread 09-25-2002, 07:10 PM   #48
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

Alrighty,

Id like to get everyones opinion about one component which is the reservoir. Originally I had planned to take some plexiglass and build a small aquarium-like tank (1.5 litres capacity) that will rest in a niche on the top, right side of my case. Id use a T pointing upwards towards the "tank" so it can serve as a fillpoint, help with removing air from the system, and add a cool look to the case.


Im noticing some people including their reservoirs in the cooling loop though, so the water goes through the reservoir to the pump after the radiator... Does that achieve any extra cooling? Wouldnt the water in the reservoir just rise to the overall system temperature with time?
mo is offline   Reply With Quote
Unread 09-25-2002, 07:27 PM   #49
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

Ah, the old "in-line" VERSUS "submerged" question!
bigben2k is offline   Reply With Quote
Unread 09-25-2002, 07:30 PM   #50
mo
Cooling Savant
 
Join Date: Aug 2002
Location: Canada, Montreal
Posts: 136
Default

hehe. I can completely understand the benefits of having a submerged pump. Im just wondering what good it would be to have a reservoir IN the loop when using an inline pump. I cant come up with any except kill the waterflow which is a bad thing.
mo is offline   Reply With Quote
Reply


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

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 02:42 AM.


Powered by vBulletin® Version 3.7.4
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
(C) 2005 ProCooling.com
If we in some way offend you, insult you or your people, screw your mom, beat up your dad, or poop on your porch... we're sorry... we were probably really drunk...
Oh and dont steal our content bitches! Don't give us a reason to pee in your open car window this summer...