So confused... Help me please!

KlevrPhelo · 09-16-2003, 02:35 AM

Hi everyone,

Hope someone out there can help me out. Correct me if I'm wrong but, I always thought a water cooled computer was supposed to have very good temps, at least better than teh conventional heatsink/fan method. For some reason the temps I'm getting are kinda high (I think). Here're the relative specs of the machine:

Gigabyte GA-7DPXDW-P MOBO
2 x AMD Athlon MP 2600+ (Water cooled) w/ Arctic Silver III Thermal Adhesive
ATI All-In-Wonder 9700 Pro (Water cooled) w/ Arctic Silver Ceramique Thermal Adhesive

I'm using 2 x Swiftech MCW5000 Rev 1.0 water blocks on my CPUs, an MCW50 water block on the video card, an MCP300 waterpump, a BIX Rad with a 120mm fan pulling air through. All connected using Tygon R-3603 1/2" ID tubing. I'm pushing Distilled water w/ Silkolene Pro CCA as an additive mixed at 1%/USGallon.

See the entire setup at:
http://members.cox.net/klevrphelo/

Anyway, according the BIOS "PC Health" area, at boot up my initial temps are 28C for the system, 37C for CPU0 and 47 for CPU1. After running the system for about 10 - 15 minutes my idle temps on my CPUs go up to 43C for CPU0 and 53 for CPU1.

Shouldn't those be lower? Especially since it's at initial boot, and the comp isn't really doing anything? Or am I off my rocker? Or am I doing everything completely wrong?

I'm so confused... =(

sevisehda · 09-16-2003, 03:35 AM

I hope that you didn't use the adhesive varient like you stated. You should have used the thermal compound not the thermal adhesive. Either way my guess is that your coolant is heating up because your radiator isn't disapaiting the heat.

Nice rig by the way, where did you get the UV window trim?

g.l.amour · 09-16-2003, 05:46 AM

awesome!!

KlevrPhelo · 09-16-2003, 09:06 AM

Quote:

Originally posted by sevisehda
I hope that you didn't use the adhesive varient like you stated. You should have used the thermal compound not the thermal adhesive. Either way my guess is that your coolant is heating up because your radiator isn't disapaiting the heat.

Nice rig by the way, where did you get the UV window trim?

Oops... did I say that? =) I meant to say compound, not adhesive.
The UV window trim is actually translucent molding I got from pcmods.com. The glow effect was actually an "accident" to tell you the truth.
As far as the heat dissipation issue, what can I do? Should I add another fan to the rad to get a Push/Pull effect?

trit187 · 09-16-2003, 10:00 AM

before you try 2 fans you might want to try making a shroud for the radiator (or 2 fans w/2 shrouds), and if that doesn't work, maby try getting a heatercore for a local autoshop and see if that helps your temps.
also keep in mind the inaccuracy are amd temp probes

MadHacker · 09-16-2003, 11:35 AM

trit187 i think is right about heater cores. you may need to get a biiger one....

my sugestion is to move the power supply to the top...
then you would have room for a large heater core and have 2 120mm fans blowing through it. the case is definatly large enough for it.

why kind of case is it?

Boli · 09-16-2003, 12:42 PM

I think its a Lian Li PC-76 I was thinking a ut getting one myself but I told myself not to be stupid and got myself a PC-70 instead.

I agree with the above comments that a better was to dissipate the heat would be the best approach.

You have two real solutions, improve the radiator... or improve the water blocks. Improving the water blocks would get the heat from the cores easier into the water but you will still need to get rid of the excess heat even of you did see lower temps it would only be temporary as the water tem.

OK improving the radiator, you have to look at several factors:

- The speed of the air entering the rad

So in essence making your fan run faster will increase the airflow and thus lower the temps. but that makes the computer noisy and no one wants that. sop we look at the other factors.

- Restrictions in the airflow

At the centre of the radiator a you have it now is the so-called "dead zone" the is where the hub of the ran restricts the airflow and thus the radiator isn't being actively cooled at this point. Moving the fan away from the rad about 1"-2" will compensate for this. This is called the shroud technique and is much favoured, as you do not need more powerful/faster fans, but you still achieve better temps.

- The temperature of the air entering the rad

Another favoured technique (and one I will be using) is to make sure the radiator draws in the coldest air possible through its fins. Many rads are mounted on the front of their case for this very purpose air but you can construct a "shortcut" to the air by a number of means. Even a fan on the top of the case pulling in fresh air will help with your temps but the best way is to "seal off" the rad from the rest of the case in a wind tunnel. turbokeu has a fine example of this

What I am saying in essence is that you don’t run out and get some new hardware every time you discover an irregularity, you find out exactly what you can do to improve the existing hardware and that way you learn more than if you always buy the upgrade.

~ Boli

superart · 09-16-2003, 12:52 PM

a couple suggestions:
1.) you should try moving your res a higher. This will help protect against bubbles getting into the loop.

2.) Try wireing the two blocks in "paralell" instead of in "series" You will get better efficiency if you split the waterflow into 2 lseperate lines going to and from the blocks.

I really like the system. Nice Job!!

Quarter · 09-16-2003, 01:08 PM

The high temp difference between the two chips might point to poor and or restrictive flow in the system.
A better pump and or running the CPU blocks in parallel may help even the temps.
Is the cooler of the two chips the first one in the loop?

That radiator is rather small for two 2600+'s and a 9700 Pro.
Building a shroud would definitely help, but may not be enough.

Boli · 09-16-2003, 07:25 PM

This is true.... you should have AT LEAST a 120.2 rad(s) in that setup... though I would look more into making the thing as efficient as possible if more for a learning curve than anything else.

~Boli

BrianW · 09-16-2003, 07:52 PM

Bigger rad, or more rads. More airflow over rads. I would run the cpu blocks in paralel, and add another small pump after the GPU block.

BrianW

nOv1c3 · 09-17-2003, 01:14 AM

Very nice set up KlevrPhelo ..I did notice one thing on your bottom block
looks like the inlet and outlet is plumb wrong ..I think the intake should be on the bottom and outlet on top ..Dont know if it maters that much but it would help bleeding

But i have to agree with everybody here that you just need more rad to cool those 2 CPU,s

KlevrPhelo · 09-17-2003, 01:52 PM

You guys are great! All your comments and suggestions have been extremely helpful.

One more thing: How about a custom made inline water chiller? Think that'll make a difference?

BrianW · 09-17-2003, 04:18 PM

Whole new can of worms....

BrianW

09-16-2003, 02:35 AM	#1
KlevrPhelo Cooling Neophyte Join Date: Aug 2002 Location: San Diego, CA Posts: 18	So confused... Help me please! Hi everyone, Hope someone out there can help me out. Correct me if I'm wrong but, I always thought a water cooled computer was supposed to have very good temps, at least better than teh conventional heatsink/fan method. For some reason the temps I'm getting are kinda high (I think). Here're the relative specs of the machine: Gigabyte GA-7DPXDW-P MOBO 2 x AMD Athlon MP 2600+ (Water cooled) w/ Arctic Silver III Thermal Adhesive ATI All-In-Wonder 9700 Pro (Water cooled) w/ Arctic Silver Ceramique Thermal Adhesive I'm using 2 x Swiftech MCW5000 Rev 1.0 water blocks on my CPUs, an MCW50 water block on the video card, an MCP300 waterpump, a BIX Rad with a 120mm fan pulling air through. All connected using Tygon R-3603 1/2" ID tubing. I'm pushing Distilled water w/ Silkolene Pro CCA as an additive mixed at 1%/USGallon. See the entire setup at: http://members.cox.net/klevrphelo/ Anyway, according the BIOS "PC Health" area, at boot up my initial temps are 28C for the system, 37C for CPU0 and 47 for CPU1. After running the system for about 10 - 15 minutes my idle temps on my CPUs go up to 43C for CPU0 and 53 for CPU1. Shouldn't those be lower? Especially since it's at initial boot, and the comp isn't really doing anything? Or am I off my rocker? Or am I doing everything completely wrong? I'm so confused... =( __________________ I'm a newb... can you tell?

09-16-2003, 05:46 AM	#3
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	awesome!! __________________ yo soy un tiburón

09-16-2003, 11:35 AM	#6
MadHacker Cooling Savant Join Date: May 2003 Location: Okotoks, A.B. Canada Posts: 726	trit187 i think is right about heater cores. you may need to get a biiger one.... my sugestion is to move the power supply to the top... then you would have room for a large heater core and have 2 120mm fans blowing through it. the case is definatly large enough for it. why kind of case is it? __________________ "Great spirits have always encountered violent opposition from mediocre minds" - (Einstein)

09-16-2003, 12:42 PM	#7
Boli Cooling Savant Join Date: Apr 2003 Location: Blackburn / Dundee Posts: 451	I think its a Lian Li PC-76 I was thinking a ut getting one myself but I told myself not to be stupid and got myself a PC-70 instead. I agree with the above comments that a better was to dissipate the heat would be the best approach. You have two real solutions, improve the radiator... or improve the water blocks. Improving the water blocks would get the heat from the cores easier into the water but you will still need to get rid of the excess heat even of you did see lower temps it would only be temporary as the water tem. OK improving the radiator, you have to look at several factors: - The speed of the air entering the rad So in essence making your fan run faster will increase the airflow and thus lower the temps. but that makes the computer noisy and no one wants that. sop we look at the other factors. - Restrictions in the airflow At the centre of the radiator a you have it now is the so-called "dead zone" the is where the hub of the ran restricts the airflow and thus the radiator isn't being actively cooled at this point. Moving the fan away from the rad about 1"-2" will compensate for this. This is called the shroud technique and is much favoured, as you do not need more powerful/faster fans, but you still achieve better temps. - The temperature of the air entering the rad Another favoured technique (and one I will be using) is to make sure the radiator draws in the coldest air possible through its fins. Many rads are mounted on the front of their case for this very purpose air but you can construct a "shortcut" to the air by a number of means. Even a fan on the top of the case pulling in fresh air will help with your temps but the best way is to "seal off" the rad from the rest of the case in a wind tunnel. turbokeu has a fine example of this What I am saying in essence is that you don’t run out and get some new hardware every time you discover an irregularity, you find out exactly what you can do to improve the existing hardware and that way you learn more than if you always buy the upgrade. ~ Boli __________________ 1800+ @ 2247 (214x10.5) - STABLE, 512MB PC3700 TwinX Cosair RAM, NF7-S v2.0, GeForce3 Ti200 Parallel BIM, 120.1 Thermochill, Eheim 1048, Maze 3, Maze4 GPU, "Z" chipset, 1/2" tubing, PC-70: 5x120mm & 9x80mm fans. Internet Server & second machine (folding 24/7): 512MB DDR RAM, XP2000+

09-16-2003, 07:25 PM	#10
Boli Cooling Savant Join Date: Apr 2003 Location: Blackburn / Dundee Posts: 451	This is true.... you should have AT LEAST a 120.2 rad(s) in that setup... though I would look more into making the thing as efficient as possible if more for a learning curve than anything else. ~Boli __________________ 1800+ @ 2247 (214x10.5) - STABLE, 512MB PC3700 TwinX Cosair RAM, NF7-S v2.0, GeForce3 Ti200 Parallel BIM, 120.1 Thermochill, Eheim 1048, Maze 3, Maze4 GPU, "Z" chipset, 1/2" tubing, PC-70: 5x120mm & 9x80mm fans. Internet Server & second machine (folding 24/7): 512MB DDR RAM, XP2000+

09-16-2003, 03:35 AM	#2
sevisehda Cooling Savant Join Date: May 2003 Location: NY Posts: 234	I hope that you didn't use the adhesive varient like you stated. You should have used the thermal compound not the thermal adhesive. Either way my guess is that your coolant is heating up because your radiator isn't disapaiting the heat. Nice rig by the way, where did you get the UV window trim?

09-16-2003, 10:00 AM	#5
trit187 Cooling Savant Join Date: May 2003 Location: shanghai, china Posts: 200	before you try 2 fans you might want to try making a shroud for the radiator (or 2 fans w/2 shrouds), and if that doesn't work, maby try getting a heatercore for a local autoshop and see if that helps your temps. also keep in mind the inaccuracy are amd temp probes

09-16-2003, 01:08 PM	#9
Quarter Cooling Neophyte Join Date: Aug 2003 Location: San Diego, USA Posts: 15	The high temp difference between the two chips might point to poor and or restrictive flow in the system. A better pump and or running the CPU blocks in parallel may help even the temps. Is the cooler of the two chips the first one in the loop? That radiator is rather small for two 2600+'s and a 9700 Pro. Building a shroud would definitely help, but may not be enough.

09-16-2003, 07:52 PM	#11
BrianW Cooling Savant Join Date: Jun 2001 Location: Desert City in California Posts: 631	Bigger rad, or more rads. More airflow over rads. I would run the cpu blocks in paralel, and add another small pump after the GPU block. BrianW __________________ Water Cooled Inwin Q500 (Dual Rads: Rad1 = DTEK Pro Core \| Rad2 = Blick Ice Estreme, Hydor L30, Dangerden Maze2, Bay Res Typhoon Reservoir, 1/2 " DD Tygon Thick Wall Hose). Flow: Res, Pump, CPU watervlock, Y into both rads, both rads into res independently. Athlon XP 1800+ (@ 1731 - 150mhz fsb.), on a Asus A7N266-c, and a Radeon 9000 *waiting for RMA'd Saphire 9800 ultra from Newegg)

09-17-2003, 01:14 AM	#12
nOv1c3 Crazy Stupid Join Date: Nov 2002 Location: Dallas texas Posts: 149	Very nice set up KlevrPhelo ..I did notice one thing on your bottom block looks like the inlet and outlet is plumb wrong ..I think the intake should be on the bottom and outlet on top ..Dont know if it maters that much but it would help bleeding But i have to agree with everybody here that you just need more rad to cool those 2 CPU,s

09-17-2003, 01:52 PM	#13
KlevrPhelo Cooling Neophyte Join Date: Aug 2002 Location: San Diego, CA Posts: 18	You guys are great! All your comments and suggestions have been extremely helpful. One more thing: How about a custom made inline water chiller? Think that'll make a difference? __________________ I'm a newb... can you tell?

09-17-2003, 04:18 PM	#14
BrianW Cooling Savant Join Date: Jun 2001 Location: Desert City in California Posts: 631	Whole new can of worms.... BrianW __________________ Water Cooled Inwin Q500 (Dual Rads: Rad1 = DTEK Pro Core \| Rad2 = Blick Ice Estreme, Hydor L30, Dangerden Maze2, Bay Res Typhoon Reservoir, 1/2 " DD Tygon Thick Wall Hose). Flow: Res, Pump, CPU watervlock, Y into both rads, both rads into res independently. Athlon XP 1800+ (@ 1731 - 150mhz fsb.), on a Asus A7N266-c, and a Radeon 9000 *waiting for RMA'd Saphire 9800 ultra from Newegg)

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