Helppp... Water cooling temps too high!!

babybunny · 07-09-2002, 07:21 PM

Alright, I finally got my setup working with the two AquaVia 1300's in a row.

I've got a black edge waterblock from becooling, a chipjacket form becooling, a ultimate radiator from dtek, an enermax 120mm adjustible fan blowing at high speed onto the radiator, 3/8' tygon tubing, and an Athlon 1.333 GHz with ACII.

The order goes: pump, pump, radiator, CPU, Mobo, and back to pump in a closed system.

My temps are 46C on idle. I dare not try what it would be on load. This is worse than when I had my delta at half speed. The ambient temp is 28C. I'm using the onboard temp sensor from my Shuttle AK31A.

This is my theory: The Aquavia's make a crapload of heat when working, so I think they're raising the temperature of the water. So I shut one off and now the temps are 45C at idle.

What the heck is going on?

bigben2k · 07-09-2002, 07:29 PM

Nah... The Aqua's are not the problem.

Here's the roundup of the first set of questions:
1-did you lap the waterblock?
2-did you use AS?
3-How did you clamp the WB?
4-do you notice any air in the system?
5-what did you add to the water?
6-is your CPU overclocked?
7-is your CPU overvolted?

Let's go!

babybunny · 07-09-2002, 07:39 PM

The waterblock was lapped to a near mirror finish, and very flat.

I used ASII.

I clamped the WB with becoolings's lexan clamp kit, and I made sure that it was flush with the CPU and the shim.

There is no air in the system, I did a very thorough job of bleeding it which took me a couple of hours.

I added Royal Purple to the distilled water. 4oz for a gallon, as per instructions.

The CPU is DOWNclocked to 1.266 GHz (from 1.333 GHz).

The CPU is UNDERvolted at 1.66V, and I can't go lower.

bigben2k · 07-09-2002, 07:41 PM

I'd try the bucket test, to check your flow, unless you can somehow see it.

invariant · 07-09-2002, 07:47 PM

Try top measure your water temp at any point in your system. That will tell you a lot!

babybunny · 07-09-2002, 07:49 PM

I tested out the whole system before mounting it and the water flow rate was pretty fast. The two pumps sent a two foot high column of water into the air. With one pump the water column was a few inches over a foot. Perhaps the water flow is too fast?

ambient · 07-09-2002, 09:36 PM

have you tried reversing the order of the system, i.e. pump>pump>cpu>nb>rad?

toddalaska · 07-09-2002, 09:38 PM

the hagged edge difinitly isnt the problem I vcan tell you that much.

the oc system in my sig never gets close to 46 and its oced over 500 mhz

with a similar setup

bigben2k · 07-10-2002, 08:32 AM

...then it's a mounting problem. Has anyone used that clamp before?

airspirit · 07-10-2002, 02:13 PM

Same thing here:

46-49C @ 41 Case in 85-90F ambient

Black Edge 3/8"
1/2" Reinforced SurgTube
Badazz NurseryPro pump (equiv 2xVia 1300 or more)
BIXr2
1900+@1766 (12.5x141x1.825v)
MSI KT333Ultra
256 2100 PNY DDR
Radeon 8500LE w/ retail BIOS @ 290/290
edit: HyperLube surfecant

Diagram of system:

bigben2k · 07-10-2002, 02:24 PM

Quote:

Originally posted by babybunny

I clamped the WB with becoolings's lexan clamp kit, and I made sure that it was flush with the CPU and the shim.

Did you say SHIM? Try without it.

(I like shims, but they can be a source of problems for good contact)

airspirit · 07-10-2002, 03:13 PM

Shims a source of problems? Nah ... mine just dug trenches in my processor and fried it. Otherwise my system was fine except for the fact that it wouldn't boot up ...

edit: I am using that clamp and it provides good pressure/contact on the die if tightened correctly. I found that I needed to wrap the bottom of the bolts a couple times around with duct tape to space them from the board and prevent misplacement.

bigben2k · 07-10-2002, 03:22 PM

Quote:

Originally posted by airspirit
Shims a source of problems? Nah ... mine just dug trenches in my processor and fried it. Otherwise my system was fine except for the fact that it wouldn't boot up ...

I tend to agree.

The more I think about it, the more I think that this whole shim thing is a joke.

I mean, we're trying to apply pressure to this CPU, but we put a shim, which reduces the pressure on the 5 contact points (core + 4 pads). If the shim is too thin, then it won't do anything at all (might as well not have one). If it's too thick, then it's trouble.

I don't believe that AMD manufactures its procs with exactly the same height, there's got to be a margin of error here. I don't believe that AMD has ever issued physical dimensions on their CPUs, as far as how much the core is above the base, so how is a shim suppose to fit my CPU exactly?

It seems to me that shims defeat the whole purpose of mounting something on a CPU.

airspirit · 07-10-2002, 03:29 PM

I wish I could get my hands on the Becooling bastids that sold me that thing. First, I ordered the mica shim but got the steel one, and once I got the steel one (damn me for not inspecting it closer) it had sharp edges on all of the interior cuts. It hacked 1/3 of the way through the card inboard of the L1 and L10 (?) bridges when I tightened down my water block.

babybunny · 07-10-2002, 03:53 PM

Do you think I really should try it without the shim? I'm scared. I've cracked a core really easily before and have no desire to do it again. Also, how much heat is an aqua via 1300 supposed to add to the water inline? Mine's adding 4C.

bigben2k · 07-10-2002, 03:58 PM

Quote:

Originally posted by babybunny
Do you think I really should try it without the shim? I'm scared. I've cracked a core really easily before and have no desire to do it again. Also, how much heat is an aqua via 1300 supposed to add to the water inline? Mine's adding 4C.

I need to see that clamp. Do you have a pic, or link?

babybunny · 07-10-2002, 04:07 PM

http://becooling.safeshopper.com/6/58.htm?21

bigben2k · 07-10-2002, 04:11 PM

What the heck is that grey part, a spring or a spacer?

chazz469 · 07-10-2002, 04:18 PM

Springs, I'm pretty sure.

bigben2k · 07-10-2002, 04:20 PM

Ok, now let's hear it from the guy that has it in his hands...

bigben2k · 07-10-2002, 04:33 PM

I've searched the site, and found this:

"The lexan clamp will allow more pressure, but be careful. More than the 20 pounds AMD recommends, may result in damage to the cpu.. With the lexan clamp option, tighten the 4 thumbscrews in small increments in a criss cross pattern. "

found here:
http://www.aquastealth.com/install.html

It's no wonder that they recommend shims!

Consider replacing the mount with this:
http://www.caseetc.com/cgi-bin/caseetc/WC-423.html

or, even easier, get the springs. (anyone know where?)

airspirit · 07-10-2002, 04:33 PM

Those are springs and they give little resistance to tightening. I found that if you tighten the bolts until the springs are almost fully compressed, you get good CPU pressure without cracking it.

bigben2k · 07-10-2002, 04:37 PM

All right then. If you want to play it on the safe side, I'd lap the shim. Otherwise, try it without (at your own risk!)

babybunny · 07-10-2002, 04:49 PM

yeah, they're definitely springs.
The order of assembly goes from top to bottom: top of long screw, thumb screw, spring, lexan, and the long screw travels to the other side of the mobo.

07-09-2002, 07:21 PM	#1
babybunny Cooling Savant Join Date: Jun 2002 Location: USA - Ohio Posts: 120	Helppp... Water cooling temps too high!! Alright, I finally got my setup working with the two AquaVia 1300's in a row. I've got a black edge waterblock from becooling, a chipjacket form becooling, a ultimate radiator from dtek, an enermax 120mm adjustible fan blowing at high speed onto the radiator, 3/8' tygon tubing, and an Athlon 1.333 GHz with ACII. The order goes: pump, pump, radiator, CPU, Mobo, and back to pump in a closed system. My temps are 46C on idle. I dare not try what it would be on load. This is worse than when I had my delta at half speed. The ambient temp is 28C. I'm using the onboard temp sensor from my Shuttle AK31A. This is my theory: The Aquavia's make a crapload of heat when working, so I think they're raising the temperature of the water. So I shut one off and now the temps are 45C at idle. What the heck is going on?

07-09-2002, 07:29 PM	#2
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Nah... The Aqua's are not the problem. Here's the roundup of the first set of questions: 1-did you lap the waterblock? 2-did you use AS? 3-How did you clamp the WB? 4-do you notice any air in the system? 5-what did you add to the water? 6-is your CPU overclocked? 7-is your CPU overvolted? Let's go! __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-09-2002, 07:41 PM	#4
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	I'd try the bucket test, to check your flow, unless you can somehow see it. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-09-2002, 09:36 PM	#7
ambient Cooling Savant Join Date: Jan 2002 Location: Planet Earth Posts: 312	have you tried reversing the order of the system, i.e. pump>pump>cpu>nb>rad? __________________ water cooled 1.4 tbird @ 1580 143x11 vcore 1.88 vio 3.55

07-09-2002, 09:38 PM	#8
toddalaska Cooling Neophyte Join Date: Jun 2002 Location: ALASKA Posts: 66	the hagged edge difinitly isnt the problem I vcan tell you that much. the oc system in my sig never gets close to 46 and its oced over 500 mhz with a similar setup __________________ Epox 8k5a3+ 2400+ @ 2340 (213*11.5) 512 corsair 3200 213 fsb turbo settings geforce 4 ti4200 -13619 3dmarks 2 ibm 120gxp raid 0 400 watts WINXP PRO BIG BLACK ANTEC CASE Water cooled tc4 - custom res and shroud by webmedic

07-09-2002, 07:39 PM	#3
babybunny Cooling Savant Join Date: Jun 2002 Location: USA - Ohio Posts: 120	The waterblock was lapped to a near mirror finish, and very flat. I used ASII. I clamped the WB with becoolings's lexan clamp kit, and I made sure that it was flush with the CPU and the shim. There is no air in the system, I did a very thorough job of bleeding it which took me a couple of hours. I added Royal Purple to the distilled water. 4oz for a gallon, as per instructions. The CPU is DOWNclocked to 1.266 GHz (from 1.333 GHz). The CPU is UNDERvolted at 1.66V, and I can't go lower.

07-09-2002, 07:47 PM	#5
invariant Cooling Neophyte Join Date: Jun 2002 Location: One step beyond... Posts: 6	Try top measure your water temp at any point in your system. That will tell you a lot!

07-09-2002, 07:49 PM	#6
babybunny Cooling Savant Join Date: Jun 2002 Location: USA - Ohio Posts: 120	I tested out the whole system before mounting it and the water flow rate was pretty fast. The two pumps sent a two foot high column of water into the air. With one pump the water column was a few inches over a foot. Perhaps the water flow is too fast?

07-10-2002, 08:32 AM	#9
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	...then it's a mounting problem. Has anyone used that clamp before? __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-10-2002, 03:13 PM	#12
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	Shims a source of problems? Nah ... mine just dug trenches in my processor and fried it. Otherwise my system was fine except for the fact that it wouldn't boot up ... edit: I am using that clamp and it provides good pressure/contact on the die if tightened correctly. I found that I needed to wrap the bottom of the bolts a couple times around with duct tape to space them from the board and prevent misplacement. Last edited by airspirit; 07-10-2002 at 03:16 PM.

07-10-2002, 03:29 PM	#14
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	I wish I could get my hands on the Becooling bastids that sold me that thing. First, I ordered the mica shim but got the steel one, and once I got the steel one (damn me for not inspecting it closer) it had sharp edges on all of the interior cuts. It hacked 1/3 of the way through the card inboard of the L1 and L10 (?) bridges when I tightened down my water block.

07-10-2002, 03:53 PM	#15
babybunny Cooling Savant Join Date: Jun 2002 Location: USA - Ohio Posts: 120	Do you think I really should try it without the shim? I'm scared. I've cracked a core really easily before and have no desire to do it again. Also, how much heat is an aqua via 1300 supposed to add to the water inline? Mine's adding 4C.

07-10-2002, 04:07 PM	#17
babybunny Cooling Savant Join Date: Jun 2002 Location: USA - Ohio Posts: 120	http://becooling.safeshopper.com/6/58.htm?21

07-10-2002, 04:18 PM	#19
chazz469 Cooling Savant Join Date: Jun 2002 Location: San Diego, CA Posts: 150	Springs, I'm pretty sure. __________________ "I believe that the moment is near when by a procedure of active paranoiac thought, it will be possible to systematize confusion to the total discrediting of the world of reality" -Salvador Dali AMD XP1600+ @1790MHz and rethinking my cooling.... MAZE3 --> DD Cooler Cube --> Eheim 1048 512MB PC2700 DDR MSI KT3 Ultra-ARU 2x 20GB Maxtor in RAID 0 1x 30GB Maxtor --> WinXP GF3 Ti200 @ 240/499

07-10-2002, 04:20 PM	#20
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Ok, now let's hear it from the guy that has it in his hands... __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-10-2002, 04:33 PM	#21
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	I've searched the site, and found this: "The lexan clamp will allow more pressure, but be careful. More than the 20 pounds AMD recommends, may result in damage to the cpu.. With the lexan clamp option, tighten the 4 thumbscrews in small increments in a criss cross pattern. " found here: http://www.aquastealth.com/install.html It's no wonder that they recommend shims! Consider replacing the mount with this: http://www.caseetc.com/cgi-bin/caseetc/WC-423.html or, even easier, get the springs. (anyone know where?) __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-10-2002, 04:37 PM	#23
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	All right then. If you want to play it on the safe side, I'd lap the shim. Otherwise, try it without (at your own risk!) __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-10-2002, 04:49 PM	#24
babybunny Cooling Savant Join Date: Jun 2002 Location: USA - Ohio Posts: 120	yeah, they're definitely springs. The order of assembly goes from top to bottom: top of long screw, thumb screw, spring, lexan, and the long screw travels to the other side of the mobo.

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