Sign in with
Sign up | Sign in
Your question

Help! Installed water cooling; computer not happy

Last response: in Overclocking
Share
March 23, 2007 3:01:20 AM

Hey all,

I just installed a water cooling system in my case. The parts I used are:
Apogee GT block, Swiftech heat exchanger, a swiftech MCP655 pump, 1/2" cooling lines. I am using a T-line reservoir system.

My motherboard is a ASUS A8N32 Deluxe. I also installed the Northbridge fan needed for passive cooling.

The problem I am having is the fact that my motherboard thinks my CPU is running too hot and shuts the computer down. Is there something I am missing? I used an infrared temp gun on the processor and the temps were fairly stable. I tried ignoring the temp sensor in the BIOS, but it keeps shutting down after 30 seconds or so. Is there anything else I should do to fix this problem? Thanks
March 23, 2007 3:52:10 AM

"I used an infrared temp gun on the processor and the temps were fairly stable"

Errrr as in it was always 80deg C and didnt fall off a table???

Can you just tell us what temps you got???

Pics?? Is the water block seated correctly with some thermal paste? Can you see water movin though the tubes? Do you have some fans blowin on the rad? Did you disable CPU fan detection in BIOS?
March 23, 2007 6:53:11 AM

You'll also need to check a few other things.

Some water coolers have a link to the main case power switch, used to shut the system down if it gets too hot or some other alert is triggered. Check that.

wun mentioned checking the block to see if it is properly seated. Check that. Also check thermal paste.

If your CPU temp is over 60ºC or so, you're going to run into issues.

Err, could you also post us your temps (idle/load)?
Related resources
March 23, 2007 8:08:35 AM

Are you sure you have the outflow from the pump going to the top of the radiator and not the bottom????It is a lot harder to pump water up a rad than just to pump it to the top and let gravity help move the larger volume down.Goodluck.

Dahak

AMD X2-4400+@2.6 S-939
EVGA NF4 SLI MB
2X EVGA 7950GT KO IN SLI
4X 512MB CRUCIAL BALLISTIX DDR500
WD300GIG HD
ACER 22IN WIDESCREEN LCD 1600X1200
THERMALTAKE TOUGHPOWER 850WATT PSU
COOLERMASTER MINI R120
3DMARK05 13,471
March 23, 2007 10:08:20 AM

I couldn't get a read on the temps after the computer started shutting off constantly.

Before that however, I used the infrared temp gun to get a read off the cpu and it wouldn't go over 38C. However, in the BIOS (before the comp. shut down) I saw a temp of 70C. I set the temp reading of the CPU to "IGNORE" and I stopped receiving the CPU fan warning during boot up.

I made sure that the CPU block was seated properly several times. I'm using AS5 as paste. I also seated the voltage regulator fan that came packaged with my A8N32.

The radiator is a Black Ice Xtreme 120. It is mounted on the top (underside of the top that is) of my case with the barbs facing downward. I have a shroud and fan attached to it to SUCK air out through the radiator. As an intake, I modded my case side panel to house a 250mm fan.

I'm aware of the fact that the temp sensors on the motherboards themselves are fairly inaccurate and with water cooling in place I was expecting the temps to go up. Is there a safety feature that I've missed that is automatically shutting down the PC? I put so much energy into cutting the top of the case, and money into the parts, and time in making sure that are no leaks and airbubbles in the loop. I'm just afraid that for some reason I won't be able to get this rig up and running properly with water cooling.
March 23, 2007 10:12:36 AM

Quote:
Are you sure you have the outflow from the pump going to the top of the radiator and not the bottom????It is a lot harder to pump water up a rad than just to pump it to the top and let gravity help move the larger volume down.Goodluck.

Dahak

AMD X2-4400+@2.6 S-939
EVGA NF4 SLI MB
2X EVGA 7950GT KO IN SLI
4X 512MB CRUCIAL BALLISTIX DDR500
WD300GIG HD
ACER 22IN WIDESCREEN LCD 1600X1200
THERMALTAKE TOUGHPOWER 850WATT PSU
COOLERMASTER MINI R120
3DMARK05 13,471


I'm sorry Dahak, I didn't mean to completely ignore your question. My radiator is directly above the pump and the coolant is flowing quite efficiently through the loop. I have the pump set at 3 for noise and power consideration and it seems fine. The pump I believe has quite a few feet of head so I wasn't worried about the rad's location (9 in. directly above pump).
March 23, 2007 11:08:22 AM

Quote:
Are you sure you have the outflow from the pump going to the top of the radiator and not the bottom????It is a lot harder to pump water up a rad than just to pump it to the top and let gravity help move the larger volume down.Goodluck.

Dahak

AMD X2-4400+@2.6 S-939
EVGA NF4 SLI MB
2X EVGA 7950GT KO IN SLI
4X 512MB CRUCIAL BALLISTIX DDR500
WD300GIG HD
ACER 22IN WIDESCREEN LCD 1600X1200
THERMALTAKE TOUGHPOWER 850WATT PSU
COOLERMASTER MINI R120
3DMARK05 13,471


I'm sorry Dahak, I didn't mean to completely ignore your question. My radiator is directly above the pump and the coolant is flowing quite efficiently through the loop. I have the pump set at 3 for noise and power consideration and it seems fine. The pump I believe has quite a few feet of head so I wasn't worried about the rad's location (9 in. directly above pump).

did you ran your watercooling kit for around 2 hours to remove the bubbles?
note that a single big bubble in the cpu block means HOT TEMPS.
March 23, 2007 11:13:32 AM

Quote:
did you ran your watercooling kit for around 2 hours to remove the bubbles?
note that a single big bubble in the cpu block means HOT TEMPS.


I didn't run it for two hours. I did run it for a little over an hour. I tipped the case a number of times and took my time filling the t-line. I hear no cavitation noise coming from the pump or the loop. I guess I can hook it up to my test power supply unit and run it seperately and see if there is a bubble in the block itself.

I'm at work so I can't do anything atm but I'm going to try reseating the block when I get home (tighten screws) and test out my Venice 3500 as well to make sure that the IHS on my X2 isn't at fault (it's been reseated many times and there are scuff marks).
March 23, 2007 11:26:51 AM

So the IHS has been removed and reseated several times??? Sounds like it was damaged or the IHS is no longer making contact with the core. Why don't you get an Opteron 165 from the Egg and plug it into that board. They are dirt cheap and overclock like banshees whne watercooled.
March 23, 2007 11:40:27 AM

Quote:
So the IHS has been removed and reseated several times??? Sounds like it was damaged or the IHS is no longer making contact with the core. Why don't you get an Opteron 165 from the Egg and plug it into that board. They are dirt cheap and overclock like banshees whne watercooled.


I'll test the temps using the Venice after I try tightening the screws and if the results are nice, I'll definitely know that the IHS has been indeed damaged.

However, the Arctic Cooling Freezer 64 works great. I just don't get it. On the other hand, the Thermalright 120 SI refused to cool my chip when I tried it a few months ago. So...my head's spinning now.

I'll look up the Opty you mentioned. OC'ing like a banshee sounds like fun.
March 23, 2007 12:33:05 PM

Maybe i misuderstood. I thought you had said that the IHS had been reseated several times. Maybe we aren't taliking about the same thing. Regardless, I ran an Opte 165 on that board for months at 3.0GHz 1.47v on water. The Egg is shipping some really good stepping right now an dthe price is below $160.
March 23, 2007 12:46:35 PM

Edit- Sorry Merc, I'm an idiot. I meant the entire processor unit has been reseated several times. Maybe while twisting and pulling on the HSF during one of those operations I might have pulled a little too hard and maybe tweaked the IHS away from its proper seated place. Is that possible?

[Ignore]Heya Merc, I meant the IHS as in the nickel covering that is essentially the surface that gets the thermal grease is applied.[/Ignore]

1. I'm going to try out my old Venice chip and see if the temps shoot up on that as well. If the temp does not shoot up causing my system to shut down, I'll know there's something wrong with my X2 4800's IHS. Maybe all those times I tried different air cooling HSF did indeed damage the IHS or maybe pulled it off away from the procs themselves. Arcting Cooling 64 Freezer does clamp down quite hard on the processor and therefore maybe that's why it's the HSF that works - #2

2. I'm going to tighten the screws on the hold down plate for my Apogee GT and see if that'll fix it.

3. I'm going to unmount the waterblock and run the WC loop and see if I can get more air bubbles to come out.

4. I'll order the Opty 165 if none of those measures fix my little problem here.

Let me tell you, I love this hobby, but I'm going through a lot of tylenol because of it. :lol: 
March 23, 2007 12:55:59 PM

Ok, I thought that you had actually removed the IHS in the past. A lot of guys do that to get better cooling. If it was never removed he it should be fine. The fact that you got good cooling before means it should be fine. The Apogee needs to be tightend fairly hard. Sometimes one side will get caught up on the screw thread and get cockeyed. Here is a good thread where a buddy and me overclocked a couple of 165s on that 32 board. It is a remarkable chip. http://asustech.15.forumer.com/index.php?showtopic=2135
March 23, 2007 1:09:11 PM

Cool, I'm glad we got that cleared up. Sorry for the confusing barrage of words. I've registered and I'm reading through the posts now. Great deal of delicious information in there.

Now I can't wait till I get home to test everything out. Thanks Merc, you rock!
March 23, 2007 1:21:17 PM

Quote:

Before that however, I used the infrared temp gun to get a read off the cpu and it wouldn't go over 38C.


I'm curious....
How can you use an infrared gun to take a reading off a CPU that has a waterblock on it? Did you take a reading off the edge on the side?

IIRC the heat is in the middle of the CPU, not the edges, but I might be wrong.
March 23, 2007 1:25:59 PM

Quote:

Before that however, I used the infrared temp gun to get a read off the cpu and it wouldn't go over 38C.


I'm curious....
How can you use an infrared gun to take a reading off a CPU that has a waterblock on it? Did you take a reading off the edge on the side?

IIRC the heat is in the middle of the CPU, not the edges, but I might be wrong.

I did it from the side. It's not a huge gun style reader. It's about as big as my thumb so I used it from the side while aiming at the copper block. I'm planning on attaching a temp sensor to the side of the processor housing and reseating the waterblock when I get the chance. Currently I'm using speedfan to get my temp reading and the information I get off the BIOS. Unfortuntely, I can't get the PC to stay on for more than 15 seconds after it starts shutting down from the heat. I can leave it on for as long as I want but once I start loading it, the temp shoots up and stays there. That's why I'm going to run down the steps I listed above to solve this issue.
March 23, 2007 4:06:51 PM

Yeah if you're getting 38C on the side of the copper block, I can imagine that your CPU's center is much much hotter.
March 23, 2007 4:33:24 PM

Then I guess I'll try reseating the CPU and waterblock one more time. I'm so frustrated. I had a similar problem with the Thermalright 120SI but NEVER with the Arctic Cooling Freezer 64. Something's amiss here and I just don't know what. Why does one thing work and the other not? Maybe it's a matter of tightening the hold down plate screws on the water block? What do you think Stefx? Mer did mention that the Apogee GT needs to be screwed in fairly tight.
March 23, 2007 4:46:37 PM

Hmm. 38c on the side is pretty warm. A watercooled 4800, at stock, should be at or below 32c I'd guess, if not less. Is the block itself warm? I ask because I am wondering if the block is conducting heat. How much AS5 did yuou use? Only a bb sized dot in the middle of the IHS is required. Put thqat little dot in the middle and then screw the block down.
March 23, 2007 5:36:27 PM

Hi Merc. I touched the hold down plate last night while it was doing the shut down dance and it was a *wee* bit warm. It wasn't hot or anything. At the same time, I did feel the heat being pulling out of the radiator by the fan attached to it via shroud. Perhaps I did put a bit much AS5 on. I guess I can try again with a bb's worth. I'm just worried because the IHS has scuff marks on it. What the hell, might as well try it. Thanks Merc!
March 23, 2007 5:39:28 PM

Quote:
Then I guess I'll try reseating the CPU and waterblock one more time. I'm so frustrated. I had a similar problem with the Thermalright 120SI but NEVER with the Arctic Cooling Freezer 64. Something's amiss here and I just don't know what. Why does one thing work and the other not? Maybe it's a matter of tightening the hold down plate screws on the water block? What do you think Stefx? Mer did mention that the Apogee GT needs to be screwed in fairly tight.


If you decide to re-seat, while you're at it, check the flatness of both your waterblock and your CPU. Both should be quite flat, otherwise they'd have a bad contact patch and heat transfer would suffer.

Like said by others above... the problem is one of three:
-Bad heat transfer (thermal paste, bad contact patch, etc) fixable by re-seating and verifying flatness, but you say you have re-seated many time and you are using a good thermal paste
-Low/no water flow - you say you have a good water flow
-Mobo error when reading the CPU temp - unlikely... if you get 38C on the side of the CPU, odds are the the temp at the center is indeed very high.

Kinda running out of possible root causes here...

Some brainstorming:

How do you apply the thermal paste? Good procedure is usually a single rice-grain in the middle, or a thin line. Wrong application of thermal paste could cause air bubbles between the CPU and heat sink (or waterblock) and cause higher temp (although not in the magnitude you're talking about)

tried lowering the CPU multiplier? Might allow you to run a bit cooler and boot up, and it would confirm it's a CPU problem

Are you BIOS settings at stock or at overclock settings? Maybe Vcore and FSB are way too high and causing a higher thermal load
March 23, 2007 5:53:21 PM

Hi Stefx,
I'm going to check the flatness of the CPU and block. I'm sure the block's flat...I do recall checking that. CPU however is another story. I'm going to take that sucker off and check it. I might have used a tad too much grease. Although it was not that much more than a bb's worth and I spread it out using a razor. Maybe that's what I did wrong? Well, it's almost closing time for me at work so I can't wait till I get back to try out the possible fixes.

So...I'm going to reseat the CPU and block. I'm going to test out the block for more air bubbles. I'll check flatness of the the block and CPU. Oh, I have not overclocked anything yet. It's all stock atm.
March 23, 2007 6:05:43 PM

Isn't that GT block convex by design or something?
March 23, 2007 9:15:10 PM

I don't believe so. I just checked :) 

I also reseated the block and cpu with just a tiny bit of AS5. Guess what? It ran great till I hit the 5 minute mark then shot up to 70C and shut down. I'm at my wit's end. For kicks I threw in my Athlon 3500. Guess what? The temp will NOT go over 34C with 100 load and it has been about 15 minutes. I'm guessing that there's something wrong with my X2 4800. For some reason it will only work correctly with an Arctic Cooling 64 Freezer and NOTHING else. What gives? Aaargh!

I'm hoping that it's just the processor that's all messed up somehow and not my general setup. I ordered the Opteron 165 and I'm hoping that I'm receiving the good step/revision. Ah what the heck, I'll update this post once I try that chip out.
March 25, 2007 3:26:59 AM

Let us know the stepping when you get it. If it is CCBWE then RMA it. Anything else is probably good.
March 28, 2007 11:08:24 AM

Okay, I received my Opteron 165. The stepping is LCB9E 0704. From what I've seen, it's not one of the jackpot steppings. It's fairly new as well. So far, I have upped the clock speed to 2.7 at 1.4v. I've had to up the voltage on my OCZ's as well to 1.7v. That's actually what the tech guys over at OCZ recommends for my model but the stock I believe is 1.6v (I'm on my work computer atm). I don't have a screenshot of the setup but if my memory serves me right, the HT is 1300, RAM is at 193Mhz, and the bus is set at 300. I usually have CPU-Z opened when I use my home PC so those numbers should be correct.

I'm still very new to overclocking but this is so far a fun venture. To test out my configuration thus far, I ran CPU burn in and dual instances of Prime95 for a few hours and the temperature at max peaked at 58C with an ambient temperature in the room which varies in certain areas from 19C to 21C. When gaming with let's say Command and Conquer 3, I've hit 40 to 41C average which I don't think is too bad.

I'm glad I have a watercooling setup now. I remember overclocking my X2 4800 to 2.6 once and at max load the temp shot up to around 59C with air cooling. Oh, and I believe the Opteron is also a Toledo like my X2.

Well, there you have it Merc. Just thought I'd give you an update. Let me know of any feedback you can provide. Thanks.
March 28, 2007 9:16:43 PM

2.7GHz at 1.4v is excellent. That is a good chip and I would expect it can do 3.0 easy. The only thing that concerns me is the temps. 58c at that freq and Vcore, on watercooling, is high. My 165 on water at 1.47v 3.0Ghz ran at 43c-46c in a 73f room on Orthos.

58c would be the limit I would except for that CPU. Unless you can get the temps down I wouldn't push it any further. Is it just the CPU in the loop or do you have other things on it?

EDIT: Looked back at your post and your watercooling is way better than what I ran my Opte on. I used 3/8" tubing on a 2x120 rad with the CPU AND both chipsets in the loop. The pump was a MCP350 and I had a straight Apogee on the CPU. With the setup you have the temps should be much better than they are. Something is wrong there.
March 28, 2007 11:34:42 PM

Heya. I'm running Orthos right now and it's peaking at 55C. Just one instance however. I only have a single radiator. It's a 1x120 Black Ice Xtreme. I guess having a second radiator or a double sized one makes a pretty big difference huh? When I put my hand on the intake barb of the radiator, I can feel the warm water flowing in so I guess it's doing its job so far.

On a side note, I noticed reddish brown crud floating around in the loop before I turn on the PC. Looks like oil from the MCP655. I know what algae looks like (brown and green varieties) and that stuff isn't it (thankfully). I'm going to flush all that stuff out and replace the coolant. Right now I'm using FluidXP since it's non-conductive. I'm moving on to distilled water with pentosin and biocide mixed in. It's probably the oil that's circulating in the loop. Merc, you seem much, much more informed than I on this stuff. What do you think? I really appreciate all the feedback.

Edit: Oh, but my idle temp is pretty darn good. Goes down to 28-33C like a sack of bricks.
March 29, 2007 12:53:36 AM

What is the room temperature? The loop can only cool to ambient temp plus a few c. As a comparison, my 4300 at 1.49v 3204MHz is running at 26c Tcase in a 75f room with the fans set to low.

I've read some things about fluidXP turning into goop. On the other hand some guys swear by it. Your temps are to high for that setup. The single fan rad should be plenty to cool that CPU, especially with that pump and the 1/2" tubing. Since you can see crap in the lines I would guess that the CPU block is clogged up and maybe blocking the flow.

Just a guess, but regardless, I would definitely flush the system very thoroughly and get all that foreign matter out of there. You definitely don't want to see anything like that in the loop. Major PITA but make sure is it cleaned throughly before you refill. You may even want to take the rad and block out and rinse them with a vinegar and water mix a few times in the sink and then flush with tap water and maybe a final flush with some distilled water if you want to be anal ( I just use tap water to flush out the V&W mix).
March 29, 2007 1:37:20 AM

Sounds good. I'm planning on flushing out everything. I'm guessing the vinegar flush isn't supposed to be done during an extended period of time? I thought vinegar ate away at copper.

Anyway, the ambient temperature in the room is 70F. The block and rad is probably clogged up. I'm flushing the system on Sunday. You don't think waiting will harm the loop do you? I mean, sure the foreign matter is making the fluid thick and thus reducing the system's ability to remove heat but I don't think it's enough to stop the pump or anything.

I'll be using tap water to flush vinegar/water mix or alcohol (I've heard either would do) and then flush the tap with distilled water. I'm going to be anal because dammit I'm going to make this work! :lol: 
March 29, 2007 1:54:51 AM

I think it eats aluminum
Anyways just put it in there and slosh it around for awhile. The thing should be fine till Sunday just don't let those temps get any higher.
March 29, 2007 2:04:46 AM

Will do Merc. Thanks for everything. I'll do that and see what happens afterwards.
March 30, 2007 11:53:00 AM

Okay, I flushed my system and got rid of the stuff floating around. However, my processor is still giving off a temp reading of low to mid 50's C. This is infuriating. Shall I try using even less arctic silver? I know about applying just a 1 1/2 uncooked short grain white rice's worth or 3/4 of a BB but maybe I'm still applying too much? The radiator only reaches temps of 23C using my infrared temp sensor. The side of the heatsink still reaches up to 35C. Ugh.
March 30, 2007 12:30:49 PM

Quote:
You'll also need to check a few other things.

Some water coolers have a link to the main case power switch, used to shut the system down if it gets too hot or some other alert is triggered. Check that.

wun mentioned checking the block to see if it is properly seated. Check that. Also check thermal paste.

If your CPU temp is over 60ºC or so, you're going to run into issues.

Err, could you also post us your temps (idle/load)?


Baltar is right :lol:  , and you may have an air bubble in your CPU block or radiator. So even though your pump is going and it kind of looks like its circulating...its not. Your motherboard is doing the right thing unless you want your chip fried. Trust it.
March 30, 2007 1:28:34 PM

I've had a similar problem with the Thermalright 128SI where the temp would shoot up to 60C and the computer would shut down. So I do know that my MB would protect itself. I know that the water inside is flowing. While bleeding my system, I use a seperate power supply specifically used in electronics testing. When I flipped the switch to off I saw some airbubbles stop after popping out of the cpu block. When I turn the pump on again, I can't even see the groups of air bubble because the flow's fairly quick.

There's a few things I want to try when I get home tonight (at work atm).

1. Both block and IHS are fairly smooth with no chips, scratches or anything. So I'm going to try using even less than I am now.

2. I have a second radiator just sitting around which I was planning on adding to the loop as soon as I got my hands on front bay mounting brackets. My 1st rad is mounted on top of my case (but inside) with the barbs facing downward. It is connected to a shroud which is in turn connected to the fan and grill. I might have to forego the brackets and mount the rad anyway somehow.

3. I'm going to put my case on its side and try every which angle in order to help move the damn bubbles out if there are any. I'm not hearing any cavitation coming from the block but I'm willing to try anything at this point.

3. I might flush the entire system again when I received additional tubing in the mail. I currently use Primoflex but I don't like the fact that it's not clear and it seems to attract dust on the outside. I'm going with Masterleeker instead.
March 30, 2007 1:43:19 PM

How is your loop set up? I suggest pump to radiator, radiator to cpu, cpu to reservoir, reservoir to pump. You can go rad to res and then res to CPU as well but you want that pump going directly into the rad. What fan do you have on the rad? Is it spinning?

The temps you mentioned. Is it mid 50s under load, at stock, idle, overclocked or what?
March 30, 2007 1:55:22 PM

My loop does this: T/line -> Pump -> Rad -> CPU Block -> T/line

I don't have a reservoir. The fan I have is an Antec Tri-cool 120mm. It was on sale and I run it on med because the high setting is just too friggin loud! At medium the fan spins at 1600rpm but I forgto the CFM. I have it sucking air through the rad out the blowhole. It's obvious that it's operational since all I have to do is look down and left to see it as clear as day.

The idle temp is 33-34C but sometimes jumps to like 42C for a split second for no good reason. The room temp was 19-20.9C. Those numbers are already sad. I helped my brother install a 3.8" Swiftech kit on his FX-55 and it refuses to go over 35C at full load Orthos. At load my temp takes no time to jump up to 49C then hits 55-56C from time to time while running Orthos. All temps I mentioned for my setup is with the Opteron 165 clocked at 2.7Ghz. I lowered the clock speed to around 2.59Ghz but I received the same temps. When I stop loading my CPU, the temps drop like a sack of bricks to 32-34C.

I'm starting to think that I just really SUCK at using the proper amount of AS5. I did have the same problem trying to get the Thermalright 128SI working on my X2 4800. Jeez, am I that big of a goofball and I wasted my money on buying a new CPU when I actually did have a perfectly good processor? If so, why did the Arctic Freezer 64 Pro work well enough for me? Was it more tolerant of my clumsy application of AS5? *sigh* sorry for the rant. I'm just frustrated and can't get this issue out of my mind.
March 30, 2007 2:03:14 PM

You may be puting a tiny bit too much on but it is hard to tell. From what you said you applied it wouldn't be the problem. Don't spread it aroung, just a grain of rice oin the middle and mount the block. heta and pressure will take care of the rest.

Like I said, I don't thing that is the issue. Your rad is using case air to cool itself? That is fine as long as your case is cool. Can you explain your exact setup with the rad and fan. Also ho many case fans are feeding the rad etc. If your case is hot and the fan is sucking the air through the rad then that is your problem.
March 30, 2007 2:33:11 PM

Maybe your system is shutting down due to a fan RPM failure... my board went absolutely nuts because it thought a fan had failed... I had no fan when I went to water cooling... and everything was running fine. Somewhere in my BIOS I had to disable to the fan RPM alert.
March 30, 2007 2:41:25 PM

He'd get beeps at startup if he hadn't disabled CPU fan monitor.
March 30, 2007 2:43:52 PM

Quote:
Maybe your system is shutting down due to a fan RPM failure... my board went absolutely nuts because it thought a fan had failed... I had no fan when I went to water cooling... and everything was running fine. Somewhere in my BIOS I had to disable to the fan RPM alert.


My PC stopped shutting down because the temps aren't high enough to cause it now. However, I'm still working on trying to get my temps down.

Merc-
I have 2 intake fans. One on the bottome blowing through my hard drive cage which houses 3 hard drives which are mounted vertically to let more air through. I have the other intake on the side panel blowing directly over CPU socket area, voltage regulators, and video card. It's a 250mm fan that I modded onto the case. I have an exhaust mounted directly over the MB headers and the second exhaust is the one mounted to my rad. The case does have some hot spot such as the RAM sticks and my 8800GTX but there's plenty of air blowing around in there and ambient temps aren't that bad. Regardless of whether I have my panel on or off or a little off doesn't affect my temps at all. I also have an 80mm fan sucking air out of the power supply. You have a PCP@C PSU so you know the deal.

I have a Scythe Kama Bay just sitting around and I was thinking of setting a duct from the top 3 5.25 in bays to the rad so that only outside air is being sucked in and blown out the top but I can't find a fan duct for 120mm fan anywhere.
March 30, 2007 3:08:00 PM

That should be fine. I am stuck on this one. Maybe try having the fan blow through the rad. Your temps remind me of when my pump stops running (have a loose connection on a molex I need to fix) but you say the pump is running and there is good flow. Even with the fan off your temps shouldn't be that high with good flow so I don't think the configuration has anything to do with it.

On my Opte 165 I originally ran both chipsets and the CPU through a single fan rad and it was only about 3c-4c higher than a dual fan model. The reason is there just wasn't enough heat for the dual fan to make much of a difference. You have just the Opte 165 with less voltage and at a much lower freq on a much more robust watercooling system yet your temps are way higher than I ever saw in a much warmer room. 82f and I may have hit 53c at 3.0GHz 1.49v running Orthos.

There is something very wrong and I can't figure it out. I am thinking that the system was not bled well enough with just that T-Line in there. I always use a res just because it is so much easier to bleed the system of air. A T-Line bleed takes hours to do, from what I have read, and I fail to see the advantage of not using a res.

Call Swiftech and ask to speak to a tech. They are very, very good about customer support.
March 30, 2007 3:17:11 PM

I guess I'll start thinking about a reservoir now. I'm going to try a few things tonight before I start buying more parts. I might try out the other radiator. I have enough coolant to try that out. The other rad is a BIX GT I believe. It's not as think as the Xtreme so it should allow for better airflow as well. Regardless, I'm going to reseat the heatsink, and make sure the CPU block is free of all airbubbles.

I didn't buy a kit from Swiftech. I picked out my parts one by one. My brother purchased the kit. Of course he's very happy with it. :roll:
March 30, 2007 4:22:49 PM

I really don't think a bigger rad is going to have any effect. Maybe 1c as the loop couldn't be anywhere close to its limits with just a CPU in it. An air bubble seems likely and without a res you are just going to have to keep moving the case around and upside down etc. to get the bubbles out of there. It may take awhile buut I really can't think of anything else that would help. A bigger rad would help only because you dislodged an air bubble when you drained and refilled. If you reread my post above I went from a single to a dual and it dropped the temps about 2c. You're looking at a 15c drop. Can I ask who convinced you that going without a reservoir would be of enough benefit to offset the major hassle of bleeding on a T-Line?
March 30, 2007 6:26:45 PM

Yeah....that would be me. I thought what the heck. I don't want a reservoir. I'm just going to have a t-line and I'll be fine! I'll see if there is a bubble in there.

I just want the day to be over so I go home and work on this problem. Gah! Why do I always have to do something different?
March 30, 2007 6:36:40 PM

HA!, I can so relate, I first tried a modified funnel reservoir attachment on a t-line on my system and the air would simply not escape out of the t-line but get lodged somewhere else and the flow was restricted. I think a secondary release t somewhere on the other end of the flow would have fixed it but I got tired of fooling wit it. I just bought me a nice reservoir from Danger Den and problem solved.

And YEA, :x its cold, rainy and Friday and I just want to get the heck out of this place and go play some stalker or something on my system.
March 30, 2007 7:08:55 PM

The problem is, I can't hear the water cavitating inside the CPU block which leads me to believe that it's clear of bubbles. At the same time, I can definitely feel the coolant flowing when I gently squeaze the outlet tube of the CPU block. I'm so frustrated right now.

I did place an order on the Swiftech micro-res from Petras. I hope my rig works out like yours did warezme. It's time for me to punch out. I'm going straight home to continue troubleshooting my severely disabled cooling system. :|

PS- How's Stalker anyway? I'm enjoying C&C 3 myself. I just need my pc to stay cool :cry: 
March 31, 2007 1:18:54 AM

Ok. I did some stuff and looks like I'm getting decent temps. Chipset is 30C. Room is 20C.

My processor is showing 46C peak with Orthos and Folding running together. My one issue now is the fact that my 1 core is running at 46C while my 2nd core is getting much higher temps. It's at 50-52C. What is up with that? I made sure the TIM spread out even and made sure there were no bubbles in there. I bled my loop dry of bubbles (took hours!). Jeez, I can't win! Any suggestions?
March 31, 2007 2:27:35 AM

What is the CPU running at speed and voltage-wise?
March 31, 2007 3:13:51 AM

I would definitely try a resevoir. It may help dissipate some more of the heat. What I would also try is reversing your 250mm fan. Have it suck air out of the case. Your two intake fans should be sufficient enough to provide cool air, but you need more fans to suck out all that hot air. That should lower your case temps and allow more heat to be sucked away. Another thing you can try is removing the CPU and polishing the top of the CPU with brasso or some other metal polishing compound. That will allow for more surface area to be incontact between the cpu and waterblock. It will also help with the thermal paste.
!