old thermal paste?

I've been racking my brain trying to figure out how to get better temps with my newly built wc.

45º is unacceptable if I can achieve the same temps with air.

I tore the system down and by looking at the block and cpu's thermal paste, it seems like it was making good contact. There was excess squeezed out around the sides. The thermal paste I used (Artic Alumina) is at least 2 years old, does thermal paste have a shelf life? I've used it numerous times over the past 2 years (syringe type dispenser) but none the less it is aged.

Does thermal paste "go bad" after so long?

Also, I was using a gasket type shim, do ppl often use shims on wc rigs?

TIA

you would get better temps with better thermal paste, maybe 2-3C, but I'd be more worried about using the shim

Thermal paste is all about application. Is your cpu using a IHS? go to the arctic silver website and follow their directions. The less thermal paste between cpu and HS the better so thermal paste squezed out to the sides is good, you want very little on the die. Also if there is alot squeezed out that means you probably put too much on and less would be better next time.

to sum up.....the thermal paste you use will not make a difference, its how you use it.

the temp diff between pastes will be less than you can measure, the difference in the quality of the application can easily be several C.

The shim; cheap insurance, keep it, no effects on temps there.

shims do nothing, but insulate the cpu.

Check the mounting with the shim on. If its not perfect, then the shim could be causing poor contact with the heatsink.

dicecca112 said:

shims do nothing, but insulate the cpu.

Click to expand...

I beg to differ. it does help prevent chipping corners and what not and doesnt stand in the way of the thermal interface to the HSF/WB.

sweetness thx for the reponses ppl

I've went ahead and reseated the wb, and used the same thermal paste again. I def did not have too much (there wasnt TOO MUCH squeezing out from the sides, just a small amount). I tried it without the shim this time, and I'm averaging the same temps. So something isn't right, but I dont think its the contact between the die and wb that is the issue.

My friend who was using this waterblock before he gave it to me was seeing temps around lower 30's. thats 10-15º cooler than im getting. He was using the same model pump, exact same rad/fan, and he was also cooling a 9700pro on the same loop. This is insane that I'm getting 45ish, especially considering im not even cooling a gpu.

I have decent ambient temps in my room, so that def isnt the prob (case temp in mbm reading 28º atm)

Anything at all that you kind folks can think of would be wonderful for me to hear.

Don't always trust temperature sensors, they aren't the most accurate things in the worse. What kind of airflow do you have in your case? Because a lot time if the socket temperature sensor needs good airflow to get semi-accurate temps. Try pointing a fan @ your cpu socket, and if you notice the temps dropping, then something isn't wrong more than likely.

citronym said:

sweetness thx for the reponses ppl

I've went ahead and reseated the wb, and used the same thermal paste again. I def did not have too much (there wasnt TOO MUCH squeezing out from the sides, just a small amount). I tried it without the shim this time, and I'm averaging the same temps. So something isn't right, but I dont think its the contact between the die and wb that is the issue.

My friend who was using this waterblock before he gave it to me was seeing temps around lower 30's. thats 10-15º cooler than im getting. He was using the same model pump, exact same rad/fan, and he was also cooling a 9700pro on the same loop. This is insane that I'm getting 45ish, especially considering im not even cooling a gpu.

I have decent ambient temps in my room, so that def isnt the prob (case temp in mbm reading 28º atm)

Anything at all that you kind folks can think of would be wonderful for me to hear.

Click to expand...

Ya listen to lowfat... his sensor could be 7C low and yours could be 8C high making it seem to be 15C higher when in all reality you have the exact same temp!

Ditto on the fan, if the mobo is not reading the temp straight from the cpu its gonna be affected alot by air flow.

Go by how your OC is affected by WCing, not the temps. Also 45 is not bad, depending on your setup....is that OCed or not? If yes, then I would say let the thermal paste set and make some minor improvements, if no, then continue troubleshooting.

Is this running stock ? If not what is the cpu and what do you have the Vcore set to ? fsb ? all these have a bearing on your temps

whew, ok

heres the rundown.

I'm using an A7V333, with an XP2400+. I have the cpu running at 167*14>[email protected].

afa the temp indications go (mbm5), the CPU Diode is reading 42-44 idle and 45-48 loaded (depending on time of day). The CPU Socket is reading 36-41, again, depending on time of day (ITS GETTING HOT IN FL!!!). If I point a fan at the socket, the CPU Socket temp does go down, but it does not affect the CPU Diode reading.

Before I got this WC system going, I was running an Alpha PAL 8045 with a TT SmartFan II running full throttle, as described in this post http://www.ocforums.com/showthread.php?t=388321 . While I was using this HSF, The best speed that was stable was 163*14>~2280, still at 1.8v. I was getting temps around 46 idle, 49 load.

The radiator (blackice extreme 1) is mounted on the back of the case, wich is pointed at a window banger a/c less than 2ft away. I have a Sunon 108cfm fan on the inside of the case, sucking in nice cool air from outside of the case, thru the rad, coming from the a/c. Im using a Hagen Powerhead 802, supposedly rated at 400gph in a prescribed application. The pump has appx a 5/8 ouput, and I have 1/2 tubing stretched over the outlet. after ~4 inches i have a 1/2ID>3/8ID adapter in the line, and then it goes to the water block as 3/8, as the wb has 3/8 fittings. Then there is 3/8 tubing out of the wb, into the rad, which has 1/2OD fittings. The 3/8 is stretched over the fittings on the in and the out, and its a 3/8 hose going back the the res. There def isnt any kinks or 90's, all the hoses are nicely set. Im running 90/10 distilled/af, with a small amount of water wetter and ~12 drops of Hi-Liter ink.

I would have expected at least 10-15º, but I'm rarely seeing a 5º drop. Even if I'm not getting uberflow, I have to be getting sufficient flow, because when I feel the wb, its not even warm to the touch. When I feel the air coming off the rad fan, its nice and cool. Nothing feels warm to the touch. If I turn the rad fan off, it goes up to 49 idle, but it takes a few. At that poing everything is noticably warm to the touch. I'm certainly not a WC expert, but if I didnt know any better it would seem like every module is functioning properly.

Ok, so consider the temp readings I'm getting from MBM (and probe BTW) are faulty, and my cpu isnt even running at 45, its in the 30's. Theres problems tho. That would put my previous HSF at upper 30's lower 40's, which could be possible, but I would still expect more difference than less than 5-10º when I upgraded to water.

thx for the help

Well it comes down to splitting hairs on the numbers here so lets add up the facts...

from what I see you are seeing a 4c drop 49-45c

take into account it is getting hotter, thats worth prolly 2c, im in FL too so I know what you mean

take into account the extra couple mhz, not much but adds up.

Thats a pretty good drop to me. Keep in mind that its probably 26-27c in the room, case temps are a bit higher, 29-30, that puts you about 15c above ambient according to the temp probe. that means that 5c you lost was 25% of your dT on air, sounds like a pretty nice drop to me. The 15c drop you expected would have put you at 34c, 5-6c over case temps, we would all like that but its just not gonna happen.

thank you ls7, I wanted to hear that!

What would you suppose cause the difference in temps from my friends computer to mine? The only differences were he was using a larger reservoir (not huge) and he was also cooling a 9700pro gpu. From what I understand the 9700 is a scorcher. Correct me if I'm wrong, but a larger res only makes the temp climb to max take slightly longer, no? I seriously doubt his ambient is much more than mine.

I never saw his temps of lower 30's with my own eyes, is he just completely wacked? If u say the temp drop I've achieved isnt actually that bad, then that would put his 30º temp at an amazing level, wouldnt it?

If you or anyone could agree with that previous statement it would put me at ease.

thanks a mill

Res benefit depends on materials of construction. I have a metal (aluminum) res situated in line with one of my 120 fans. Makes for a nice air cooled heatsink.

citronym said:

thank you ls7, I wanted to hear that!

What would you suppose cause the difference in temps from my friends computer to mine? The only differences were he was using a larger reservoir (not huge) and he was also cooling a 9700pro gpu. From what I understand the 9700 is a scorcher. Correct me if I'm wrong, but a larger res only makes the temp climb to max take slightly longer, no? I seriously doubt his ambient is much more than mine.

I never saw his temps of lower 30's with my own eyes, is he just completely wacked? If u say the temp drop I've achieved isnt actually that bad, then that would put his 30º temp at an amazing level, wouldnt it?

If you or anyone could agree with that previous statement it would put me at ease.

thanks a mill

Click to expand...

yea either he is not reading at load or the sensor is off. Its not uncommon for a sensor to be off by 10c so I wouldnt paint him a liar.

alright, thats cool

thx ls7

I'm about to turn it off and clean up the clutter and take some nice pics. watch for a new thread