ok, so this whole time i have been running my custome WC, i have been using the Bios temp from my DFI LP mobo. under full load it will go up to about 35C

now tonight i just installed a aero cool gatewatcher and attached the temp probe to the side of my RBX block. the temp readout says 11.5 under full load....


which should i believe???

11.5 is impossible unless your running a p1 or your computer is in a fridge , beleive the other 35c even the seems abit low.. if its at full

i believe that the block temp is prolly 11 but you cpu temp is 35 the deltaT between cpu temp and block temp and be quite high . btw whats your ambient?

i doint know about being impossible as i live in Minnesota and its ****ign cold here. but it does seem very low to me. but my old WC kit did only 29ish to 33ish and that was a cheesy all in 1 kit.


RBX block with black ice xtreme rad and a L40 water pump.

i would say the room temp is a cozy 65ish degrees.

12C is only 53 Fdegrees, but ya it does seem low... also a AMD 64 2800+ slightly OCed

I'd think that the temperature probe you got is defective.

i dont think it is defective, but what i think is the RBX is kinda thick. so teh otterwall where the probe is attached is prolly cooler

what jsu said is right.

now with the RBX, it doesnt cover teh CPU fully. would there be any danger to my CPU if i attached the probe directly to the CPU ?

also to add the L40 does have a high head and GPH 740 gph with about 10 fh. i would say im only using up about 6 feet of head.

Go with the bios reading. Because of the IHS (Integrated Heat Spreader) on the 64's, the only way to get an accurate reading with a temp probe is to place it top-dead-center on the IHS (this requires routing a trench/drilling a hole in the HS...i know d!ck about water cooling....don't know if there's a work-around with a water block or not). WARNING! THIS IS POTENTIALLY DESTRUCTIVE TO THE CPU: The other alternative (if you're brave enough) is to try to remove The IHS and touch the side of the core with the thermal probe, and notch (for the probe wire) and replace the IHS: info here - http://forums.pcapex.com/showthread....+thermal+probe

i guessi forgot to mention this earlier.. the DFI boards and i should say most socket 754 boards have problems with the buios reading temps worng. on my old WC kit, the BIOS temp was about 10C off from what the temp probe said. so i cant just go by what the BIOS says. as it is about 85% wrong. its prolly wrong by only 5 to 10C, but still wrong.

sorry, not going to possibly ruin a $70 water block to attach a probe to it. plus you can really to too deep on some parts

i agree (had the same prob)....a bios flash gave me temps that are much more believable

Quote: Originally Posted by droeblek9 i agree (had the same prob)....a bios flash gave me temps that are much more believable

ya i flashed the bios as well, so the temps are more believeable as what they were at first. think it was close to 65C before the flash.

i may attach it directly to the CPU to see how that works. im just hoping ther eis no electrical discharge or anything that, that will feed back to teh CPU.

Quote: Originally Posted by crazymod i dont think it is defective, but what i think is the RBX is kinda thick. so teh otterwall where the probe is attached is prolly cooler

??????????
I donÂ’t understand. Do you think the temp probe is lying because the RBX is thick??

That would be absurd!

Quote: Originally Posted by crazymod ya i flashed the bios as well, so the temps are more believeable as what they were at first. think it was close to 65C before the flash. i may attach it directly to the CPU to see how that works. im just hoping ther eis no electrical discharge or anything that, that will feed back to teh CPU.

Here's the rub....in order to get the probe to the core, you have to remove the IHS (integrated heat spreader - the metal plate that covers the top-side of the cpu). If you try that, you very well might completely destroy the cpu.
I didn't have the b4lls to try it (or the $$ to buy another cpu if i scr3wed mine up). I seriously considered drilling a hole through my HS, but i was afraid the bit would hit one of the heatpipes on my HS (XP 120).

So, i just said "the h3ll with it". I currently monitor my case temps (air intake and ambient), etc. with a thermal monitor. That at least gives me a relative idea/comparison of how accurate the onboard monitoring is.

Another word of warning (which i'm sure you already know -not trying to be insulting....just thought it worth mentioning): Never ever, ever, place anything (other than thermal paste) between a hs and the cpu/IHS...it will keep the HS from making contact with the cpu/IHS = one very hot and possibly trashed cpu

no, i think the probe cant get a good reading what the actual CPU temp is because the sidewalls are too thick. that is why i am wondering if i should attach it directly to the CPU

hehe, ok with the RBX block, it doesnt cover the CPU fully. there is about 1/8 inch along the sides of the top of the CPU exposed. so i could attach teh probe directly to that without drilling or anything. go take a look at a dangerden RBX block, you will understand then.

you would have to watch the installation video as the pictures doesnt really show the CPU exposed on the sides,

Quote: Originally Posted by crazymod no, i think the probe cant get a good reading what the actual CPU temp is because the sidewalls are too thick. that is why i am wondering if i should attach it directly to the CPU hehe, ok with the RBX block, it doesnt cover the CPU fully. there is about 1/8 inch along the sides of the top of the CPU exposed. so i could attach teh probe directly to that without drilling or anything. go take a look at a dangerden RBX block, you will understand then. you would have to watch the installation video as the pictures doesnt really show the CPU exposed on the sides,

Still won't be accurate. You have to place the probe directly on the center of the IHS (that will put it directly over the core...that's where the highest temp will be).

ok im not understanding the logic here.


where the probe is at now, only reads 12C and thats on teh side of the block. now if i placee teh probe directly ON the CPU.

also to add. if i put it on the CPU and the temps are close to w that is exposed, i may not get 100% accurate reading, but a close one. also, you CAN NOT place a probe directly over the CPU with the RBX block

also to add, if i put the probe on the CPU and the temps are reading close to what it reads now, i would say that would be accurate. as the heat will still spread to teh sides of the CPU exposed and be able to give a reading close to what the core is

Directly over the core is the only place to get an accurate reading. If you place the the probe on the edge of the IHS, it'll still give you a temp reading that's cooler than what the core is actually running at (so you're back to the beginning: you still won't be getting accurate info on how hot your cpu is running ). I have no idea how much of a difference in temp it'll give, but it'll be cooler.

i understand that, but it is not possible without ruining the block. but wouldnt teh edge of the CPU be better than 12c from teh side of the block??


basicly, i guess all i really need to know is there any danger to frying my CPU by connecting the probe to it? electrical charge to teh CPU or anythign?

Quote: Originally Posted by crazymod i understand that, but it is not possible without ruining the block. but wouldnt teh edge of the CPU be better than 12c from teh side of the block?? basicly, i guess all i really need to know is there any danger to frying my CPU by connecting the probe to it? electrical charge to teh CPU or anythign?

Edge better...i imagine so, but i'm not sure (haven't tried it).
Dangerous? Not that i've ever heard/read (as long as the probe isn't placed between the HS/block and the IHS).


Here are a couple of pics taken from the AMD pdf that's in the thread that i linked to in one of the above posts (#9). This is what AMD recommends for manufacturers who need to use a probe to monitor CPU temps on Processors that have an IHS (first pic is the correct placement of the probe on the IHS, second pic is the groove and hole machined into the base of a HS - the hole is to allow clearance for a drop of epoxy that is attached to the end of the probe wire)

Attached Thumbnails