• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

What's your WC fluid temp?

Dew itt right

2[H]4U
Joined
Oct 28, 2005
Messages
3,314
Alright, I'm sick of having a tight setup and getting idle CPU temps upwards of 38C! I've replaced the radiator with one twice as expensive, I've switched from T-line to res to make sure there where no air pockets, and now I just ordered a $37 bottle of coolant to see if it'll help (I wanted new coolant anyway). I know the flow isn't being blocked either because my pump is only running at 40% and the fluid in my reservoir is going crazy! Definitly plenty of flow. Removing the side panel from the case only drops the temp about 2C.

Well one guy in my last thread HERE said he thought my Abit board may be giving be bogus temps. So to check that, I figured I'd measure the temp of the coolant in my reservoir to see if it's really as warm as the board says it should be. But I need something to compare it to - to see how fluid temps typically relate to component temps. That's where you come in! :D

PLEASE post your fluid temp along with corresponding CPU temp (and chipset & GPU if you'd like).

IDLE:
CPU: 38C
GPU: 38C
Chipset: 37C
Estimated room temp: 25C (77F, Holy crap it's warm in here!! :eek: )
Corresponding fluid temp: 34.6C

LOAD (1hr of Prime):
CPU: 48C
GPU: 42C
Chipset: 47C
Estimated room temp: 25C (77F, Holy crap it's warm in here!! :eek: )
Corresponding fluid temp: 38.0C
 
I have an Abit AV8 3rd Eye. Yes, the temps are screwed up on these boards. Bios 24, I think it was, fixed a lot of the temp issues, but they are still higher than actuality. My temps with water cooling are 37-38 Idle then 56 under load. Yet, my gfx temps are 31 idle and 38 load. I'm sure that my Abit is falsely elevated; abit is notirious for poor temp monitoring.
 
FLUID temps are the most important part of this. Can you guys get your fluid temps (w/ a temp probe or thermometer) that correspond to those readings?

My temps have been updated! Load temps to come in an hour or two!
 
I just checked my inline temp sensors. I have four sensors, Res is 28C, CPU 1 29C, CPU 2 30C, RAD 27C, GPU 2 29C. I just returned home and those are idle temps for about 10 hours.
BTW, I have 2 RADS.
 
theseeker said:
I just checked my inline temp sensors. I have four sensors, Res is 28C, CPU 1 29C, CPU 2 30C, RAD 27C, GPU 2 29C. I just returned home and those are idle temps for about 10 hours.
BTW, I have 2 RADS.

Sorry I'm a bit confused. :confused: Are the CPU/GPU temps of the fluid at those components or of the components themselves? I'm assuming it's all fluid temps because of the "inline sensors". If so, do you have component temps that correspond to these fluid temps?
 
Know what I think? i think you have a 120mm radiator trying to dissipate the heat from a processor, chipset, and a vid card, thus causing your temps to be high.

Your best bet would be either a second rad or a dual radiator

----pump---------------rad------------------CPU-----------------chipset------------GPU-------------------------------Reservoir-------------------------->
--MCP655-------B.I.Xtreme-------MCW6000K-------DD Maze4--------MCW55-------------Innovatek Tank-O-Matic Micro---------->
 
If your just using a 120mm radiator I am under the suspicion that is your problem. With a small radiator and a high heat load you need a large delta T between the coolant temperature and the ambient temperature in order to dissipate the amount of heat your loop is producing.

You have a hot CPU, a relatively hot GPU, a hot pump, and a chipset all in the same loop trying to be cooled by a single 120mm radiator.

Also you say you upgraded to a radiator which was twice as expensive...so did you upgrade from the BIP to the BIX? Depending upon the fan you are running that may have done more harm than good. The added depth of the BIX while providing marginally better cooling capacity also greatly increases the back pressure. This means that with weaker fans the BIP will perform better than the BIX.
 
Dew itt right said:
Sorry I'm a bit confused. :confused: Are the CPU/GPU temps of the fluid at those components or of the components themselves? I'm assuming it's all fluid temps because of the "inline sensors". If so, do you have component temps that correspond to these fluid temps?

I only trust fluid temps. The inline sensors are on the out-let side as I want to see the highest temps. IMHO you cannot trust temp sensors on "the chip" as everyone mounts them differently and temps from any MB monitor are basically worthless.
 
It's true about the radiator being a relatively small source of heat dissipation, however, if you read the thread I linked in my original post, my temps were in the mid-20's when I first put this system together. Exact same WC setup except w/ a T-line and my previous rad (swiftech MCR120). I ran them in the same configuration over a DFI LanParty Ultra-D, A64 3500, and the same vid card. I fried my mobo and replaced it with this one and switched to a non-conductive coolant around the same time. :p Since the temps have been high, I've tried reseating my blocks, flushing my coolant a couple times, & upgrading rad's. I've even removed my HDD from the case to help cool the ambient temps a bit! Now it boots my PC from the floor behind the desk via 6-foot long SATA and power cables! :D

My fan should be able to handle the extra back pressure. My system has three 120mm Scythe S-Flex's I just put in last month. (1) intakes through the front (thru the rad) , (1) exhausts thru the back, and (1) in my PSU exhausting. I've also sealed off all the other air wholes in the front of the case to make sure as much air is coming thru the rad as possible. Besides, with that deeper rad, I have a considerably greater surface area for convection.

I just don't think that's the problem. The coolant seems to be getting cooled properly enough. My posted temps are from the res just before the pump and rad so it should be the higher end of the loop. When my components go up 10C during load, my coolant only went up 3.4C! It seems to me that (a) the coolant isn't doing it's job pulling enough heat from the components or (b) the temp readings are off.

The only reason I doubt the temps being off is that my GPU is right up there with them and it has it's own sensor apart from the mobo obviously.

I just ordered some coolant yesterday. When it comes in, I'll see if I can't squeeze my old rad back in the case (perhaps on the rear exhaust fan) to see if the two rads help at all.

**hey theseeker, I still would appriciate the component temps if you have then available. I understand what you're saying about the onboard sensors, but humor me ;)
 
with a Scythe S-Flex (relatively low cfm fan) the Swiftech QP-120 is most likely going to outperform the Black Ice Xtreme. So if you cant fit it in the case as a second radiator at least use it instead of the BIX.
 
I just bought a new rad I will try that tonite, I hope it lowers my temps. In another post the guys suggested I go. pump>rad>cpu>rad>gpu>res>pump
 
catalystsupreme said:
I just bought a new rad I will try that tonite, I hope it lowers my temps. In another post the guys suggested I go. pump>rad>cpu>rad>gpu>res>pump

Actually what was suggested in that thread which was the best suggestion was to pipe it whichever way provides for the cleanest, simplest, shortest tubing runs. Put the components in your system and figure out which way will use the least amount of tubing without kinking. The only thing you want is to have the resevoir feed the pump (res > pump) aside from that, anything goes pretty much.
 
Dew, what type of CPU block do you have? Hopefully there's a good review on it with DeltaTs, so you can see if the range of DeltaT's they got in the review matches with your DeltaT (CPU Core temp - coolant temp).
 
Erasmus354 said:
with a Scythe S-Flex (relatively low cfm fan) the Swiftech QP-120 is most likely going to outperform the Black Ice Xtreme. So if you cant fit it in the case as a second radiator at least use it instead of the BIX.

This is true, however this is the only clear opening (when I block the side window fan hole) through which (3) 120mm fans are pulling air. If you hold you hand outside the opening in the front of the case you can feel some serious air moving when the fans are at full speed. Besides, even when the rad fan is on full blast and the side panel is removed ( so the ambient temps are basically equal to the room temp) it only drops my CPU temp by 1-2 degrees.

I should hopefully be able to fit the old rad on the rear exhaust fan if I mount it upsidedown, the top of the fan may be too close to the PSU to even do that! I'll give it a shot as soon as the new coolant comes in. The setup will be res>pump>rad>CPU>rad>chipset>GPU>res.

Erasmus354 said:
Actually what was suggested in that thread which was the best suggestion was to pipe it whichever way provides for the cleanest, simplest, shortest tubing runs. Put the components in your system and figure out which way will use the least amount of tubing without kinking.

I'm assuming this is to help maximize your flow? I know this wasn't directed toward me but my pump is only running at 40%, if I put it above 60% the flow in the res becomes so turbulent that it can begin to throw air bubbles into the loop! So in my case I'd assume that the more tubing the better (ie. more heat loss through the tubing). Not that anybody want a spiderweb of tubing in their case anyway! :eek: Irrelevant point I guess, but great advice for one who has a smaller pump. When I had a T-line I had the flow at 100% for a while and the temps only dropped 1C compared to pumping at 40%.

penguin said:
Dew, what type of CPU block do you have? Hopefully there's a good review on it with DeltaTs, so you can see if the range of DeltaT's they got in the review matches with your DeltaT (CPU Core temp - coolant temp).

All my components are in my sig in the original post. The delta-T varies from 3.4C to 10C between load and idle temps. I'm sure it's not the CPU block (or atleast not ONLY the CPU block) because the chipset and GPU temps are right up there with it.

BTW, I still don't have anybody's Fluid & component temps!
 
Spring is here, so the temps are starting to rise. Which means comp temps start to rise, also. In the mid-70's (F) outside right now. We were in the low 80's yesterday, and the following idle temps were about the same. And it's humid, also. In cooler weather, the idle temps are significantly lower.

Here are my comp temps:

Ambient case temp: 30.6°C (Crystalfontz temp probe)
Northbridge temp: 41.5°C (Crystalfontz temp probe)
HD1 temp: 31.5°C (Crystalfontz temp probe)
HD2 temp: 30.3°C (Crystalfontz temp probe)
CPU temp: 35.1°C (Crystalfontz temp probe)
GPU temp: 35.1°C (Crystalfontz temp probe)
Radiator inlet water temp: 33.2°C (Crystalfontz temp probe)
Radiator outlet water temp: 32.9°C (Crystalfontz temp probe)

MB temp reported from Speedfan: 35°C
CPU temp reported from Speedfan: 38°C
GPU temp reported from ATI Tray Tools: 37°C
Video Card PCB temp reported from ATI Tray Tools: 41°C

CPU: Intel 3 GHz Northwood @ 3.45 GHz.
GPU: ATI X850Pro @ 580/1180

Is this what you are looking for?

Jim
 
CPU 36
Fluid Temp (IR Thermometer on tubing) 28.8C (about 1 deg F cooler before CPU and vid card than after and 2 deg F cooler in T-line)
 
Since the arrangement of the loop doesn’t matter, at least with an extra 120mm rad there should be a drop in temps right? I would assume
 
theseeker said:
I just checked my inline temp sensors. I have four sensors, Res is 28C, CPU 1 29C, CPU 2 30C, RAD 27C, GPU 2 29C. I just returned home and those are idle temps for about 10 hours.
BTW, I have 2 RADS.

OK, since you have been so persistent.
Component temps are as follows: TM=TYAN monitor, CS=Component Sensor, NS nvidia

CPU 1 TM=31 CS=28
CPU 2 TM=33 CS=29
RAD 1 CS=29
RAD 2 CS=26
GPU 1 NS=Doesnt register CS=25
GPU 2 NS=DR CS=26
The inline temp sensors are still the same and I have not used the computer since the last message.
 
Back
Top