• 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.

Prescott temps under Water Cooling

B

BLiTzKRiEG

Limp Gawd
Joined
Apr 22, 2005
Messages
295
Ok... I have an older swiftech kit that I bought from someone and installed it on my second rig:

MCP600 Pump
MCW5000 P4 waterblock
MCW50 VGA Block (not on the card yet)

I coupled these 3 with my Danger Den black Ice Pro Micro II (dual 80mm radiator) and a dual 3.5" bay reservoir.

I used 3/8" ID Primoflex tubing overtop of 1/2" barbs ont he blocks...

Heres what I know thus far:

-Abit 478 boards tend to display temps 8 to 10*C hotter than the actual temperature
-Prescotts run unbearably hot.
-I am running an older Swiftech kit and shouldn't expect top of the line cooling temps


So judging from what I stated above... I am currently idling at about 39*C, and loading at about 50 to 53*C (as in my abit utility is reporting them at 49*C and 60 to 63*C)... This is with the processor overclocked from 3.0GHz to 3.6GHz.

I am in a little warmer than normal ambient room right now, its about 75*F in here, but that shouldn't attribute much...

I was just wondering if these temperatures seemed reasonable? I haven't toiled in the pentium world and overclocking before, and not with Water Cooling...
 
Sounds a bit high....

On my other box I have a p4 2.8ghz northwood running at 3.4ghz on a Thermaltake Bigwater SE. Idles around 33 degrees c, load around 40 degrees c.
 
yes, but northwoods run significantly cooler than prescotts...

I just got off the phone with a buddy of mine...

He asked me to start touching around the waterblock... asking if it was hot to the touch...

The waterblock itself is not hot at all... feels like room temperature metal would feel...


I decided to do the max heat and max power test in prime95 like my buddy said...

He says to let it run and touch the waterblock every 30 minutes or so... If it doesn't feel hot, it would be a messed up temperature probe.

So I start prime95 like he says... and am watching my temperatures... obviously they are hotter than normal... now reporting up to 75*C (so like 65*C)... but it also says the PWM area is hitting 75 to 80*C... But the thing is, I can reach in and touch the caps in the PWM area... yes its warm, but not 75 to 80*C hot (that would be enough to burn the fingers)...

I'm kinda just wondering if maybe the motherboard is what is taking the dive, rather than it being a processor issue...
 
I have an Abit with a 3.0c Northwood clocked at 3.5 Ghz. It idles at 45C accorfding to Abit (I assume it is more like 35C). It loads at 55-56C (Again, I assume it is more like 45-46C). I have a had a few instances where the temp went sky high, like a thermal runaway, but I felt the heatsink and it felt almost room temp. I think the temp sensors on these boards are crap, so I pretty much ignore the readouts, and use them only as a reference.

Mine is on air, and I have read umpteen threads about Abit temperature monitoring problems. Cruise around at Abit's forums if you want to see all the threads posted about it.
 
Thanks for the heads up ewb302 (btw, hows the raptor running? ;) )

I decided to test a couple of things...

I dropped to stock clocks and monitored the temps... I got 45*C idle and 52*C load in prime95 at stock settings...

So My guess is that yes, the Abit temp probes are way off target...

I am now testing the machine in the prime95 max heat/power consumption at 3.3GHz (220FSB) at stock vCore (which is 1.3875 in the bios, but the Abit utility is reporting it at 1.30V)... It has the same idle and load temps as stock timings (according to abitEQ: 45*C and 53*C load)...

My only cause for concern is the PWM area... the abit probe is reporting it from 55*C to 65*C still... I am slowly going to work my way back up to 3.6GHz while being at least 8 hours prime stable...

I am also thinking about adding a fan over the PWM area to push air across those mosfets and caps in order to lower that temperature...


I also need to realize that the thermal paste needs about 50 hours of burn-in in order to settle and cool at its maximum efficiency...

Thanks to everyone for their input... I'll post here later tonight and tomorrow to see where it stands.
 
The raptor is great, I appreciate it :)

I have heatsinks on as many mosfets as I could. I couldn't get heatsinks on every mosfet because things are in the way, but I get PWM temps of 38-40C if I remember correctly.

Have fun, I don't think you have much to worry about, except that Abit reports temps higher than anyone else. There are also some websites out there that show thermal probes being attached to the heatsinks on Abit boards, and the temps were verified lower than what was being reported by the motherboard. It's all relative.
 
Well... I think I figured out the problem...

The PWM area was getting so hot that the efficiency of the coils and mosfets was decreasing to the point where I needed to increase the voltage in order to obtain a stable overclock...

The higher voltages lead to the outragous temperatures...

Last night and this morning saw at least 8 hours of prime 95 stability at 230 and 235 FSB (3.45GHz and 3.52GHz)...

Right now Prime95 is blazing away at 240FSB (3.6GHz) on stock vCore (1.3875V) under the long FFTs test.

My idle temps were 45*C by the abit utility (so 35*C) and loads at 53*C by the utility (43*C).

Heres the kicker... The 60mm fan I have rigged overtop of the PWM area has dropped the temperatures by 23*C!!!!! After talkin with my buddy again... he says that probably did the trick. By lowering the PWM temperature, I increased the efficiency of the PWM coils and resistor caps probably back to the 85 to 90% mark again. This in turns means more stable power to the processor. aka better overclocks with a lower voltage due to its stability...

:cool:
 
Nice. I need to water cool. I can get pretty damn high on air, with stock vcore even, but I don't want to push it for no reason. I have run as high as 3.71, and I know it will go higher... but there is no reason honestly.

I don't have a real desire to run at max overclock, hence leaving mine at 3.45-3.5 or so on air :)
 
BLiTzKRiEG said:
Well... I think I figured out the problem...

The PWM area was getting so hot that the efficiency of the coils and mosfets was decreasing to the point where I needed to increase the voltage in order to obtain a stable overclock...

The higher voltages lead to the outragous temperatures...

Last night and this morning saw at least 8 hours of prime 95 stability at 230 and 235 FSB (3.45GHz and 3.52GHz)...

Right now Prime95 is blazing away at 240FSB (3.6GHz) on stock vCore (1.3875V) under the long FFTs test.

My idle temps were 45*C by the abit utility (so 35*C) and loads at 53*C by the utility (43*C).

Heres the kicker... The 60mm fan I have rigged overtop of the PWM area has dropped the temperatures by 23*C!!!!! After talkin with my buddy again... he says that probably did the trick. By lowering the PWM temperature, I increased the efficiency of the PWM coils and resistor caps probably back to the 85 to 90% mark again. This in turns means more stable power to the processor. aka better overclocks with a lower voltage due to its stability...

:cool:
Click to expand...

my temps are about the same also....at 3.2 my idle is 45C and load is 50C
and at 4.0 idle is 47-50C and load is 55C.....and i run my fans at 40% all the time
 
spyder_flex97 said:
my temps are about the same also....at 3.2 my idle is 45C and load is 50C
and at 4.0 idle is 47-50C and load is 55C.....and i run my fans at 40% all the time
Click to expand...


What voltage are you running it at?
 
Well... I think I have found more of the limit of my RAM than I have the limit of the CPU... its older... and only rated to 233FSB... and with me at 242FSB, i think that is more logical...
 
You must log in or register to reply here.
Back
Top