Celeron D LOW temps (41C) at full load??

Stinky_1

n00b
Joined
Jan 29, 2006
Messages
4
I have a celeron D 2.4 processor. I bought it just over 1 year ago now. When I first got it I had issues with it going to 70 degrees and even higher (closer to 80) when I played games such as HL2. I this was during the initial burn in period. So I grabbed some thermal paste. (arctic silver I think) and scrapped off the stuff that came with the stock heatsink and put that in its place. Then my temps would go to 60 - 65 under full load.

Now my problem is that here I am 1 year later. And under full load (playing HL2) I am going only to 41 degrees. Some people would think this is NOT a problem. But I suspect that my processor is burning out and not giving me the full potential of what it would have when it was new. I notice that my FPS in HL2 (again) are sitting pretty low to what I would expect them to be. And since my Video card is a ATi 9600 pro (not the best, but not bad IMHO) I would expect better than 30 fps in game.

Now my question is this:

Is there some setting or somewhere that I can check my stats against someone elses PC with similar hardware? Or can I adjust something somewhere? Is it possible my proc is just slowly dying? or have I discovered every PC builders dream setup where the temps actually DROP over time without sacrificing clock speed or performance? Any help is really apreciated on this guys.

Also I used CPU-Z. It says that I am running at 2399 mhz. Multiplier is X18, and bus is 133. Total bus is 532mhz. Voltage is 1.328.

Everything it shows me on there seems fine. I wondered if maybe it had Processor Throttling (which I thought these CPU's did) and it was maybe downclocking the cpu to keep the heat lower (for some unknown reason). Sop I grabbed an app that checks if throttling is enabled. It said it detected the presence of it. So I started logging data and played my game again for a half hour. I went back through the log and it NEVER once throttled the CPU!

One other thing. When I use apps such as DVDshrink it used to take me 30 minutes to do everything. From start of the disk, to final buring. (15 - 20 mins for shrinking and 10 minutes for burning).

Now it takes me closer to 45 minutes or even 1 hour and more (depending on the movie). ANd the same thing too, the temps only go to 41. But its a SOLID 41. It doesnt variate above or below that. Its just a solid straight line. It did that before when it went to 51, as well as when it went to 61, 71, and 79 (80 was the point I told the mobo to shut the PC off)

I am getting all the temp ratings off my mobo's monitoring app called ASUS PROBE.

Even if I overclock it in the bios the clock speed goes up. But the max temp it will allow me is 41 still!

Is there some sort of setting I could have in the Bios to tell it to go higher than 41 degrees? I have a P4SP-MX mobo. Bought that 1 year ago with the rest of the system. I built it all myself too so nothing is locked in the mobo settings.
 
that particular cpu doesnt have throttling i dont believe. you'll probably see loss of frames from time to time in hl2 due to steam patches and updates( hl2 is not a good program to use to judge performance over any timeline because of this) I have no idea why your temps would keep getting lower (though being that its winter now and sumer 5 months ago). Overclock it until it wont go any further is what i would do...if you never see your temps rise, hoora! Your just using the wrong tool to benchmark, hl2 is not the answer. As for the dvd thing, coincidence, file size, compresion, etc could be altering this as well as extra programs running on your pc now that weren't before i assume could have an impact.
 
I made a norton ghost image of my pc after I loaded all the programs I initially needed. Including HL2 files and validation. I did that 1 year ago when I built the PC. I also thought perhaps it was just extra junk, possibly spyware that wasnt being picked up by the scanners and such.

So I loaded the original image back onto the pc and tried again. Even with the same drivers I had before (originally) and all that it still shows the same signs.

I am not looking really for a benchmarking program to see what the PC is capable of. I am looking to find out what is causing my HL2 to not perform the way I would have expected. I figured that the reason could be linked to the lower temps of the CPU. Certainly with ALL the eye candy on and 1024X768 resolution my processor should see a lot more load than what 41 degrees would show. I mean I idle at 34 - 35 degrees. So raising only 6 degrees over idle is NOT normal ny any means.

I just stepped up my FSB speed to overclock it. It is now at

2699 mhz
multiplier X18
FSB 150 (150 X 18 = 2700 mhz)
And 600 bus speed.

Now my voltage has lowered from 1.328V down to 1.28 V.

Now correct me if I am wrong, but shouldnt that voltage stay close to 1.3V? I did actually first put the FSB to 160 ( that has always been the highest stable clock speed I could get) So that clocked me in at 2880. But once I was loading into windows the PC reset.

I am now wondering if possibly my PSU is not good enough. So as I try to squeeze more power out of the CPU I am losing overall power output to my 1.3V bus?? ANd if the voltage is lower then the heat dissipated would be lower. Is that right?

But if that was true I would expect that my clock speed would lower as well??

My PSU is a 350W generic one that came with my case. I paid $45 CDN for the case and PSU so I dont expect it to do great things. But I didnt think I would be anywhere near 350 W or even 250W of power draw.

This is all I have in the PC

Asus P4SP-MX mobo
generic ram one 512 and one 256 stick
ATi 9600 pro 256 VC
NVTV single tuner TV card (pci)
on board sound
120 gb Seagate hdd
12X sony DVD +/- RW

Cooling includes an upgraded thermal paste on the CPU heatsink
One 60mm fan in the side of the case
one 60 mm fan on the back of the case
one standard (stock) fan on the video card
one standard fan on the PSU
and of course the standard heatsink and fan on the CPU.

A side note here. While I am typing this I am running an HDR demo (High dynamic range lighting) in the background. My Temps are not locked solid at 49 degrees.

As soon as I shut the demo off it drops right away down to 37 where it is idling (a little higher than normal but prolly from the OC).

Now I also nitice that when I turn the HDR demo off my voltage goes right back up to 1.328V

So does this confirm what I think? That my PSU is not strong enough?

Maybe I just talked myself through the problem?? heh, go figure.

or my temp probe is no good? But with it clocked at this speed now I can get the temps to go as high as 48 degrees. So it deff is showing higher now. But I would expect it to go MUCH higher. Unless I really did just somehow manage to get it running cooler.

Also about the ambient tempos being lower now. Its always possible. But I have always idled at 34 degrees. Even in the summer. And it was still summer when I started having problems with it not getting as hot (some may not consider this a problem)

I guess I will have to look for a new PSU and see if that fixes it.
 
Just found smart guardien and ran it. As soon as it started all of a sudden my asus probe started flashing and beeping saying my CPU temp was 87 degrees. And smart guardian says its 76 degrees.

Its also saying that I have NO +12V and that my -12V is actually at -16V and -5V is really at -8V.

Something seems fishy here. Is it possible my mobo isnt supporting properly the outputs that this app is looking at?
 
Or a motherboard issue...That aside, if you overclock your cpu (just overclock, not including overvolt), the power consumption still raises higher, meaning that your cpu would still get a tiny bit hotter without overvolting. Just a thought.
 
yes, you are right. But if my voltage drops down from 1.3 to 1.26 then my current draw is higher, but my total power usage is probably about the same.

That is why I am thinking its more a psu problem. It doesnt have the current capacity to keep the full 1.3V AND allow a higher current draw. So when it draws more current I get a voltage drop.

Maybe somebody else can verify what their coreV is at while running ANY app that makes the CPU go out of idle for more than a few seconds? I just want to know if its normal to see the coreV drop like this and STAY low while running a cpu intensive app?

I would think that a GOOD PSU will be able to keep the voltage constant even under larger loads.

I checked a website that shows you what the MINIMUM PSU requirments are of your PC based on the components you have in it. The thing said I need a 290W PSU. Mine is a generic 300 (maybe 350 I dont remember for sure) watt unit. BUt I suspect I am only able to get about half of that out of it. If thats the case I am running on less than 200 watts total output. Which would for SURE explain my problems.

I guess Ill have to check the threads to see what a good (budget priced) PSU would be.
 
Back
Top