Printable View
Quote:
Originally Posted by JeffLo
It is common for sensors on some of the cores to get stuck, but Core# 2,3 should not reach a higher temp than core# 0 while iddled. If your supplier allows you to return in based on sensor temps, then I would advise you to go for it.
I wouldn't bother. As has been said above stuck sensors seems to be common with the new 45nm chips, however I should note that this is in no way a major issue. The DTS has never been designed to provide accurate readings at idle - it's calibrated to trigger a thermal shutdown reliably, and nothing more. Your stuck sensors will work perfectly fine for that purpose - eg. my Cores 1&2 are stuck at 44 degrees, but when I load up my 9450 with OCCT the readings hit and display above 44 degrees precisely in line with the other two cores that read below 44 degrees at idle - as is happening above in your example.
thanks for the info, so what temps should i look out for in terms of max temp? should i base my temps on cores 1&2 or should i based my max temp on 3&4?
Both. As you can see, your stuck sensors are working perfectly fine in giving you readings above their stuck value - they're just not reading below that value. Ignore them for looking at idle temps, but they're still useful at load. In terms of max temp, you need to go on RealTemp numbers, as CoreTemp uses the old, incorrect offset value resulting in temps that are 10 degrees higher than they should be. Judging by those, anywhere over 60 degrees is dangerous from a longevity perspective.
I'd look at doing a re-mount of your TRUE to be honest. Discounting idle temps due to stuck sensors, there's just too great a variation in the temps of the first die (Cores 1&2) and the second die (Cores 3&4), indicating to me that you have an uneven mount.
ok thanks for the info i'll try remounting my hsf, thanks
Hiyas, new guy and new computer here :)
Asus Rampage Formula
Q9450 VID 1.2500
4 sticks Mushkin 2GB PC8000
700w PSU
Noctua NH-U12P air heat sink.
I must say, I'm impressed by the mobo and chip, it booted right off the bat with 450Mhz FSB with 1.5v NB and 1.34v FSBT with everything else on auto and it's prime95 stable. :)
Now, I have one problem and that is the north bridge. It is hot, and I mean very HOT! First time I OC'd my machine I set everything on auto and booted with 450 FSB, once in windows I started Prime95 and then a minute later started Asus Probe. It went haywire and told me that NB had a temp of 77c and it was climbing :shocked:
Nedless to say, my machine crashed and burned shortly after that.. ;)
It turns out the reason to that is due to my big heat sink that covers NB's heat sink blocking air flow and heat dissipation. Temporary fix was reducing NB voltage plus zip tie a 80mm fan to my case and have it blow air to the heat sink. It's a ugly fix but it works..
Now, anyone here that's knowledgeable about the heatpipe cooling system that is mounted on the Rampage formula board? I want to replace it two Thermalright HR05-SLI sinks but I'm uncertain in how I should remove the heatpipes and if that would have any effects on the cooling of the voltage regulators.
The heatpipe cooling on the board is splitt in to 2 peices 1 that cools SB + NB and one that cools PWN area. So u can safely remove that part that cools NB + SB and keep the part that cools PWN, hey it should even reduces the heat on the Pwn part.Quote:
Originally Posted by Nordmann
I have done this on my Maximus formula and I think they used the same idea on the rampage formula, but I fitted watercooling on the NB(35c load now) and keept the SB cooling as it was. I used a hairdryer to seperate the NB cooling from the chip it sits with some stupid glue but it works if u heat it up or freeze it then it is easy to remove remember done forces it off it will come of gently with the right amount of heat/freezing.
http://b3lial.rahulsjohari.com/GRAPH...T/xs_C2Q_1.jpg
Got my new Toy :D
Can anyone tell me if this a good batch or anything about it - looking at this: (Someone earlier had asked me to post batch/info from the sticker)
http://b3lial.rahulsjohari.com/GRAPH...T/xs_C2Q_2.jpg
Thanks, that sounds great!
New coolers ordered and when I receive the new heatsinks I'll lap the cpu and it's heatsink while I'm at it. CPU temp's are a bit funny with one core at 34c, second at 20 (!?!) and the other two at 40 when idle.. at full load temp's are at 57, 48, 58 and 58. while it seems that atleast two of my core's have stuck sensors the other two have a fairly large gap in the temperatures which makes me believe I messed up when I fitted the heatsink. On a side note.. IC Diamond thermal goop is a big pain in the behind to apply evenly, it's almost like clay. ;)
Hi! I'm having trouble getting 3,95 Ghz or even 3,9 Ghz with Q9450.
Here are my specs:
Q9450
Striker II Extreme bios 0704
9800 GTX
2x2gb 1600 Mhz
Vista 64bit
I tried 3,9 with these settings but my computer wouldn't start:
Vcore 1,6 V
VTT 1,62 V
PLL 1,5 V
NB 1,7 V
SB 1,7 V
Memory 2,0 V
Could someone help me with this? Soon I'm going to get water so I can raise the voltage even further but I wonder why Vista won't start with those voltages and FSB?
Lower the vcore a bit eh, you're killing it!Quote:
Originally Posted by Joonas612
Hmm. Would it help if I lowered it to 1,45 V?
I would try to keep it below 1.4v under load in Windows. The same goes for VTT too.
Very few Q9450's will do over 3.8ghz on current boards, so 3.9ghz is very optimistic tbh.
Im going to chime in and say the NB is incredibly high also...Something's lifespan has been considerably shortened!
So it might work better with lower voltages? Thankfully I tried those insane voltages only once :D
No, it won't do any better at lower voltages... but it sure won't work any better with MORE voltage because of FSB walls... Most Q9450s, as previously said can't get more than 3.8GHz on a DECENT board.Quote:
Originally Posted by Joonas612
All you're doing by giving it more volts is killing the chip.
You should never go above 1.4v Vcore and 1.4v VTT. You might have killed your chip already or severely degraded it.
Be happy if you can get 3.8GHz, as many people can't even get that.
I ve an Asus p5k deluxe, and tested my 99450, the best OC is 3,76ghz at 1,31v in cpuz and 1,33v in bios, but at 3,8 rhw welcome screen of the XP freeze, y put 1,34,34,or 37v and dont star again welcome screen i dont know what happen but im hapy with my quad...ah my VID is 1,215.
http://img179.imageshack.us/img179/1...lockyg9.th.jpg
I have just overclocked the Q9450 to 3.6 Ghz stable. But volts seem high, haven't tweaked any other setting, just vcore. Maybe its due to having a bad VID = 1.25.
Q9450
Rampage Formula
XFX 9800GTX
4 GB of G.Skill Ram (2GB x2)
For the above pic, imageshack didn't have my username available so don't be surprise if it says shadowsaviour
That's good to know :P I managed to get 3,9 Ghz and run SuperPI in safe mode but nothing else. Would it make any difference if I had water cooling?Quote:
Originally Posted by Loser777
Here's a sceenshot:
http://img66.imageshack.us/img66/3314/39x0fr3.th.png
It seems that Q9450 is not a good processor for overclocking?
Quote:
Originally Posted by Shadowthor
May be vid 1,25 is not ideal to do and oc beyond 3,6ghz, may be is a little hi if u look at other q9450 vids, mine is 1,215, is not the best to but not teh worst, i see somee hi temps in your pic,, whats your cooling system??
Is VID 1.2500 really that bad for OC's above 3.6?
I had no problem booting Vista at first try on my Q9450 with VID 1.2500 at 3.8Ghz with these settings
Vcore 1.3v
NB 1.6V
VTT 1.35v
PLL 1.52v
SB 1.1
it wasn't prime stable and due to my north bridge cooling problem I mentioned earlier I decided it wasn't worth pushing my system until I get it properly fixed. It is going to be a while until I can do that as the shop I ordered from is sold out on the chipset coolers and wont get it restocked in two weeks so I have to find me another shop that sells them.
is it just me or are these 45nm procs may not demand a lot of vcore, but greatly requires a lot from the board (especially NB & even RAM) :(
That's exactly the current situation.Quote:
Originally Posted by emoners
Nope water-cooling won't make any difference to these chips. They can usually be clocked to their current limits on decent air-cooling.Quote:
Originally Posted by Joonas612
And a 3.9ghz SuperPi suicide is the best I can manage too. Anything over 490mhz refuses to get past the Windows boot screen. I'm waiting for the Abit P45 boards to hopefully beat that wall.
I'm using the TRUE.