Q6600 unable to run at stock speed

bmapar

Junior Member
Mar 30, 2008
13
0
0
Just put a new build together (1st time myself, but I've watched others do it before), and I'd been having troubles with it restarting/throwing errors all over the place. I decided, on a whim, to underclock it all the way down to 1.8GHz (200FSB @ 9 multiplier), and it magically decided to start working on me.

If I try and put it up to stock speed (2.4 GHz, 266FSB@9x), I can't even load windows half the time. Even safe mode crashes eventually at stock speed. I'm 99% sure it's not due to voltage, as I've set the Vcore all the way up to 1.325V, right below the spec'd limit. I'm pretty sure its not overheating: I've got a Tuniq Tower 120 put on it, and I'm running it with an open case, so the processor is idling at room temperature.

One of the few times I've managed to get into windows with the processor at stock speed, I brought up CPU-Z. I haven't bothered to disable speedstep yet, so the processor idles at 6x multiplier. I was watching CPU-Z, and right as the multiplier jumped from 6x to 9x (I right clicked the desktop to cause that) the damned thing restarted itself.

I've run several memory tests at this clock speed, all come up clean. I'm running my memory at what the memory is specified at, 2.1V, even though the default on my MB is to run at 1.9V. I'm also running the memory at specified 5-5-5-12 timings. Right now I'm running the memory at DDR2-667 instead of DDR2-800, just in case that's an issue too.

I'm completely lost. If anyone can offer any suggestions or ideas, that'd be awesome.

Parts:
Antec P182 case
Intel Q6600 processor
Tuniq Tower 120 cooler
DFI Lanparty LT P35R2T Mobo
OCZ DDR2-1000 (5-5-5-12, 2.1V) 2x2GB RAM
Corsair 620HX PSU
EVGA 8800GTS 512MB graphics card
Creative X-FI XtremeGamer
Samsung Lightscribe DVD burner
Western Digital Raptor 150GB hard drive
Vista 64 Ultimate
 

Idex

Senior member
Oct 18, 1999
435
0
0
Did you check your temps? I'm guessing the HS isn't mounted properly.
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
running open case, processor at 23C, motherboard at 31C (Northbridge)
running closed case, processor at 26C, motherboard at 35C (Northbridge)

that's from the BIOS at idle. I'm trying to get SpeedFan or something installed on it, but it's not cooperating at the moment.

EDIT: I've tried like 5 times now, and every time I attempt to run SpeedFan the computer decides to instantly restart. argggh...
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
At this point you are better off just removing your CPU heatsink and reinstalling. Just as Idex posted, while the cause of your issues could be any number of things the most likely cause is a overheating CPU and the most likely cause of that is an inproperly installed heatsink.

It can take quite a few tries to ensure you get all four clips to latch thru the mobo holes. Just because it clicks doesn't mean it is intalled, you really need to flip the mobo over and inspect it from the backside to confirm all four clips are latched thru the motherboard.
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
Clips? my cooler has no clips or pushpins, its a Tuniq Tower 120. It's got a backplate and thumbscrews (Right? Am I being a dumbass here?). I'll try removing and reinstalling it anyway though. Thanks for the ideas so far, I appreciate you guys taking the time to reply to a random newb like me.
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
I also have a tuniq. I don't think that it is possible to accidentally not install the screws properly on it, but you might have a problem with thermal compound. It is also entirely likely that you just have a bum chip. I recommend removing the tuniq, reapply thermal compound (use 2 lines instead of one so that you have thermal compound over all 4 cores) and seeing what transpires. realistically, however, with a tuniq the cpu shouldn't heat up so fast that a reboot happens nearly instantaneously. Have you tried bumping up vcore to ensure stability? If you need additional vcore to be stable at stock speed then I believe that an rma is in order...
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
Yea, I put it up to 1.325V, and that still wont make it stable at stock speed. I just took off the Tuniq and looked at it, and while its possible i didn't put enough TIM on, I think my BIOS would show temps higher than 26C if I'd messed up. There was TIM spread around everywhere, but patches were still bare metal. I'll put more on and see what happens. Sadly, I have a feeling you're right bryan. I was hoping I could avoid spending a week RMA'ing it...

Its just really weird, because if I let my computer sit at the Vista desktop once it gets running, It won't crash. However, the moment i open something intensive, It restarts immediately.

Thanks again for all the ideas.

EDIT: Corsair 620HX Power Supply. I thought it was that too at first, but i've quadruple checked all my power lines and they seem to be in order. in the BIOS, the 12V ATX is at like 11.77V and similarly for the other voltages, but I assumed this was normal drop through the wires.
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
Originally posted by: bmapar
Clips? my cooler has no clips or pushpins, its a Tuniq Tower 120. It's got a backplate and thumbscrews (Right? Am I being a dumbass here?). I'll try removing and reinstalling it anyway though. Thanks for the ideas so far, I appreciate you guys taking the time to reply to a random newb like me.

Ah, sorry, my bad. I looked right at your system specs and failed to register you had a Tuniq. So do I, so I know exactly what you are saying.

My next guess would be mobo...as in a bad VRM causing excessive Vdroop as soon as your chip is loaded and speedstep bumps the speed back up.

I say this because I had a bad Gigabyte DS3L board. It had a Vdroop if >0.2V, and a slightly charred spot the size of a dime visible only from the backside of the motherboard.

I though I had a bad Q6600 too so I swapped it with another, and then I thought I had a bad PSU so I swapped it with another. Still had the same problem. Then I pulled the motherboard and had to ridiculously closely inspect it, but finally I saw the signs of a burned out component. RMA'ed and the replacement gave me no issues.

So if you have a bad mobo like I did then what you can try and do is do your trick where you underclock so you can get into windows, load CPU-Z and while idle note the Vcore. Then have the computer do something so it is kind of loaded, don't stress it so badly you know it will reboot but you need to stress it and then take note of the Vcore.

Let us know what the Vcore readings are like. My money is that you have a bad mobo and the Vdroop is >0.2V at stock speeds.
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
Alright, the voltage in BIOS was set to 1.325V. I ran it underclocked at 200Mhz and normal speed 266 MHz (took me in the area of 10 tries before I could get CPU-Z up and running @ 266, and it bluescreened, then restarted before I could read the bluescreen about 5 seconds after I loaded the CPU). When I say loaded, I mean I dragged windows around the desktop really fast in random directions. That was enough to bring the multiplier up to 9x and make the CPU do something.

Voltages at the different clock speeds were identical:
No Load: 3.28V
Load: 3.12V
Vdroop of 0.16V....talk about ambiguous. I've gotta agree with you about the motherboard though (see below), and thats pretty impressive that you can just diagnose things like that with my poor description of things.

My friend who was helping me put the computer together earlier just reminded me that the Driver CD that came with the motherboard was horribly cracked (almost cut in half). Honestly, I just downloaded the drivers, used a thumb drive to install them, and forgot about that. I guess it would make sense if the motherboard itself was damaged. I'll check the motherboard for defects tomorrow, as it's getting late.

Thanks a lot for your quick and insightful replies!
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
I hope you meant 1.28 at no load and 1.12 at load... and yes, .16 vdroop is very excessive, even for a Quad.
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
oh wow, yea, too many problem sets and I can't remember numbers anymore. i meant 1.328, 1.312 and 0.016. I guess that means its not the MB? Still though, that driver CD was kinda fishy...I could just RMA both I guess?
 

RanDum72

Diamond Member
Feb 11, 2001
4,330
0
76
Try running the memory at SPD only. Disable EIST. Upgrade the mobo BIOS to the latest. I highly doubt the CPU is bad; I have yet to see a bad Intel CPU since socket 478 and especially if its brand new. Its more likely the mobo more than anything else.
 

flexy

Diamond Member
Sep 28, 2001
8,464
155
106
some bios setting, IMHO.
Also...please test your mem with bootable memtest CD first. Don't guess whats the problem, find the problem methodically.
 

flexy

Diamond Member
Sep 28, 2001
8,464
155
106
btw. is this the XTC OCZ?

Run them 5-5-5-15 (or 18)...the "12" seems odd to me!
TRFC 48
 

PhantomKnight

Junior Member
May 23, 2006
24
0
0
Hey hows it going. I have exactly the same m'board as you (assuming a typo) and for me it didn't even install Vista without updating the BIOS. Do you have all the sticks in. I'd reccommend if you did, take one out and flash the BIOS. DFI motherboard's shouldn't have ANY vDroop. There is a setting in the BIOS to disable it. LOL. I know, I thought it was weird to. But I read it in This Article. Good luck and I hope you succeed. BTW The Vista installation may have only been a problem for 32 bit versions and 4gb. I dont know.
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
@RanDum: According to CPU-Z, I'm running slower than SPD at 667, but I'll set it to exactly SPD. I'll disable EIST and flash the BIOS to the newest version. I don't have a floppy drive, so hopefully the thing doesn't crash mid flash...

@flexy: Nah, its not XTC OCZ. its this: Text

I've run a couple other memory tests that came out fine. The main reason that I don't think its memory is that I'm running the memory (rated at ddr2-1000) at ddr2-667. Raising the CPU speed from 200MHz to 266MHz while holding memory constant at ddr2-667 results in a huge drop in stability.

My old computer (that I'm posting from) keeps eating CDs when i try to burn memtest, so I'll have to get a friend to do it for me. I'll get that test done as soon as I can. I'll also try running at ddr2-667 5-5-5-15 just to see if that does anything.

@PhantomKnight: Awesome link, thanks a lot. That'll come in handy even more when I get this running properly and start tweaking. I'll try flashing the BIOS. as I mentioned above, I'll have to initiate the flash from in Windows, so if that crashes, I'm SOL.

And yea, its that motherboard from the article (LT not UT though). Sorry for all typos, trying to remember transient data from Cadence and info from my new computer at the same time isn't working out very well.
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
You should NOT flash your bios when the system is unstable. If you do this you are a complete idiot, I need to be frank with you so you will take this seriously. Do NOT flash your bios if your system is not stable enough to do things in windows.

Regarding memtest, etc...why not just put it on a thumbdrive and load from there? (same goes for BIOS updates too, even if you are stable in windows you should avoid flashing your BIOS from windows if at all possible)
 

Drsignguy

Platinum Member
Mar 24, 2002
2,264
0
76
Originally posted by: Idontcare
You should NOT flash your bios when the system is unstable. If you do this you are a complete idiot, I need to be frank with you so you will take this seriously. Do NOT flash your bios if your system is not stable enough to do things in windows.

Regarding memtest, etc...why not just put it on a thumbdrive and load from there? (same goes for BIOS updates too, even if you are stable in windows you should avoid flashing your BIOS from windows if at all possible)





Well put! I would'nt do it........:shocked: Btw here is a link to your Ram bmapar. OZC Looks to me that they are supposed to be set @ 5-5-5-18 @2.1v.
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
an easy way to test would just be to put in a known good chip and/or try the Q6600 on another mobo.

100% agree with idontcare re flashing bios. if you think the system is unstable NOW, try flashing the bios for problems. I just spend 15 hrs getting my computer to POST again after flashing my bios, and it had been 100% stable for 9 months at the old settings.
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
@Drsignguy, oh wow, looks like CPU-Z was telling me the CPDs things then. Still though, I was running them at DDR2-667, not DDR2-1000, so you'd think that would get me some slack on the TRAS. Yea, I was about to flash the BIOS, then reconsidered and checked this thread. Memtest is running off my USB drive right now (awesome idea idoncare! I didn't even think of doing that). I'll see what that comes up with.

I would put in another chip or try another motherboard, but I don't have any others available (my current computer is an old laptop). One of my friends has a desktop he put together, but he'd rather not take it apart and move it through the rain over here(can't say I blame him).

I'll avoid flashing the BIOS at all then, after hearing that horror story from you, bryan.

I'll post back with the results when Memtest is all done. Thanks everyone, and please keep the great ideas coming!
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
Yea, as I thought this is almost certainly not a memory issue. Memtest shows no errors through 4 passes, around 3-4 hours. If I shouldn't flash my BIOS, where does this leave me? Is there anything I haven't tried yet? Or is it time to RMA?
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
well, you have to decide if it's the cpu or the mobo. based on the info about the driver cd I would think that mobo is PROBABLY the issue. unfortunately, I don't know of a good way to test it.

Did you ever reinstall the tuniq?
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
Can you put your Q6600 into anyone else's computer system to determine if the CPU is bad? Doing a CPU swap doesn't require hauling gear around to a friend's house, just take the CPU with you and maybe some thermal grease to replace the stuff your friend will lose when he reassembles his computer after checking out your CPU for you.
 

bmapar

Junior Member
Mar 30, 2008
13
0
0
yea, I doubled the amount of TIM and put the tuniq back on. Nothing changed. I'm running this open case anyway so there's no way overheating is the issue. I really wish I had a spare cpu or mobo lying around, since as you put it, thats the easiest way to troubleshoot, but no such luck.

I'll probably just go ahead and RMA both cpu and mobo to be safe (easy way out). If the motherboard is messed up, it might have done something to the processor anyway, better to be safe than sorry. That about wraps this up then.

Thanks everyone for all the help you've offered me. Everyone was really friendly and came up with tons of good ideas. I really appreciate you guys taking the time to answer and follow up!

edit: Apparently my friend's computer is an AMD rig anyway. Everyone else I know around here just has a laptop or non-socket 775 (to horribly over-generalize and stereotype, poor college students with dells and rich college students with macbook pros). that about sums up my luck.