• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

E6400 suddenly won’t overclock.

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

boead

Registered
Joined
Sep 13, 2006
The last few years I’ve been using a LianLi case, Enermax PSU (cost around $120) with a Swiftech water cooler. Twin radiators in the front only cooling the CPU.
Mainboard is an Asus P5BDelux-WiFi, CPU a C2D 6400, HighPoint SATA PCIe RAID controller, ATI X1900xtx video card, Maxtor 400GB drives, 2GB G.Skills PC2-6400phu1
Vista Ultimate 32 retail, ran the beta till February – its been great. Viva La Vista except OGL doesn’t work :rolleyes: .

I had the machine at 445x8 for the longest time (since October of last year). Voltage fairly high at 1.55v, mem at 1:1 (4-4-4-12) 2.35v, NB/SB voltage up too. Again, been doing this for a long time – tried dozens of combos till I settled on this on. Prime stable.


After having issues suddenly a week or so ago, I had to restore the system due to horrible system instability (crashes constantly, hard drive errors, services wont start and stop suddenly) after the restore and setting the ststrem at the stock clock (SPD) all was perfectly fine again, completely back to normal, rock solid and not overclocked at all, I ran it that way for two days without any problems at all – very happy.
I then bumped up the OC to modest 420Mhz which was below the usual 445+ I had been running for months. System became unstable immediately. I monitored temps which never went much above mid 40’s. I dropped memory timings at 420 (1:1) from 4-4-4-12 to 5-5-5-15 which the system SPD’ed at when at stock clock at 400Mhz (2:3). Same instabilities. The system has sporadic explorer and IE failures, services fail or wont’ start and eventually the system will re-boot. Once then twice then more frequently till it won’t boot again!
It will NOT run overnight running lets say, iTunes and I didn’t even bother running Prime95.
If its not overclocked it’ll run iTunes overnight fine and Prime endlessly. So this is definitely a sudden overclock problem.

I tried all different and mild OC’s but NONE will work. System simply doesn’t’ want to be overclocked anymore yet at a measily 2.1 GHz its fine.

What’s also strange is that when loading Vista, the graphical horizontal bar that swipes across the screen when loading does so exactly 5 to 6 times before proceeding to the next screen. If I OC at all, the swipes go to 13 to 16 times. Obviously its encountering a problem, overall boot time is longer at 3+ GHz then at 2.1 GHz.

None of this was so a month ago. Visual inspection shows the computer looks fine, all working and running cool.


Lastly, after a few hours of OCing and crashing, the system becomes corrupt and needs to be restored from backup!


My first guess would be the motherboard is going bad, my second guess is that the CPU is failing. Both have been given way more voltage then they were designed to have. Cooling was never an issue for me. CPU always ran relatively cool and the case is very well ventilated and has giant fans blowing down and around the board.


At stock clock;
E6400 - 8x266=2133 1.3v
G.Skills at 400MHz (2:3) SPD’ed 5-5-5-15.
P5B-Deluxe bios 0711 2.4, 9/29/2006 8.00.12

OCCT v1.10 - 100% 2X CPU load
Intel TAT temps 36º - 39º C * not surprised, my water cooler is really good and has always been very efficient. Even with 1.55v (3.6GHz) double prime (beating the snot out of it) temps hardly hit mid 60ºC.
Completes test over and over again.

Running Orthos now just fine. I’ll let it go overnight.

So if its completely stable at stock clock, what the hell could it be? If the CPU and/or chipset was damaged then I’d expect an unstable system at any clock.

Some have said that it’s been observed with the C2D’s that as the CPU and chipsets degrade from overclocking they need MORE voltage to obtain the same OC. This is one thing I have not tried since 1.55v is already fairly high.
Of all the CPU’s I’ve overclocked (and I’ve overclocked many dozens with no exaggeration) I’ve never had this happen, that is a stable OC suddenly gone bad for no reason. I’ve had CPU’s and chipsets die, dead, never to work again at any clock. I’ve had memory fry and all sorts of other stuff but nothing as alluding as this.


I’m at a loss.


Photos of current system before liquid filled.
Not much has changed since these photos except I added a PCIe RAID card.

Core2Duo003.jpg


Core2Duo004.jpg


Core2Duo005.jpg


Core2Duo006.jpg


Core2Duo007.jpg
 
Last edited:
inkfx said:
Your memory may have gone bad. One of my Crucial sticks was bad, I didn't notice for months.

Will ORTHOS find it from its 'blend of CPU and RAM' ??

How did you find it?
 
Well my system just plain wouldn't post anymore so it was pretty easy to narrow it down at that point. Memtest86 is probably the best thing out there to test your memory. Sure you have to make a boot diskette but it's worth it in the end. Orthos may find the problem, not sure.
 
So I ran Orthos (Edition SP2004) for nearly 9 hours. It was still running when I woke up this morning. Vistas screensaver was running. I woke it up, looked at the progress on the screen. Orthos still running, temps at about 39ºC. I launched IE and I watched the RAM gauge in the Vista gadget jump to 100% and then BAM, it re-booted.

I then installed and ran MemTest 3.5, a windows based memory tester. It ran to about 70 something % and failed. Message said memory test failed, system didn’t lockup or re-boot.

I took the machine to work and threw in a floppy drive and ran memtest86 v1.7. It ran on one stick (1GB G.Skills) for just over 2 hours till I stopped it, then it ran another 2 hours on the other stick of memory. Then I let it run for well over 4 hours with both sticks in. No errors! That’s at stock clock, stock voltage, memory at 400Mhz (2:3) at 5-5-5-15 timings.

I’m running MemTest 3.5 in Windows right now. We’ll see if it runs. I also have Memtest on a bootable CDR that I can run over night.

In them mean time, I ordered a new Intel E6600 which I can use anyway at work. I’d like to try a higher multiplier and a bugger cache.

I’m also looking at motherboards. Not shure which one I’d lie to try. Seems like the 975 is on its way out. Many more boards with the 965. However, I kid of like the 975’s reputation and I have a PCIe RAI controller so I don’t’ really care about the SB raid support. I DO care about compatibility with my ATI X1900, I’d like to try crossfire with a pair of 1900’s at some time. I don’t’ need WiFi but I do use Firewire.

The 680i chipsets are a bit expensive.

Another Asus? Abit? 965? 975?
 
Orthos and Memtest 3.5 have been running for over 10 hours now. Go figure.
 
So I left Memtest-86 3.3 running since this morning. The system boots off a CD and runs the app - NO windows.

With both stick in produced 83 errors over 10 hours and 5 minutes before I stopped it.

Test/Errors
1/0
2/0
3/0
4/40
5/7
6/23
7/11
8/2
9/0
I stopped it...


I'm going to pull a stick and run the test again overnight.

If just one of the sticks produced errors then its obviously a bad memory module and I can get it replaced.

If they both fail, I can also suspect that its more then the memory and likely the north bridge. I can also swap the memory with sticks I have at work.

* in Windows, memtest ran over 11 hours without any errors. Orthos was ALSO running at the same time and the system appeared stable - no errors.
 
inkfx said:
Your memory may have gone bad. One of my Crucial sticks was bad, I didn't notice for months.

That was it!

Problem turned out to be a memory stick. A 1GB G.Skills PC2-6400 4-4-4-14 module. BTW: The memory errors didn’t’ appear within Windows, apparently a memory test can NOT test all the system ram while running. Memtest-86 3.3 boots off a CDR and can test ALL the system memory, it found errors in one module regardless which mainboard slot it was plugged into, the other stick passed all tests in all slots (each test was over 10 hours, it took days to test everything).

Still, I have a E6600 coming my way today. I’m interested in seeing how well it OC’s. With the E6400 I get 445 FSB easily and can get 455 with some extra voltage. Memory bandwidth is over 7300mb/sec. Not sure what the E6600 will give up but we’ll see.
 
Back