Oh no, possibly fried my new build :(

[dogfeet]

THE PROBLEM: Well now i've done it. Brand new system, just a couple days old. Tried to OC it to 4ghz but prime95 crashed within seconds regardless of how i configured the settings. so i backed it down to 3.8GHz. booted up fine, ran prime95 for at least 15 minutes but temps were getting high, around upper 80s. Computer crashed and shut down, when i attempted to restart i got no response.

WHAT I'VE TRIED: unplugged power cord for a few mins and tried again, still no response from either the case power button or the mb power button. sometimes i get a fraction of a second of leds, sometimes i don't. pushed the "clear cmos" button on the mb, and also tried the process to reset the bios with a jumper. also no effect. swapped out the power supply with a known working power supply, no change. took out 2 sticks of ram, no change. tried with one of the other ram sticks instead, no change.

SYSTEM SPECS
Core i7 920 D0 stepping
Xigmatek Dark Knight HS & Fan
ASRock X58 Extreme
PC Power & Cooling Silencer 750 quad
OCZ Platinum 6GB kit DDR3 1600MHz cas 7 OCZ3P1600LV6GK
80 GB Intel X-25M 2nd gen

i'm at a loss. now i have no screenshots as my system wont boot or even post to the bios. but the settings i was using to the best of my memory are as follows.
cpu 200*19
vcore 1.325
uncore 3200mhz
dram 800mhz (1600)
ram timings 7-7-7-18

i'm just so disappointed, my previous pc was an old amd 3200+, apparently an oc'ing beast, but when i tried to oc it i had similar results, i had to "hot flash" the bios chip to revive the system. now my i7 system is failing to oc. everyone else in the known universe just tries diff settings till they find something stable... when i test my limits i end up with doorstops.

any help is greatly appreciated.

 

[Boulard83]

Reseat the CPU and at the same time, watch for burned CPU/socket. Some P55 socked have issue but it dont seems to happen often on X58.

Just be sure !

 

[Brutal-Force]

i'm just so disappointed, my previous pc was an old amd 3200+, apparently an oc'ing beast, but when i tried to oc it i had similar results, i had to "hot flash" the bios chip to revive the system. now my i7 system is failing to oc. everyone else in the known universe just tries diff settings till they find something stable... when i test my limits i end up with doorstops.

any help is greatly appreciated.

One reason why you work from the bottom up and not the top down, and in small increments.

Does your Mobo have a battery in it? If it does you might try removing it, although I don't see how it would help much if by jumping the CMOS didn't clear it. Also on some computers you have to double boot. Hold the power switch down then let go, then hold down again. Sometimes works for me.

Might also want to make sure your CPU fan isn't being stopped by a wire, and that your heatsink is properly installed. That could cause it to not boot, at least not for long.

 

[dgb303]

Yeah I'd try pulling the battery out and letting it reset.

 

[kkpudge7]

1.32v isn't really enough voltage to fry an i7. My vote is to pull the battery and let it sit for about 10 minutes. Report back if that had an effect.

 

[rainless]

One reason why you work from the bottom up and not the top down, and in small increments.

What's wrong with Top-Down? Won't his 4ghz overclock trickle down to the rest of us?

I agree: It's REALLY harder to kill a CPU than you think. I mean a BOARD is pretty easy to kill... but not a CPU.

Just mess with the CMOS a bit and you should be fine.

 

[Mega_Death]

1.32v isn't really enough voltage to fry an i7. My vote is to pull the battery and let it sit for about 10 minutes. Report back if that had an effect.

Agreed, reset the CMOS by pulling the battery with the power supply unplugged. No way you fried anything.

And the burnt cpu sockets are on P55 motherboards not X58.

It crashed due to an unstable overclock.

 

[SanDiegoPaul]

One reason why you work from the bottom up and not the top down, and in small increments.

.

Yep - always start with the very basics - everything at default.

 

[dogfeet]

Thanks for all the suggestions guys, they're much appreciated. So as per the above recommendations I just finished unplugging the computer, and removing the mb battery and the heatsink as well. The battery was out for about an hour. I pulled the CPU out and there doesn't seem to be any visible damage (burn marks, bent pins, etc) to either the processor or the socket. I reseated the heatsink and put the battery back in.

Unfortunately, still no change. I tried the double boot idea that was suggested, also no effect. I still get only a split second of power and then nothing. The only indication that i'm even getting power is that the LAN light is on throughout it all.

Anyhow, hopefully you're not all tapped out of ideas, i'm open to more thoughts, thanks.

 

[tinymouse2]

Have you got a spare PSU? What exactly happens when you press the power button? Any LEDs turn off? or on? Basically does anything at ALL happen?

Also I'm assuming you've tried just a basic boot? (1 stick of RAM, CPU, PSU, Heatsink, GPU and MOBO)

An hour is plenty of time to clear the cmos. It takes about 15 mins to clear so your WAAAY past that (I usually leave it for 20 mins max)

So yeh, do you have another setup where you can test PSU, GPU, RAM, or anything else? (obviously excluding HDD and optics)

 

[bda1967]

Test your PSU http://www.overclockers.com/forums/showthread.php?t=403837 using the paperclip trick. http://aphnetworks.com/lounge/turn_on_psu_without_motherboard_the_paperclip_trick Do you have a digital multimeter?

 

[vgta88]

The lan should run off 5vsb. 5vsb is always on and is separate from the other power rails so the light is no way to rule out a psu(or motherboard receiving power) except when you used a known working psu.

It really sounds like the board croaked.(or cpu but that is highly unlikely)

I've had a board blow caps but the cpu was fine (should be noted the psu was working, if the psu went too then everything would probably been over volted/amped) 80 degrees you say...I'm not sure if that is the max for i7 systems. °F or °C? edit: nvm you're Canadian. its c.

As for bios reset, I don't know the proper procedure but what works for me is unplugging the psu and doing the cmos clear jumper for 1 second with the battery in. That works right away for me.

 

[OnDborder]

Welcome to the Forum! http://www.ocforums.com/showthread.php?t=491697 dogfeet and SanDiegoPaul

 

[dogfeet]

tinymouse - yes, as per my first post i do have a spare psu, i tested it and i get the same result out of the new system... after pushing the power button i get a fraction of a second of fan LEDs and the digital read-out on the mb blinks but not with an actual number or letter, just a random pattern. the fans do begin to spin but only get maybe a quarter rotation at best. then fans stop, and LEDs go out (except for the LAN LED). unfortunately i dont have another setup that i could swap parts with. i'm currently on an AMD 3200+, ddr ram, agp graphics... nothing's compatible.

bda1967 - no i don't have a multimeter, and the paperclip trick is very interesting, i'll be sure to use it in the future... however for my present dilemma i'm not sure it would be useful since i did test a known working psu and that didn't solve my issue. thank you though.

vgta88 - i tend to agree with you, all signs are pointing to a blown mb. i'm quite sure that 85°C is not enough to nuke a 920 i7.

now i've not just been resting on my laurels either, i have an IT friend whose brain i picked today, i wanted to know how much i could remove from the motherboard and still get some activity. he confirmed that i could remove the bios chip and the cpu/ram/vid card/hd/burner and have only the psu and mb. granted there would be errors/beeps but it should still be active and powered.

i tried it. i pulled everything out, even 3rd party usb ports and all. since my mb has a built in power button i also disconnected my case wire harness. in the end i had only the PSU and the MB (minus it's own bios chip). after all of the that, i got the exact same result. a split second of power which i could then only see in the mb's digital debug read-out.

so now that every variable possible has been removed from the equation, does everyone here agree that my mb is now a slab of charcoal? is there anything i've missed? honestly, i dont really think this had anything to do with my OC attempt, my settings were not that aggressive and there should have been built in protection against this sort of catastrophic failure. since this was the first and only time i ran a torture test for more than 1 minute, i suspect it would have happened whether i OC'd or not. a shame my warranty is now void for what may have been an existing defect.

oh, and thx for the welcome ondborder

 

[Brutal-Force]

Why is your warranty voided? Overclocking is part of the game, and all Motherboard manufacturers know it. If they didn't want you to be able to overclock, they would lock all of your settings like Dell does. In addition (and I am not saying be deceitful) the Motherboard manufacturer doesn't know that you are overclocked, much less how high. Just RMA the motherboard. It is quite possible you just got a bad one. It happens.

 

[dogfeet]

i agree completely, when manufacturers market their products for their OCing abilities then they should assume some responsibility for the OC risks. but alas, a common clause is to void the warranty in case of OC, my MB has this clearly written in the user manual. but perhaps you make a good point, far be it from me to be nefarious, but indeed, how will they know? unless of course my MB was too far gone to properly clear the CMOS and my OC settings are all still there, just waiting for asrock to put my bios chip in a reader and deny my RMA claim. i have nothing to lose from trying though, perhaps they will be forgiving.
but i would like to be reasonably certain it's the MB before i send away for a replacement/buy a new one.

 

[Brutal-Force]

You don't think someone actually checks out your motherboard.

You also don't think they honestly think you DON'T try to overclock. After all its why we bought the i5s and i7s to begin with.

They know what happened, and they will bite the bullet on a cost of maybe 20 bucks to them. Its more expensive to pay someone to "Check out your Motherboard" then it is to just send you a new one.

Think about it. They have to pay a RMA Phone Jockey, Keep a Building or Room to "Test" your Motherboard, and Pay someone $15 dollars an hour to sit around and wait for RMA'd Motherboards that they "know" have been overclocked.

Honestly, I see them more like McDonalds.... its more of a hassle, so they throw the Motherboard in the trash, write it up as a loss and send you a new one. McDonalds would rather take the loss, then have you stand out at their cash register and complain to all the other customers about poor service. Word of mouth travels, especially for us internet users.

 

[WonderingSoul]

Try booting the board out of the case. If that doesn't work, I would RMA it. There are clear issues wrong with that that would have probably arisen later even if you weren't OCing.

 

[visbits]

I had 3 ASUS boards, the first just died randomly, the 2nd did what yours is doing, I had to bend the board in a certain way for it to work. that board died and the 3rd one killed 3 set os crucial ballistix, at that point I sold the 4th RMA. lol

 

[CompuTamer]

Try pulling all of the RAM out and unplugging everything from the board and turning it on. See if it'll do anything more for you. If it doesn't board's probably dead.