so strest testing OC and computer shuts off???

[the1320god]

new system for my lil girl- msi 770-45c MB
amd 965be c3
mushkin mem 1333 7-7-7-20
600watt psu ocz
ok, lets keep these crazy probs going... lol. so I'm playing around with this new computer, and starting to overclock it. so I'm moving up the speeds, a lil at a time. i had it at t 3.8ghz with 1.46vcore. so i move it to 4.0ghz, and bsod 124. so i move the vcore to 1.5v and i move my pcie from 100 to 106. ok now this thing is doing good. made it though 3 pass with LinX maxing out all the mem. temps at 56C max. and then the computer shuts off? im like wtf! and it wont turn on at all. so for ****s and giggles i unplug the 4 pin power connector that powers the cpu part of the board, and try to turn it on. it turns on fine. so i plug it back in and the same thing happeneds. i can see all the led's in the fans try to turn on, then they turn off..... so what is it? cpu? board? or maybe psu? also unplugged it and turn it back on, went to plug it in and seen some sparks from the opposite side of the connector (the side u see when its installed) anybody!lol... please help wit this computer from hell!

[CrimInalA]

pc's for little girls don't need to be overclocked . :confused:

Regarding your question : clear cmos first .

[the1320god]

pc's for little girls don't need to be overclocked . :confused:

Regarding your question : clear cmos first .

i know they don't. but its like having a fully built drag car, and just cruising around. if its there, see what it has. lol. i did clear it. same thing. it has to be hardware. I'm just trying to pin point it as i just took a new asus motherboard to the ups store to send it back. i can get it back in the morning if i need to, I'm thinking its the motherboard but i not 100% sure.

[hydrotoxin]

also unplugged it and turn it back on, went to plug it in and seen some sparks from the opposite side of the connectorSame thing happend to me benching a crappy F2 Windsor @1.7V on a DFI M2RS ->
http://www.abload.de/thumb/dscf6563rs808l6.jpg (http://www.abload.de/image.php?img=dscf6563rs808l6.jpg)
.... it keeps running :eek:

[the1320god]

i think its my fault for using a :banana::banana::banana::banana:y board. well u live and learn. im going to get my crosshair III back from ups in the morning. lol

[xman01]

if you see sparks try a different connector

had a similar problem with a 9800pro, switched molex and was good to go

[Particle]

If your system turns on when your aux 12V connector is disconnected but not when it is connected, it almost has to be a PSU problem with a blown (or tripped) 12V rail or something similar. That would be exceptionally rare.

I would recommend you test the system using a different PSU as a diagnostic procedure.

[PatRaceTin]

thermal safety shut the pc down

[the1320god]

if you see sparks try a different connector

had a similar problem with a 9800pro, switched molex and was good to go its the only one on the psu.


If your system turns on when your aux 12V connector is disconnected but not when it is connected, it almost has to be a PSU problem with a blown (or tripped) 12V rail or something similar. That would be exceptionally rare.

I would recommend you test the system using a different PSU as a diagnostic procedure. it was under a bad load when it happened. (LinX max mem). and when it is connected, and you try to turn it on, you hear this high pitched sound. it only will turn on for lil 1/4 of a sec. but unplug it and its fine???


thermal safety shut the pc down how do you do that?

[Particle]

it was under a bad load when it happened. (LinX max mem). and when it is connected, and you try to turn it on, you hear this high pitched sound. it only will turn on for lil 1/4 of a sec. but unplug it and its fine???

My reply to this information would be to quote what I already said. It's still valid, and my suggested test is still what I think you should do.

[the1320god]

My reply to this information would be to quote what I already said. It's still valid, and my suggested test is still what I think you should do.

sounds good. sucks i just sold my extra psu... i went and bought a new psu, a way better one. (Corsair CMPSU-750TX 750-Watt TX Series 80 Plus) and ordered the asus crosshair iii as well. (called the ups store and the one i returned shipped out last night) :(. oh well, replacing with better is always good!

[amd64majin]

i think that should be the powersuply

[the1320god]

i don't know why i didn't think that first. maybe cuz i HATE this motherboard, just wanted a reason to get the asus. when i built my i7 rig with the x58 motherboard, i had a enermax rev 1050watt and it did the same thing. turn on for about 1/4 sec LED's on fans would try to turn on and then shut off. the light on the bad was then red. like it had a short... and this is doing the same thing.well i ordered both. they will be here tomorrow. i will just change the psu and see if that was it. i will post the results. thanks all.

[the1320god]

so new parts came in 2 days ago. got it put together, works like a champ. i smelled inside the psu and oh yeah, smells of burnt electrical! nice. thanks for the help

[[XC] Oj101]

its the only one on the psu.

it was under a bad load when it happened. (LinX max mem). and when it is connected, and you try to turn it on, you hear this high pitched sound. it only will turn on for lil 1/4 of a sec. but unplug it and its fine???

how do you do that?

I'm sorry to hijack the thread but that's almost the exact problem I'm having. Firstly, my specs:

Phenom 2 965 @ 3.7GHz
MSI 790FX-GD70
2GB OCZ SLI DDR2000 RAM
Radeon 4670 OR GeForce GTX260 (tried both)
AOpen (rebranded FSP) 700w quad 15a 12v rail

I can crunch for weeks without a single issue, but as soon as I add ANY 3D load, be it a game, Furmark, 3DMark or GPUGrid the PC goes into a "suspended" mode, exactly what it is like if you hold in the restart button. I hear the DVD and hard drive power cycle and the screen switches off. If I hold the power button in for five seconds the power LED on the case continues to flash and pressing the power button doesn't do a THING. It remains dead until I pull the plug and allow the caps to drain. Certain things such as the flashing LED tell me it's motherboard but everything else says PSU. I have tried two cards, one stick of RAM at a time, reset the BIOS, reseated the CPU... Everything except a difference board or PSU as I don't have either. The PSU was "abused" several years back, I benched an E8600 under LN2 and dual 4870x2s as well as dual 9800GX2s. Where would you say the problem lies? At the moment I don't have access to replacements for either and can't afford to buy the wrong component.

[BeepBeep2]

I'm sorry to hijack the thread but that's almost the exact problem I'm having. Firstly, my specs:

Phenom 2 965 @ 3.7GHz
MSI 790FX-GD70
2GB OCZ SLI DDR2000 RAM
Radeon 4670 OR GeForce GTX260 (tried both)
AOpen (rebranded FSP) 700w quad 15a 12v rail

I can crunch for weeks without a single issue, but as soon as I add ANY 3D load, be it a game, Furmark, 3DMark or GPUGrid the PC goes into a "suspended" mode, exactly what it is like if you hold in the restart button. I hear the DVD and hard drive power cycle and the screen switches off. If I hold the power button in for five seconds the power LED on the case continues to flash and pressing the power button doesn't do a THING. It remains dead until I pull the plug and allow the caps to drain. Certain things such as the flashing LED tell me it's motherboard but everything else says PSU. I have tried two cards, one stick of RAM at a time, reset the BIOS, reseated the CPU... Everything except a difference board or PSU as I don't have either. The PSU was "abused" several years back, I benched an E8600 under LN2 and dual 4870x2s as well as dual 9800GX2s. Where would you say the problem lies? At the moment I don't have access to replacements for either and can't afford to buy the wrong component.


It could be the PSU, my OCZ GameXStream 700w (FSP Epsilon) powers my 965BE @ sig and 5770's fine, it gets a little warm but nothing major.

However, my PSU is rated @ 18A per 12v rail...your AOpen seems to be an overrated FSP 600w.

Could also be the motherboard, the GD70 is only a 4+1 Phase design and it's been known to fail in the past.

[[XC] Oj101]

You're not helping... :p:

Internally it is an underrated (current-wise) FSP 700w, identical to your unit. A 4670 without external power causes the shutdown even though its power draw is in the double digit numbers, a GTX260 with dual 6pin connectors does the exact same thing, crunching or Intel Burn Test can run forever without a single hitch?? 1.2GHz, 3.4GHz, 3.7GHz, makes no difference :(

[Particle]

At 15A per rail, that's pretty weak. It wouldn't surprise me if you either have unbalanced rails (too much stuff on one +12V). If that is the case, you might benefit from distributing the power connectors in your system among your rails a little more carefully. It's also possible that you just won't be able to get away with 15A rails. The GTX has a 182W TDP by itself after all. A 15A rail is only supposed to provide a maximum of 180W and leaving some breathing room is advisable. Combine a moderate 3D load with a moderate *other stuff* load on that same rail and presto--your PSU shuts itself off because of overcurrent. If you had a cheap PSU instead of an FSP, it would simply have let out the magic smoke.

In any case, I recommend you try to rebalance your rails a bit. Give the GPU one completely by itself as well. Ultimately, you may need to test your system on a friend's PSU that has a single large 12V rail or at least beefier split rails to determine where the problem lies.

[BeepBeep2]

...
I have a rail distribution chart if that helps any.
From OCZ:
12V1: CPU1 (EPS12V Part 1)
12V2: PCI-E Connector 2 / CPU2 (EPS12V Part 2)
12V3: M/B + Accessory (Molex, Sata, 24-pin, etc.)
12V4: PCI-E Connector 1

FSP PSU's aren't that great either. Ripple is a horrible problem on these things...it puts alot of strain on capacitors and delivers power thats not-so-clean.

[Particle]

I'd try disconnecting as many components as possible, plug in 12V1 to the mobo, 12V4 to the GPU, and 12V3 to any essential devices. See if it behaves differently. If it does, the culprit is power. If not, it's just as likely to be power. :) You're kind of in a hard place on this one. Test underclocked & undervolted or with another, beefier PSU to see if you get different behaviors.

[AMD_Freak]

a good rule of thumb .....

Here is a direct quote from our very own Chew*:

“Im sure some of you may have experienced a crash with cinebench………sometime you will blue screen, somtimes you will just black screen and sometimes the bench will just crash ( dissapear, etc just shut down ) and windows will still be up………..the blue screen is NB vid/IMC memory related, the black screen is core clocks/cpu voltage related and the just crash/dissapear from desktop is temp related………….”

[[XC] Oj101]

At 15A per rail, that's pretty weak. It wouldn't surprise me if you either have unbalanced rails (too much stuff on one +12V). If that is the case, you might benefit from distributing the power connectors in your system among your rails a little more carefully. It's also possible that you just won't be able to get away with 15A rails. The GTX has a 182W TDP by itself after all. A 15A rail is only supposed to provide a maximum of 180W and leaving some breathing room is advisable. Combine a moderate 3D load with a moderate *other stuff* load on that same rail and presto--your PSU shuts itself off because of overcurrent. If you had a cheap PSU instead of an FSP, it would simply have let out the magic smoke.

In any case, I recommend you try to rebalance your rails a bit. Give the GPU one completely by itself as well. Ultimately, you may need to test your system on a friend's PSU that has a single large 12V rail or at least beefier split rails to determine where the problem lies.


Keep in mind that the GTX260 has TWO PCIe connectors, so that's 30A or 360w to play with, ignoring the 75w from the slot. It used to work fine for a long time with the 4670, but a few hours after installing the GTX260 the problem started. I thought I killed the card but after going back to the 4670 the problem still existed. I can safely say it's not overcurrent as the system used to work and the PSU has powered both 2x 9800GX2 and 2x 4870x2 with a Wofldale at over 5.5GHz (see my Hwbot profile, all the high end stuff was done with this unit and the older AthlonXP, etc was done with an older AOpen 550w (the old gold one)). I know what I need to do but as I said, I currently don't have access to either a new board or meaty PSU.


...
I have a rail distribution chart if that helps any.
From OCZ:
12V1: CPU1 (EPS12V Part 1)
12V2: PCI-E Connector 2 / CPU2 (EPS12V Part 2)
12V3: M/B + Accessory (Molex, Sata, 24-pin, etc.)
12V4: PCI-E Connector 1

FSP PSU's aren't that great either. Ripple is a horrible problem on these things...it puts alot of strain on capacitors and delivers power thats not-so-clean.

Thanks Beep, but I already knew that from the sticker on the PSU :p: I know that by todays standards they're not great but when I got it it was one of the best PSUs available.


I'd try disconnecting as many components as possible, plug in 12V1 to the mobo, 12V4 to the GPU, and 12V3 to any essential devices. See if it behaves differently. If it does, the culprit is power. If not, it's just as likely to be power. :) You're kind of in a hard place on this one. Test underclocked & undervolted or with another, beefier PSU to see if you get different behaviors.

I've played around with all of that, 1.2GHz at 1v, stock, 3.7GHz at stock vcore and 4GHz at 1.6v - the symptoms never change.


a good rule of thumb .....

Here is a direct quote from our very own Chew*:

“Im sure some of you may have experienced a crash with cinebench………sometime you will blue screen, somtimes you will just black screen and sometimes the bench will just crash ( dissapear, etc just shut down ) and windows will still be up………..the blue screen is NB vid/IMC memory related, the black screen is core clocks/cpu voltage related and the just crash/dissapear from desktop is temp related………….”

As I've said I've eliminated that, and if you look here... I'll drop my resolution as I don't want to post a 2048*1536.

http://i140.photobucket.com/albums/r35/Oj10101/stability.png

Note the speeds and voltages, as well as CPU usage and uptime. VDIMM is 2.092v for what it's worth. Over eight days at this speed with 100% CPU usage.