PC Build issue

[Amad]

I'm building a new PC, here are the specs..

Core 2 Quad Processor
ASUS P5Q Pro Turbo LGA 775 Intel P45 ATX Intel Motherboard
4Gig DDR2 Ram
Nvidia 8600 GT Vid Card

Here's my issue..

Hooked everything up today... quadruple checked the mounts, all the cables and everything to do with the Mobo and processor. The power supply, which is a 650 watt and came recommended by newegg, is also hooked up right. When I go to power it up, the green power light on the mobo itself turns on and stays solid. The fans and case LED lights kick in, but after half a second everything shuts off except that green light. No matter what I try, removing peripherals like ram/cards, anything except the bare bones, it never does anything beyond the small half second fan start then nothing. It doesn't even do the system beeps that are associated with troubleshooting before a comp boots to bios.

any clues as to what's causing this? The internet seems to be full of stumped people on this one.

Thanks

[Xeon18]

I would say try sending the power supply back to new egg. cause if it is all doing the mini power boot and then not turning on period it just sounds like there is not enough power to go around. Not saying the power supply in not enough just that it may be a bad piece. So I would contact new egg and get a new one shipped to you. post that I can ask some techie friends and get back to you.

[alsoknownasOTiS]

yea to me it sounds like there isn't even power, but the thing is..a 650w PSU is more than sufficient for what you have going on there. And the PSU is clearly working since its booting things up, makes me thing it might be the mobo I'd RMA that and get a new one...and if you have another PC, try connecting the PSU to it and testing to make sure its not that...

good luck

[rburke]

try this first
Are you even getting a post, can you see it do the ram check and display cpu speed. What happens very very often is this. The default bios settings are reading your cpu beyond its rated speed. When it tried to post it is essentially pre-overclocked. what then happens is it will just power down without doing anything. Try to reset the CMOS jumper, should say how in the mobo instructions. I would check that first because i had the same issue when setting up my i7 its just an odd thing that happens. After the reset it should boot if that is the issue. Its actually very likely that could happen. Also besides just cpu speeds it could be reading the ram speeds wrong in which case you would get the same results.
if its not that i agree sounds like a psu issue. I had a similar situation where a psu was not giving adequate power and would shut down after a few seconds. a 650 is mor ethan enough but often psu run under their rated levels. Try it on older parts if you get the same issue rma it. If it was a cable issue it would either not start at all or throw an error message. Often people forget to plug in the gpu power connector, whihc will ussualy just throw an error.

[Amad]

Awesome these are all good approaches, most of which I've encountered in troubleshooting this stuff over the last day. Thanks for all the suggestions. Maybe I can bug you for a bit more...

I've done the following...

Tested the PS on a working computer - the PS is fine.

Taken the Mobo out of the case and worked with it on cardboard to avoid grounding issues - still busted.

Removed any peripheral ram, cards, everything including a test with and without the cpu in the mobo - still nothing.

Done the CMOST reset a few times, to no avail.

I'm now hearing that it's a DoA board, so I'm sending it back to newegg. I dont' ship til tomorrow, so if there's anything you guys can think of before I do that, it'd be awesome so I can avoid the shipping hassles. Thanks again, you guys have been a huge help.

[rburke]

only thing i could think would be dead cmos battery, but ive also had 2 DOA mobos before so just send it back, It sucks that that happens but newegg might cross ship for you. Its actually very common to get a doa board unfortunately and its even more common on that generation of boards there was more 775 socket mobos made in the last 4-5 years than can be accounted for.

[Amad]

Yeah the guy on the phone mentioned the battery to me as well. I asked if it was his opinion that that would be the case and he said 'extremely unlikely' that it would cause it. Instead of spending 2 hours reassembling and testing, only to probably have to take it apart again, I just ran to UPS and sent it back to newegg. They won't cross-ship it (believe me I asked), but they are gonna next day air the new mobo after they receive this one. I must say, I've learned more about hardware this last 24 hours than I ever knew on a hands on level... really good learning experience.

Ryan, that does put me a little behind on our project. Barring some amazing turnaround, I'll be out of commission for at least another 4 days while I try to get this whole thing going.

Thanks again all for the advice. I might have more when the new mobo comes, but I hope not.

[rburke]

yeah ill be out of commission as well, online classes wrap up on the 5th meaning i got a ton of makeup to do, and then i got a bachelor party to attend on the following weekend so that should set me back 3 days or so if im lucky lol. Then back to work.

[matthewparis]

Hehe, I don't think it'll help any but even after checking my cables several times during my first build I still managed to miss plugging in the CPU power cable. *shrug* At least my case the plug on the mobo was obscured by the PSU so i missed it for a while.

It created the same situation your describing, and since when I had asked online no one mentioned that I figured it as worth point out hehe for the sake of a through option :)

[rburke]

i gotta admit ive built 15 or so different pcs and rebuilt my existing ones many many times, and i still manage to forget to plug something in. Even when i know its there already. ill just sit there starring at it for a while before i realize what is wrong. it was a lot worse back in the day before psu's where nice and organized, you know back when there was big fat IDE ribbon cables, and your network card was not integrated into the board.

[Amad]

Haha yeah I did the same thing... staring at the board and connections... but ultimately there wasn't anything that wasn't connected. I wish it was that easy this time :/

Hey if anyone wants to drive out to Davis, I'll buy you lunch and you can help me build this thing! No takers? I didn't think so, driving 4 goddamn hours both ways really sucks :)

[rburke]

yeah i loathe that drive, i go to sac just about every weekend, and i hate the 80 i just drive late night now like 11pm or later. I actually used to enjoy it when i lived in east bay because i could drive the 160, its a pretty drive, and a fun road to boot. But the 80 thats like a lobotomy, i think the only thing worse is the 50 up to tahoe, that takes all the enjoyment out of the trip for me.

[Amad]

Meh dude, I won't even go to tahoe unless it's like... free condo stay, free gambling and free drinks.

They really should have made BART go from the bay through Sac and up to Tahoe.

[rburke]

lol your still up too, i just got up, time to go unload a truckload of shoes, cant wait to get a job that doesnt suck.

[Amad]

Lol damn dude. I feel ya, I just got back from work myself. Good news though, I found a bill for 3 grand for the upcoming semester!

Maybe this project will get bought out, every single person on the team will randomly give us all the credit, and Valve will start paying royalties before September.

[matthewparis]

*Ears perk up* What? Valve? Hehe.

[Amad]

I'm writing from the new machine now - turned out that the old power supply had a short in it that would randomly screw things up depending on the angle it was positioned. Anyway...

Now, I'm running into something disturbing.

Running vista 64bit. Everything will work fine for a while, then I'll get randomly blue screened. It's not up long enough to see what it says. When I run the Windows Diagnostic Tool from my boot menu, it runs 2 passes of the memory test, which says that I have 'hardware errors'. It won't report beyond that.

I've updated windows fully, including all of my display, gpu and system drivers. Thermostat indicates that my processor is consistently at 30 degrees celcius.

I've tried reseating the peripherals... gpu/ram and even reset the cpu heatsink. No better. I DO remember having something at boot on my old machine, with the same video card, that was an alert telling me that my card was running in '1x mode'. It's an 8600 GT. No such error now, but I'm wondering if that error on the old machine was an indicator that it had issues... but since I was running a lite 32 bit XP on it, it just wasn't ever enough to crash it like this. That's my speculation.

Another symptom... my mouse/keyboard will momentarily freeze, or programs will start ending unexpectedly within windows right around the time I'm about to get blue screened. Seems to be the only common thread to when it happens.

Any thoughts? Pretty lost at this point with the vague nature of the 'warnings'. Thanks.


***Update***

In the ddr2 ram slots 1 and 3 (as prescribed by the mobo manual), slot 3 fails hardware tests. Slot 4 also fails hardware tests in the windows diagnostic.

Slots 1 and 2 will go through the diagnostic fine, windows will recognize 4 gig of ram (the correct amount), but memtest 6.5+ will give me errors, saying that memory is not copying properly through the system. It's more stable now... not crashing, but it's definitely still having some issues.

In other words, the hardware all PASSES the windows diagnostic tool that can be performed out of the bios. It does not fully pass the third party program Memtest, however. No other symptoms now outside of that, feeling stable.

[rburke]

what brand memory is it and what product line. If you are getting errors from mem test there is a good chance the memory is causing that BSOD issue, most times ive had an issue with that was with a value ram or just low quality. All ram is not created equal when a batch is made they test the chips to see what frequency they can run at then they market the best ones as a premium line and then make value ram from the rest, and often the value ram modules can be faulty. Also sometimes ram is just bad even if it comes from a good product line.
the way it sounds that programs are crashing out before a blue screen would lead me to believe that the memory is having trouble running at its rated speeds, a good test would be to downclock the memory and see if stability improves. You could either lower the mem multiplier to lower the frequency or raise the memory latency settings which is a bit trickier. You could also lower the FSB frequency to lower the mem frequency but that would also lower the cpu frequency so try lowering the mem multiplier first. That would be an easy way to test to see if the RAM is having an issue at its current speeds.

also checking the manufacturers specs for the RAM will reveal the proper timings and frequencies to be used so you can make sure the bios has set this properly as it often does not set timings correctly using the auto settings. And that is actually very common with that generations of computers.

[Amad]

It's Corsair DDR2 800, 2 2 gig sticks. I tried changing up to 667, as you described and took my mobo from what it called 'turbo' to 'standard'. Apparently this improves stability after overclocking (which I didn't do), so I'm not sure if that made a difference. Those changes caused immediate failures in the bios test, after about 8 percent it just kinda laughed at me.

I've gone through and diagnosed with 1 of the sticks in slot 1 on of the ram seating. It PASSED both diagnostics, including memtest. I just now went through and did the same test to the other stick. It also passed. So, both pass tests in slot one individually.

Reading up on this stuff some more... trying an error test after clocking the ram up to 1066. Shrug.

What are your thoughts at this point, Ryan? Bad mobo again?

[rburke]

from the review i read on the mobo i would prob say it just has compatibility issues which is a common problem. not only that given the reviews i read it could just crap out on you one day and you'd be SOL so you might be better off returning it while you still can. ram can also be a compatibility issue sometimes certain ram and certain mobo's just dont like each other. If you can link the ram so i can see its exact specs i can maybe give you ab it more info on that but the ram is prob doing okay. Also always run the stuff on standard, turbo will attempt to reduce the latency settings and that will cause all sorts of issues, the stock bios settings can be all fucked up sometimes and require tweaking, BSOD is no good and you should not be getting that issue unless there is some sort of hardware error associated with either the memory or the mobo. I still recommend getting that gigabyte board. Asus does have a good track record but unfortunately the last year or two they have been going down in quality drastically, not sure why but they are no longer the leader or the name you look for when you want quality. its odd that the memory would not run downclocked but that could be another issue all together like certain things need to be set properly in the bios. it seems odd that he ram will pass individually but not in dual channel mode so it could be faulty ram, I cant really say anything for sure unless i had hands on with the computer. And a short in the PSU is no good try to resolve that asap even if it means using some electrical tape, but be careful those things can kill you even unplugged so don't go sticking your finger on too much stuff internally. Ill leave steam on hit me up if you got any questions or anything.

[Amad]

I did forget to mention 2 things. I ordered a gigabyte ga-ep45-ud3p, which is what's in the machine now. Also, the new PSU is working fine, has been good.

So I went ahead and left it on default turbo and overclocked the ram to 1066. I didn't read anything that indicated I should do this, but it did 8 hours of memory tests with 0 errors, which is unbelievably good from what I can tell. The voltage on the ram is 1.8. I noticed that that is correct on the bios, but the falloff voltage, or the voltage it is taking to downcycle is .9. I wasn't sure if that was an issue - if I fail this round of error tests I'm gonna try it.

What I've done now, since the thing worked at 1066, is testing it again at the standard 800.

And no I don't intend to break into the psu. Screw that :)

Would there be a conceivable reason that overclocking the ram would work on this board? Seems to me that since the board supports 675/800/1066 that it wouldn't make a damn bit of difference.

[rburke]

if the ram handled that without any fails then run it like that by all means. OVerclocking ram is just one of many things you can do to improve system performance, granted teh ram can handle it and it did so that good. Although i will admit that often what you set in the bios can sometimes be deceiving, so you can use that cpuz program i linked before to see what its actually running at in windows, which tends to be more accurate. but congrats i hope you don't get any more crashes.

[Amad]

CPU-Z is reporting all systems go... dual channel symmetric 4 gig ram. However, memory tests are going off left and right with errors. System stability is so so. Ran Maya 2k9x64 with a heavy project and started abusing with smooths and renders. One crash out of 4 sessions like that.

Right now it's on 'Turbo', the ram is at 800 with 1.9 volts per stick (minimum is 1.8).