Computer Oddity - Advice Sought

I bought a new PC at the start of the year, and since the beginning, it’s exhibited some odd behavior. I’ve been working with the tech support of the place form which I bought it, but I’m growing more and more frustrated with them.

Here are the symptoms.

The big issue is that I get periodic blue-screen crashes. The Windows 7 event log identifies the type as a critical kernel power event, but the blue screen error is an IRQL_NOT_LESS_THAN_OR_EQUAL. These crashes only seem to occur when the computer is in active use, and then only once every two to three weeks.

Another oddity is that the computer refuses to remain off if I shut it down via software. A few seconds after going dark, it revs back up. I can shut it down via the power button on the case, though, and it’ll stick.

Lastly, the beeps. When the computer starts up, I get extra beeps. These beeps do not correspond to any beep codes I can find, and the computer POSTs and boots just fine. The number of beeps seems to vary depending on how many devices are plugged into the machine - if I strip it down to hard drive, video card, and a keyboard, I get the one, normal beep. With all my USB devices and optical drives and such attached, it’s four. I’ve also heard two, and three, depending on ‘load’ while attempting to troubleshoot.

The first thought their tech support had was that the battery in my UPS could be going back, causing the power losses, but the problem continued when I hooked up the PC to a straight surge protector. The thought after that was the power supply, and so they sent me a new power supply to swap in. That was earlier this week. Issues #2 and #3 remain, but I can’t tell if #1 is gone, due to its intermittent nature.

Still, issues #2 and #3 trouble me. The only plausible explanation I can see is a bad motherboard or BIOS. Does anyone have others ideas, or has anyone seen something similar?

It’s Windows 7 Pro, and the hardware is a MSI® X58 Pro-E Motherboard
with an Intel® Core™ i7 processor i7-920, quad 2.66GHz cores, 8MB Cache, 4.8 GT/sec.

one bump!

Are the bluescreens consistently in a particular component or module? Or all over the place? If the latter you have a hardware issue, which could be anything from overheating, to bad memory or CPU.

Given it’s still under warranty, with that variety of issues and failure to fix so far, I’d be demanding a replacement machine by now.

How would I know what component or module causes the bluescreens?

I’ve tried re-seating and spot testing the memory, it doesn’t seem to be the issue. I’ve run heat-monitoring software, that’s not the issue either.

I have no ability to demand a replacement - my option is to ship it into them for diagnosis and repair, but 1.) I have to pay for shipping since we’re out of the first 60 days and 2.) they require the hard drive in its original state, so they’re going to wipe my data. I have backups of the normal files, but reinstalling everything will be a pain in the ass.

Given the symptoms you describe, I’d be willing to bet you have a bad motherboard. Unfortunately the only way to test that is to swap out the motherboard for a new one.

My suspicions are in that area as well - I just wondered if it might be something fixable with a BIOS upgrade, or if it’s an actual hardware defect.

It doesn’t sound like the bios. It sounds like the bios is detecting an error within your motherboard. Maybe too much solder shorting a circuit that doesn’t occur unless you have something attached to that circuit. Heat would make it worse as well as what you were doing at the time you got the BSOD. If you kept track, you might find the pattern (heavy audio use at the same time might indicate a problem with the audio circuits, for example). Good Luck. Problems like yours are really had to narrow down.

The BSOD always occur while I’m using the computer, but beyond that…

The one I remember most clearly happened as I was browsing my Firefox bookmarks list.

It should get called out in the bluescreen. In this one it mentions SPCMDCON.SYS . In this one it’s in storport.sys . Failing all else, are the memory addresses called out always the same?

What actual temps did you get for CPU and GPU?

Huh. I wrote down the hexadecimal stops, but I didn’t spot the module. If it does it again, I’ll look for that.

I don’t remember the actual temps - I saved the log from monitoring them, sent it to the tech support guys, who didn’t seem to think that was the problem.

This is often indicative of a hardware error.

There should be some numbers and text associated with the error message. We need the exact full message. Without those I’m hesitant to diagnose further, but I will suggest you reseat the memory.

With regard to the powering on, I’m wondering if you have a short somewhere. This may also just possibly be the cause of your BSOD.

Yeah it is. You’ve got a marginal component that’s flaking out when it gets overheated, probably on the motherboard. The blue screen “IRQL_NOT_LESS_THAN_OR_EQUAL” error when the machine is not idling is a classic indicator. Crack open your case, point a box fan at it, then run your machine that way for a while to see if the problem recurs.

Or the motherboard could be very tightly seated and thermal expansion is causing something to short or disconnect.

This used to be a significant problem with Dell machines in the early / mid 1990s. The CPUs used to rise out of their sockets. Just slightly, of course, but just enough to cause problems. We just opened the cases and pushed the CPUs home.

You don’t usually get much time to jot down the on-screen details of a BSOD - is there anywhere within Windows that stores a log of these events and their causes? If not, why not?

The problem takes 2-3 weeks to manifest itself.

I was running my PC last night after returning home with my AC off for the whole weekend - the ambient room temperature was close to 90 degrees. No crash occurred.

I have reseated the memory. It didn’t help.

As for the numbers, I can provide some of them. The ‘stops’.

0x0000000A (0x0000000000000000, 0x0000000000000002, 0x0000000000000000,0xFFFFF80002CC97B6)

If I get another before I ship it off to the company, I’ll let you know.

If you reported the problems within the 60 days, and they were unresolved, then you should demand a replacement.

And the pain you are currently enduring isn’t sufficient already? Buy an external hard drive, clone your PC’s hard drive and ship the PC back to the manufacturer.

I do have an external drive. I just wonder about the viability of a cloned drive to restore everything. what free clone software would you recommend?

They’re willing to do the repair/replacement work. The quibble I have with them right now is who’s going to pay for shipping. I think they should. They disagree, because it is NOW more than 60 days after receipt of the system.