Possible motherboard problem

[Thrall]

Hello everyone, recently I have built a new AM3 quad core rig for my family and I have had nothing but blue screens with it. They will pop up on login into windows about 80% of the time, but after that the system is stable. They include IRQL_not_less_equal_or_equal (0x0000000a), RDR_file_system (don't remember stop code), and memory_management (0x0000001a). I have tested the hard drive using seagate's free tool and windows error checking and nothing came up and the computer runs 3dMark06 fine for hours on end. However, the curious thing is that I can only get the machine to pass memtest86+ if I have the memory sticks in slot 1 and 4 or in any single-stick configuration. I tested each of the sticks individually and they're ok and I also tested each of the slots with individual sticks and they're good. I also upped the voltage on the memory but that didn't help them pass either. I suspect it is the mobo because I've had trouble running the memory in other configurations but then again I don't remember getting any bsods for a few days after I installed windows so maybe I just need to reinstall that. I've already reinstalled graphics drivers, updated the bios, and made sure windows is up to date. Below are my system's specs. Thanks for your help!

2 x 2gb OCZ Signature Seriess DDR3 1600 RAM
Gigabyte 5770 1gb GPU
Thermaltake TR2 500w PSU
Samsung SH-S223L DVD Burner
Hitachi 500gb 7200rpm HDD
Asus M4A785TD-M Mobo
AMD Ahtlon II x4 640
Windows 7 32

[sneekypeet]

did you set the voltage and timings for the memory to the specs they require?

[Swamp Monster]

It sounds like a IRQ conflict. Look at IRQ settings in BIOS. It never hurts to check.

[micropage7]

if all of hardware is OK, then check bios setting, check Ram setting, is that marked auto or re-setting according to your Ram clock, check the voltage.
Re check the irq address, does it have their own irq?
Have you try run Ram on lower speed?
Does it have sideport memory? If yes try set the clock according Ram clock

[Mussels]

those errors are likely memory errors. Make sure you manually set the voltages and timings required for that memory in the BIOS.

[Thrall]

I increased the voltage of the ram to 1.65v and set the clock speed to 1600mhz manually and this seems to have fixed it for a few hours so far. I think that the memory was unable to run at ddr3 spec latencies even with 1.65v so they sold it at looser latencies at a higher clock speed. I have to admit, I'm not a big fan of ocz products lately...

[Mussels]

Originally Posted by Thrall

I increased the voltage of the ram to 1.65v and set the clock speed to 1600mhz manually and this seems to have fixed it for a few hours so far. I think that the memory was unable to run at ddr3 spec latencies even with 1.65v so they sold it at looser latencies at a higher clock speed. I have to admit, I'm not a big fan of ocz products lately...

almost all 1600MHz DDR3 requires 1.65v


the latencies are supposed to be read from the memory itself, but if you buy 1.65v ram its gunna need 1.65v no matter what. any settings you manage to make it work that doesnt (say at 1333Mhz) is just luck.

[Thrall]

I'm still having the same problem and on "bad mornings" it will take 10 reboots to get into windows. If the memory isn't compatible with the board and/or if it is bad with memtest86+ show any errors?

I don't see any IRQ settings in the bios, but I've reset it a few times just incase a changed a value accidently. I've tried lower ram frequencies and higher ones and I've also kept the sideport memory disabled since I have a dedicated gpu.

[francis511]

Running stock or overclocked ?

[DanTheMan]

Make sure the motherboard battery is not resetting to defaults after sitting a while. You may have a bad memory slot since you stated that one stick configurations work well. I had similar problems with OCZ ram before, finally had to boost the voltage high enough to not cause the error. For testing I say use only one stick for about a week and see if any errors. If not remove that stick and try the other for a week for errors. My guess would be the bad memory slot.

[Thrall]

It's running at stock clocks currently. Wouldn't memtest86+ show any kind of memory related error? I've tried running just one stick and it doesn't help :\

[francis511]

It COULD be the psu..

[Thrall]

Originally Posted by francis511

It COULD be the psu..

It's a thermaltake so it should be pretty high quality....but you're totally right. Is there any way to test this without taking out the psu? Maybe I should check the voltage that I'm getting in the bios?

[Swamp Monster]

You could analyze dump files that appears after BSOD. They can tell you more about the problem.

[Thrall]

Originally Posted by Swamp Monster

You could analyze dump files that appears after BSOD. They can tell you more about the problem.

OK, I will start to compile them. Oddly enough the problem happens in safe mode as well and also when I try to uninstall a via package manager or something.