I started getting blue screens. Most were memory management, bad pool headers, and irq not less or equal. I started troubleshooting the memory. Ran memtest for 7 passes on each of my 4 sticks. One showed bad and the other three were fine. After leaving out the bad stick I was able to boot into windows and run just fine for 2 days. Then started getting blue screens for every start-up. These were mostly ntfs.sys errors. Disconnected all but one HDD and ran chkdsk on it 2x with no errors. Formatted the drive and while trying to reinstall windows got a bsod. More memory management bsods. Ran memtest again and got inconsistent results. Same stick would show errors in the second pass then could run for 7 passes without issue. Memtest started freezing and messing up stopping the scan. Trying to boot to a live cd and got kernal panics. It will post every time without fail. I dont have a second system to interchange parts with. PSU is brand new and working fine.
Is there anyway to determine if the issue is with the mobo or cpu? What further troubleshooting can I do?
CPU: AMD 1075t
mobo: asus m4a87td/usb3
memory: corsair 4x4g 1600 defaulting to 1333, timings correctly defaulting to 9-9-9-24
First check if all the power cables are attached good in the board.
If the DVD was Vista 64 first edition is having problem installing it with more than 2 Gbytes ram, the disk driver is hiting errors during installation and the errors reported only in the events! I had to slipstream sp1, the 4 GBytes hot fix in winpe image and building a new dvd just to be able to install without removing 3 sticks!!
Remove your ram except 1 stick, try to get into the bios and see the system temperatures cpu, system if you don’t get a problem while you are in bios for some time it may be software related problem (windows). If everything looks ok proceed format, installing windows with only 2 Gbytes ram
A PSU problem can create similar issues only if the 12 Volts trail falls below 9 volts but you need a digital polymeter to see that.
Oh I forgot I always use speedfan 4.45 to see the speed of the fans on board can help me detect a voltage drop in the motherboard if their speed drops to low.
In fact I have only 140x140 of NH-D14 connected to the power fan connector and everything else to the PSU spins usually at 1250 rpm if for some reason I see 1240 rpm then I know the power delivery is not right on the board! I check with the digital polymeter my PSU of a drop and I usually find it fully running! I also have UPS that normalizes the drops and high peeks no way to get problem from the main power to my PSU anyway.. Bottom line this way you can detect if something wrong at your motherboard over the years and the heat overload can lead to capacitors malfunction and from there to what you call my motherboard is starting to die….
Selecting a motherbord with as many as you can find solid japanise capasitor (fully operational at 105 C) may help bypass the problems form overheating, I hope I’ve been helpful.