BSOD - ram/psu/mobo/videocard? - lots of testing done already

chops13

Distinguished
Oct 17, 2011
3
0
18,510
Problem:

- I am getting all sorts of BSOD errors ranging from PAGE Fault; Memory management, Stop 0x50, 0x07a, all sorts of stuff. They are never consistent.
- BSOD's at various times. Just after typing in password for logon, before the password screen comes up, in the middle of the night when nothing is happening
- Sometimes it just sits in a boot loop before getting to the logon screen
- I haven't made any significant hardware changes when this all began; it started with my rebooting my computer.

System:

Roughly 1 year old.

OS: Win 7 - 64
RAM: 8gb 4x2 (2 sets) Corsair XMS3 Classic 4GB

Case: Antec Three Hundred Gaming Case ATX 3/0/(6) 2xUSB Audio No PS
PSU: OCZ StealthXStream 2 Ultra-Quiet 500W Power Supply (OCZ500SXS2)
RAM: Corsair XMS3 Classic 4GB (2x2GB) DDR3 1600MHz CL9 DIMM, Optimized for Core i5, i7 and Core 2 (CMX4GX3M2A1600C9)
Mobo: Gigabyte GA-P55A-UD3 Socket 1156 Intel P55 Chipset Dual-Channel DDR3 2200/1600/1333/1066/800Mhz 2x PCI Express x16

CPU: Intel Core i5 760 Quad-Core Socket LGA1156, 2.8Ghz, 8MB L3 Cache, 45nm (Retail Boxed) (BX80605I5760)
Cooler: Hyper 212+ Cooler

Videocard: Asus EAH4350 SILENT/DI/512MD2(LP) ATI Radeon HD 4350 Chipset (600MHz ) 512MB (800Mhz) DDR2 Dual Display PCI-Express 2.0 Graphics Card

Case fans: (2) 140mm | (2) 120mm LED
Harddrives: 6
USB Soundcard: Cakewalk UA-101
Optical 1: DVDRW+

BIOS Settings

1. I've tried both Optimized Defaults and Fail/Safe defaults
2. I've tried hard settings as well.
a. Voltage for Memory Corsair XMS 3 - 1.64 or 1.5 @ 9/9/9/24
b. I've used the XMP profile setting
c. I've disable XMP and set my BCLK at 133 or 160 and set the multipliers for the CPU and Memory so that the memory is running at 1600mhz
d. I've also tried running the memory at 1333mhz
e. I can also run an overclock (3.2ghz; 1600mhx) with stability with Prime or Linx no problem for a few hours (cpu temp 58celsius (doesn't BSOD)
f. CPU cores run at 29 Celsius without overclock


Troubleshooting (all without overclock):

1. Update drivers / BIOS

a. Gigabyte F11 is current bios
b. update intel chipset, etc
c. current ATI driver installed

2. Memtest86 v4.0

a. Removed all RAM and tested each stick one at a time
b. Verified fault on one stick of "Pair A" RMA'd that pair
c. Both stick of pair B pass when ran on their own
d. Found faults on "Pair B" when ran as a pair in slots 1 and 3 as a pair
e. Received RMA'd new "Pair A" - ran Memtest 86 in slots 1 and 3 - memtest86 fails sometimes on a brand new pair of memory
f. I took one of the good sticks and ran it by itself in both slots 1 and 3 to make sure the MOBO slot wasn't faulty

3. Linx / Prime 95 complete no problem

4. I used extreme PSU calculator (http://extreme.outervision.com/psucalculatorlite.jsp) to identify voltage required. It gathered i would need about 430w max at any given time. My psu is 500w.

What now?

Here are some ideas/thoughts:

1. Replace the video card. I've had issues with my computer crashing when trying to load the ATI Catalyst suite, I think. Because about 6 months back the machine would crash as soon as I tried to launch a webpage (using chrome) that had flash content. The solution to this in the past has been to boot into safe mode and re-install the Catalyst suite. Recently I removed the Catalyst suite and just installed the ATI driver and have been using "ATI Tray Tools" for videocard config. These BSOD's I am getting now though don't seem to be related because my machine typically doesn't crash when loading ATI Traytools or Flash pages.

2. I have configured start-up services using CCleaner so I am not allowing a whole bunch of stupid stuff to load after logon.

3. Perhaps my PSU is failing? Maybe it is unable to supply voltage for everything even though the PSU calculator says it should be fine. Unplug the 2 fans in front of the case, unplug the harddrives except for the OS drive.

4. Casefans (3spd variable) and harddrives are daisy chained off of 2 molex lines using "Y" adapters, etc.out of the PSU. Perhaps this is a problem?

5. Brand New RAM - RMA'd "Pair A" is bad. It does fail memtest86 sometimes.

6. RAM "Pair B" - currently not in use but also tests bad sometimes
 

beenthere

Distinguished
It sounds like you're mixing RAM which often will not work, especially DDR3, even if you purchase multiple DIMMs of the same part number. It's always recommended to purchase ONE, tested, matched RAM kit so you know it will work without issues.

None of the new RAM you have should be testing bad at any time. You might try using one DIMM that tests good and running it for a few days to see if the problem still exists.
 

chops13

Distinguished
Oct 17, 2011
3
0
18,510
I am not mixing RAM.

I'm just using 1 of the matched pairs to test. Currently that happens to be the brand new "pair a". The thing I am at a loss with in regards to the brand new RAM is that it fails Memtest86 intermittently. Is it likely that I have had 3 pairs of bad RAM one of which is brand new? This leads me to believe I have a problem somewhere else.