Computer BSOD or restarting once a week?

[ViperJPB]

Recently built a custom HTPC box. Work has been hectic so I have not had a lot of time to devote to this issue. However the box keeps restarting randomly. It is doing so at least once a week now. Initially I thought it was heat related being that the system never restarted in over a two week period when the system sat on the floor before being placed on the shelf in the media rack. I added two more fans and got a 2.5" drive to cut down the heat. The heat specs seem to be in the safe zone. Memtest runs clean, HDTune and Western Digital's Hard drive scans show all good.

Browsing the Event Viewer I have been trying to narrow it down. Thought it was a conflict with Microsoft Security Essentials and uninstalled it when I saw the, trouble starting "Microsoft Security Client" error a week ago. However it still crashes and gives that error.

Another error lists "The computer has rebooted from a bugcheck. The bugcheck was: 0x000000de (0x0000000000000002, 0xfffff8a00f5f7d50, 0xfffff8a00f5d7d51, 0x00000000378c98c0). A dump was saved in: C:\Windows\MEMORY.DMP. Report Id: 040213-15865-01."

The mentioned memory.dmp file is over 444MBs in size or I would post it here. Anyway to get a smaller dump?

Any ideas on how to further diagnose this problem? How can I get a .dmp file in a size uploadable for someone to browse over. Machine runs great otherwise and records 3 HD cable streams while streaming to mobile devices and other computers without batting an eye.

Specs:
CPU: Intel Ivy Bridge Core i3-3225
Mobo: Asus P8H77-I LGA 1155 Intel H77
Memory: 8GB
HD: 1TB Blue Western Digital 2.5" Sata 3

[FONT=helvetica][SIZE=23px] [/SIZE][SIZE=23px][/size][/FONT][SIZE=23px][/size][SIZE=23px]

Will post MiniDump tomorrow....[/size]

 

[Tmagic650]

Along with any minidumps, post your power supply specs

 

[ViperJPB]

Here are my Minidumps. I originally had a 250 watt PSU that came with my small form factor HTPC case. The reviews on it were pore and it made more noise than I liked. The system crashed on that setup too....but at the time I thought it was my 3.5" hard drive overheating, as it was over the recommended temps and wouldn't crash when opened and off the shelf. I have since changed the hard drive to a cooler 2.5" WD Blue and jumped to the usual silent HTPC external PSU option....a PicoPSU 150XT 150watt PSU. Here are the stats.

[FONT=Lucida Grande]PicoPSU-150-XT, 150W Power Kit[/FONT]
[FONT=Lucida Grande]- Tiniest DC-DC ATX PSU: picoPSU-150-XT
- Power adapter: AC-DC 12V, 12.5ASwitching Power Adapter (4pin DIN connection)
- 12V-ATX cable (4 pin power)
- Fits 24 pin mini-ITX boards
- Configurable wire harness
- Power cord[/FONT]
[FONT=Lucida Grande][FONT=Lucida Grande]- NOTE: Will not work with M300/M200 enclosures due to 4 pin din connection[/FONT]- [FONT=Lucida Grande]ROHS compliant version[/FONT][/FONT]

Should be plenty of power for these super low power parts. Thanks for any help you can provide. Been chasing my tail on this. Still think it may be a heat situation on the Mobo...but starting to wonder if it is software. The PSU 24pin connector has a lot of gear on it and is very close to the memory as well....possible heat issue there too. Though again it crashed before this new PSU.

 

[ViperJPB]

Here is a link to the PSU's sight. http://www.mini-box.com/picoPSU-150-XT-150W-Adapter-Power-Kit

 

[Tmagic650]

Okay, this looks like a memory problem now according to the minidumps. Did you use memtest86 for at least 7 passes with no errors. In case you didn't use it correctly, here is a link:
http://www.memtest.org/

 

[ViperJPB]

Hmm. Thx for looking at those for me. Strange though. I ran Memtest86 with the latest self booting version for a USB stick, from there site, fresh after installing the sticks. Ran through about 5 passes and no errors. Should I run it again tonight and let it go past 7? This ram is off the Motherboards recommended list as well. Still makes me wonder if the stick closest to the transistors of the PSU connection is getting too hot and causing memory errors...

Anything else suspicious on there? Thx again.

 

[ViperJPB]

CPU gets up to 47C....which is fine.
Motherboard gets up to 48C or 49C....Asus says at 60C it would give a warning...so close but still 10 degrees off.
Hard drive is in the low to high 30s which is great.

The mobo being as hot as it is coupled with the PSU transistors being as close as they are to the ram makes me think the memory issues are due to heat. Especially since it never crashes when out in the open and not on the crammed cabinet shelf.

 

[Tmagic650]

"Especially since it never crashes when out in the open and not on the crammed cabinet shelf"...
This about cinches that it is a heat problem with the memory. I have built a few of these boxes, and your temps don't seem too high. What brand and speed are the memory sticks? Do you have 2-4GB sticks installed? Have you tried running with just 1 stick at a time?

 

[ViperJPB]

The Memory is Kingston HyperX Dual Channel DDR3 1600, 2x4GB. Typical Blue heat-sinks on. Had to set the timings in the bios to get them at their rated speed.... don't recall what the timings were without looking them up, but they weren't anything crazy. I have not tried one stick. I suppose I could do that for a few weeks and see how it goes. I could move the back fan into the case and jimmy rig it to blow across the memory. I have yet to run memtest86+ 7 times, but I will this weekend. Problem is if it's heat related I'm not sure what to do as I already think the thing is to loud at standard fan modes with Noctua's. Thing is it crashed before this power supply went in as well....so the crazy transistor 24pin connector of this external PSU that is so close to the one stick wasn't present then.