[SOLVED] New Build Issues - Random Reboots

[adrianf74]

Hello Everyone:

I've just built my first system in about 7 years and feel like a bit of a noob so bear with with me here. Thermal grease was used (as supplied by Scythe) and was applied in a "thin layer" on the entire CPU chip. CPU chip had "no damage" and was being used in a previous build without ever being abused/overclocked.

The parts that have been used are as follows - everything is brand new except for the CPU which is about three months old (and was verified working):

Gigabyte GA-H55M-D2H Motherboard (rev 1.4, Bios F3 - latest, reflashed for "fun")
Intel i3 540 CPU
Scythe Big Shuriken CPU Heatsink/Cooler
Seagate Barradcuda 7200.12 SATA HDD
ASUS 24x SATA DVD Writer
OCZ Gold Edition Low Voltage 8GB (2 x 4GB) Kit (OCZ3G1333LV8GK)
OCZ ModXStream Pro 500W Modular PSU
Silverstone GD05 HTPC Case

This machine, when I get it working, will be used as a HTPC and I'm using Intel onboard audio and video via HDMI as it serves its purpose (I know some people will say buy a standalone card but because of the cooling solution, the PCI-E x16 slot seems too close to the edge of the Shuriken).

The first time I installed Windows 7 Ultimate, I forgot to change SATA to AHCI in the BIOS so I had to re-install Windows because it became very unhappy (endless reboot loop) when I changed AHCI in the BIOS after the fact. I learned there are other ways to do this but I didn't know this until after I'd already started the re-install. It was done off a USB key so it wasn't too painful.

I have two issues that I'm experiencing:

The first is when the machine is in "start up" mode, I'm misisng a lot of text around the edges of the screen (viewing via HDMI). On any other computer (even using HDMI out, I've seen this content). When I'm in the BIOS, I can see the options but can't see the "outer border" or the BIOS information along the top of the screen. I also can't use my USB keyboard or mouse as they're not always being grabbed on boot up - less of an issue for me here.

The main issue is this:

Windows is rebooting randomly when doing basically next to nothing (or nothing) and is not displaying any BSOD; it just goes "poof" and reboots like somebody hit the reset button. On one reboot, it happened during a WEI test. On another, it was while in FireFox. On a third, it was while it was sitting doing nothing while I watched TV in the next room. Yes, I have installed all of the drivers from the supplied Gigabyte DVD. I think the next thing to do here is to look to see if any of the drivers have since been updated.

I thought maybe the RAM was flakey. So I download MemTest86+ and created the bootable USB version. Left it run overnight while sleeping, woke up, and saw that it had been running for 6h09m without a single error. All four passes had passed cleanly. This makes me a little happier. Leads me to think that the PSU and memory are okay. To see six hours without a reboot makes me happy because I'm lucky to see an hour in Windows.

Now, I haven't done anything in BIOS except have it learn its "best settings" as I've read on a couple of theads. I've left my computer on at home today with a firefox splash screen and will see what it's like in about 9 hours when I get home. I have a feeling it will be gone due to a random reboot when I get back to it.

I've used both PC Health in BIOS, the supplied Gigabyte Utility and CoreTemp to check my temperatures. Under "light use," my CPU temp is hovering around 21-23C and the system temperature is hovering around 30-31C. I'm *VERY* happy with these temperatures.

Does anybody have any input here? If there is additional data required, I'd be glad to provide it.

Thanks. I just want to get my system up and running.

***

I have posted this to the OCZ forums already to see if they can help me here. I've also run MemTest86+ again (for almost 11 hours and 8 passes, there are *NO ERRORS* being reported. I'm leaving it run at home all day (while at work) just to satisfy my curiosity.

One of the issues I've noted is that the ram is spec'd at 1.65V/DDR3 10666. My motherboard only allows for 1.64V (BLACK) or 1.66V (PURPLE).

Does anybody have any ideas on this one? :(

**EDITED TO ADD [SOLVED]: The OCZ RAM was the culprit (1 of the sticks is questionable).

[KaptCrunch]

1)what moniter you viewing from ? PC or TV

2)sounds my have shortting going on under the MB or faulty start/reset switches. have tried 1 mem stick @ a time yet

note: Modular PSU's are pita when they lose the connections within plug intermittinly. use a safty pin to spread male pins and squeaze the female with tweezers

[adrianf74]

Originally Posted by KaptCrunch

1)what moniter you viewing from ? PC or TV

2)sounds my have shortting going on under the MB or faulty start/reset switches. have tried 1 mem stick @ a time yet

note: Modular PSU's are pita when they lose the connections within plug intermittinly. use a safty pin to spread male pins and squeaze the female with tweezers

Thanks for the reply.

As this will be a HTPC (home theatre pc) I'm using my LCD television as a display. Every other machine I've ever hooked up to the TV (Acer Revo, Acer Netbook, Acer AX1800 and AX3200) have always displayed the BIOS/POST screen without issue. I'm guessing I could always run a VGA output to the TV as well for testing purposes.

As far as shorting going on under the board, there is nothing being run "under" the board and I have ensured that all of my standoffs that are installed are the only ones being used by the motherboard to ensure nothing 'touches' the active board to cause shorts. I'll triple check this when I get home tonight and pull up the board to ensure the push-pins on the heatsink are 100% connected just in case something dislodged and a thermal break is being made between the CPU and heatsink. The start and reset switches are quite heavy to press "in" to activate on the case so I'm at a loss with that as well.

EDIT: On second thought, wouldn't MemTest86+ reboot (run from DOS) if there was a short? It has run for 6 1/2 hours and an additional 10 1/2 with 4+ and 8+ passes without error. This leads me to think it's something bigger than that.

I'll check the Modular connections as well when I get in. Somehow, I think it's a timing issue with the memory itself.

[stoanee]

Set up your rig so that when it reboots it shows the error code. Once you have this information you should be able to narrow down the problem.

Disable UAC. Control Panel > System And Maintenance > System > Advanced System Settings > Advanced tab > Under Startup And Recovery > Click Settings > Clear the Automatically Restart check box > click OK. Enable UAC.

[adrianf74]

Originally Posted by stoanee

Set up your rig so that when it reboots it shows the error code. Once you have this information you should be able to narrow down the problem.

Disable UAC. Control Panel > System And Maintenance > System > Advanced System Settings > Advanced tab > Under Startup And Recovery > Click Settings > Clear the Automatically Restart check box > click OK. Enable UAC.

Thanks Stoanee. I'll try this when I get in tonight. I'm also gonna do a full re-install of Windows 7 for the third time just in case.

For the record, I've got a very sharp eye (my job demands it) and I've seen *NO* flash of a BSoD on these reboots; it's like someone just hit the reset button.

[KaptCrunch]

check the 8 pci-e and 8 cpu connections

[adrianf74]

Originally Posted by KaptCrunch

check the 8 pci-e and 8 cpu connections

Not using PCI-E power supply connections (no external video card as it's not required in the build). Those two cables are sitting in the little pouch provided by OCZ. Will check the other ones.

[Shadowmeph]

the first thing I see is that you installed your drivers and other things from the dvd , you should get all of your drivers from the hardware websites , that way you know they are up to date.
the other thing I noticed is that you said you ran memtest over night, well this tells me that the problems is very likely caused from either the operating system or something in the operating system, what I would do is turn off everything except for the basic things in your OS, then one by one turn them back on until you get a reboot.
I am wondering if you have a8utomatiuc updates set to install all updates automatically which would sometimes reboot your system, the thing is though that it shouldn't reboot to often.
have you looked in the event viewer to see if there are any errors, if you are not sure of there that is it can be found in the Control Panel, clicking System and Maintenance, clicking Administrative Tools, and then double-clicking Event Viewer.

[adrianf74]

@Shadowmeph:

Thanks for your insight.

I have downloaded all of the appropriate drivers for my next reinstall of windows. I'm definitely led to believe that the issues are with Windows itself and/or a particular service. I haven't turned off services in ages so if you could kindly point me in the right direction, that'd be helpful. I have automatic updates set to always download but allow me to ALWAYS CHOOSE when/if to install. So that has nothing doing on it.

I'll take a look at Event Viewer when I get home as well. Again, it's been forever since I've built something so I feel very rusty.

[adrianf74]

@ All:

OCZ Support feels that my voltages and settings are "good" based on what I've provided. The only "change" they suggested was to bump up my QPI/Vtt into the pink between 1.25V and 1.35V because they feel that my IMC may need help for the 8GB of memory on the two slots. I also bumped up the DRAM Voltage to 1.66V; please correct me if I'm wrong in thinking that if my DRAM Voltage at 1.64V = 1.632V, then 1.66V netted me 1.652. So far, since I rebooted my system it's been up for over 35 minutes without a reboot -- NEVERMIND, it just tanked. So I changed back to the 1.64V setting and 1.29V now for the QPI/Vtt setting. I'll try 1.66V again if somebody feels it's a "better choice."

@ KaptCrunch:

Considering I just ran MemTest86+ for the last 21h 42m 16s with 16 successful passes and ZERO errors, I somehow doubt it has to do with the PSU or the standoffs. I did check and there are no standoffs that are touch the board other than where they are screwed down.

@ Stoanee:

I've disabled auto restart, however, when the last restart ocurred there was *NO* BSoD; it just restarted for no known reason. I stand corrected. I just had a BSoD which read "PAGE_FAULT_IN_NONPAGED AREA." I had installed Norton 360 but it corrupted on me after a crash (it won't actually run) so I'm debating doing a clean windows install with a format when it does it.

@ Shadowmeph:

I just went into the event viewer .. OH WAIT (back to @Stoanee - just BSoD'd). Back in event viewer and there are a crapload of "Critical" "Kernal-Power" (source) 41 (event ID) (63) task. I just googled this and it's talking about two different audio drivers being installed and conflicting. When I check my device manager, I've got Realtek HD Audio and Intel(R) Display Audio. When I look at playback devices, I'm seeling that the TV/Receiver are being played out under the Intel(R) Display Audio and that the Speakers and Realtek Digital (S/PDIF) are the other two. I'm not planning on using the S/PDIF so I'm wondering if I should disable or uninstall something here?

@ All:

I think a good starting point here would be to gas the system one more time (as I said above) and then disable auto reboots again. I'm likely going to install just the bare minimum and avoid installing the Realtek drivers as I'm sure the Intel(R) Display Audio will be native to windows itself.

Thanks for the help so far... maybe I'll get this nailed down sooner than later.

EDIT @ 7:10pm:

Same thing just happened with the Same Error / BSoD on a clean install after first installation. Got Error 0x00000050. Now, rebooted, and I got Error 0x00000019 (BAD_POOL_HEADER).