Sign in with
Sign up | Sign in
Your question

Intel DX58SO rebooting and BSOD with Win7 Ultimate x64

Last response: in Motherboards
Share
July 10, 2012 5:17:09 PM

This is the configuration I am having problems with:

Motherboard: Intel DX58SO
CPU: Intel i7-920 (2.67GHz, no overclocking) + CPU fan
PSU: Cooler Master 620W (has fan)
Case: Cooler Master PC Case with 3 fans (bottom, back & left side)
Video: nVidia GTX285 (1GB dedicated video memory)
Hard Disks: 2x (two) 1 TB plus 1x 750GB eSATA
DVD: 2x DVD rewritters
Other: Multicard reader

The system originally ran (for a few weeks) Windows Vista Home Premium x64 and was then upgraded to Windows 7 Home Premium x64. It ran fine for about 1+ year. Then sporadically I got BIOS boot errors saying it failed to boot but after pressing Y it booted fine into Windows.

Naturally as it happens with all Microsoft Windows OSes, at some point it just became slower to boot as programs were added (something I never experienced with Linux).

Then over the 1+ year mark with the boot slowness (nothing serious but definitely slower than the clean install) I started getting sudden Blue Screens of Death (BSOD). In March 2012 I decided it was time to reinstall everything from scratch but kept postponing it until a few weeks ago. I also cleaned up all the fans as they had accumulated way too much dust.

I reformatted both 1 TB hard disks and performed a *clean* install, this time Windows 7 ULTIMATE x64 SP1. Installed all the critical updates that came up with Windows Update. Then it started developing sudden reboots, sometimes a BSOD, other times it simply went black and started rebooting all by itself.

To make things worst, after the 3 weeks I have spent reformatting, reinstalling and reconfiguring *everything* things have become worse. Sometimes it simply refuses to boot:

- Sometimes I goes into the "Starting Windows" and then I get a BSOD
- Other times I get the BIOS boot messages multiple times and goes on forever without reaching Windows
- On other ocassions it attempts to boot several times but no errors are given by the BIOS. Then the system simply leaves all the fans running and the video card powered but no other activity such as HD access, a brain dead system.

Yesterday it booted okay but gave me a bunch of sprites on the monitor, I thought my monitor was damaged but after than it booted fine and I used it without issues for more than 8 hours.

Today it booted multiple times but never managed to get into Windows even after "Repair startup". I tried booting Ubuntu Linux from CD (Boot Menu) and it also restarted multiple times until *finally* it booted into the Live Ubuntu DVD which I am using right now for a few hours without issue.

Prior to all this hell I ran the crippled Windows Memory Diagnostic that only tests 4GB (I have 6GB). No errors given after several passes. I swapped the 2 & 3 DIMM slots and ran the test again, no problems detected.

Then I also used the MemTest86 memory test suite (Extended version) that comes with Ubuntu Linux, that one does tests *all* 6GB. It ran the whole pass and did not detect any errors. Then I booted again and tried the same test and started giving errors all over the place. I stopped the test and restarted it and let it run for several hours, no errors found.

I am going crazy with this, I need my system for both development and flight simulation but it is *very* annoying when it refuses to boot or get a rando blue/black screen of death.

I am pretty sure that I can now stop Ubuntu Linux and it would boot just fine in Windows. I don't understand what's going on with this deviled system.
a c 466 V Motherboard
July 10, 2012 5:44:47 PM

i seen it with some os and bad parts like ram..some os are more ram friendly so that if there a bad stick of ram...sometimes they never use it..and run a lot better then windows. i would start off by the simple stuff. make sure that you have all your mb bios updates..some of them are for cpu code and others for ram fixes. your mb has a few bug fixes. i would then run cpu-z and see what the speed the ram is running and it voltage and timing. use the spd tab in cpu-z and see that all the sticks are the same...voltage and speed. with the dimm info email intel and the ram vendor to see if the ram be qal in the mb. some ram will work fine some wont..sometime with mb you have to set the dimm up by hand if there 1.65 ram or turn on the xmp profile to get them to run at max speed. if a dimm is real bad..dead...you may not see that till you swap it in by itself or with one other dimm.
start with two dimms in the main slot and see if the bsod get worse or are still there..if there still there swap one dimm. if they still happen or it fails memtest and the vendor say the ram should work...try picking up a good name brand 1066 ram kit or see if someone has another set of dimms you can try. if the error stop the you know it was an issue with the memory and your mb. if with another set of dimms you get no change..download a program like open hardware monitor and look at your power supply voltage. a power supply with a bad 12v leg or a lot of ac ripple can cause bsod that look like memory errors. also look at the windows error log file..see if when and what is crashing..look up to see what the error code is..it may point to a bad driver or the video card.
m
0
l
July 10, 2012 7:13:49 PM

I didn't build the PC myself, I bought it from a renowned european vendor where one could select from an approved amount of choices so they should have checked.

Anyway, when I checked a week ago both physically and with a program, all three DIMMs where exactly the same, at least in model and capabilities (those 3 packs). Following your suggestions I will verify that their voltages are the same. I just downloaded the Cpu-Z and Open Hardware Monitor as well as Intel's Desktop Monitor.

Still have to upgrade my late 2009 BIOS and hope it doesn't crap out in the process. Hopefull that fixes everything because I am going desperate.

Will post any findings. I have not yet received my TH confirmation email so I hope I can login to my TH account after I close this session...
m
0
l
Related resources
July 10, 2012 8:25:59 PM

Apparently this booting problems happens only on a cold (but on the tropics it is hardly called "cold") boot. After many times once it boots, it boots. That has been the case so far.

CPU-Z Indicates as I indicated earlier:
- A triple memory channel where all 3 memory DIMMs are the same part number: OCZ3P1333LV2G
- All 3 DIMMs have exactly the same voltage: 1.350V
- All 3 DIMMs have exactly the same operational settings
- tCL: 7 tRCD: 7 tRP: 7 tRAS 20 tRFC: 59 CR: 2T

Mainboard:
- Intel DX58SO chipset X58 rev 13
- Southbridge Intel 82801JR (ICH10R) rev 00
- LPCIO: NS PC8374L
- BIOS SOX5810J.86A.4196.2009.0715.1958 (15 July 2009)
- Graphics I/F PCI-Express x16
CPU:
- CPU i7-920 Max TDP: 130W (is this the power consumption?)
- Core voltage: 0.879V
- Multiplier x12
OS:
- Windows 7 (6.1) ultimate edition 64-bit SP1 Build 7601
- DirectX 11.0


Open Hardware Monitor reports the following:

Bus speed: 133 MHz
CPU Cores 1600MHz
CPU Core Temperatures 1..4: ranging from 68 to 94 Celsius but on average 73 Celsius
Hard Disk temperatures (1TB each): 52 Celsius
eSATA Hard Disk temperature (750GB): 54 Celsius

Motherboard BIOS Voltage Readings:
V12.0 12.000 V
V5.0 4.869 V
V3.3 3.17 V
V1.1 1.054 V
Vrcp 1.078 V
Voltage Regulator temp: 61 C
Memory Voltage: 1.350 V
Motherboard ambient temperature: 59 C
Aux Fan Speed: 0900
Front Fan Speed: 0960
Rear Fan Speed: 0960
CPU Fan Speed: 3120
m
0
l
a c 466 V Motherboard
July 11, 2012 1:13:04 PM

- CPU i7-920 Max TDP: 130W (is this the power consumption?)
it intel thermal spec and max power that a cpu of that type could use if it was put under a stress test like prime 95.
the two things so far that i seen is the ram voltage is low for the ram..the ram is 1.65v ram and the mb is set for 1.35v
look online to set the voltage to the ram to 1.65v. the other issues is your cpu cooler is not on right or the fan not working the temps should not be that high for idle system.
PU Core Temperatures 1..4: ranging from 68 to 94 Celsius but on average 73 Celsius
http://communities.intel.com/message/152003
i would check that the cpu heat sink is on and working.
also use speedfan to see that the temp is right and the bios.
the idle temps after 5 min of running should be sitting in the 30-50c range depending on the size of your case and the room temp.
http://www.directron.com/ocz3p1333lv2g.html
m
0
l
July 11, 2012 5:05:01 PM

smorizio said:

the two things so far that i seen is the ram voltage is low for the ram..the ram is 1.65v ram and the mb is set for 1.35v
look online to set the voltage to the ram to 1.65v. the other issues is your cpu cooler is not on right or the fan not working the temps should not be that high for idle system.
PU Core Temperatures 1..4: ranging from 68 to 94 Celsius but on average 73 Celsius
http://communities.intel.com/message/152003
i would check that the cpu heat sink is on and working.
also use speedfan to see that the temp is right and the bios.
the idle temps after 5 min of running should be sitting in the 30-50c range depending on the size of your case and the room temp.
http://www.directron.com/ocz3p1333lv2g.html


Will try that SpeedFan utility. As for the memory voltage that is a bit confusing as in the motherboard settings (Performance | Memory configuration) the Memory Voltage entry is configured to 1.66V.

Do note that the ambient temperature in my room (I live in the tropics) is 31 Celsius with ceiling fan on :-/
m
0
l
a c 466 V Motherboard
July 11, 2012 5:29:07 PM

Motherboard BIOS Voltage Readings:
V12.0 12.000 V
V5.0 4.869 V
V3.3 3.17 V
V1.1 1.054 V
Vrcp 1.078 V
Voltage Regulator temp: 61 C
Memory Voltage: 1.350 V ----------------------------then issue with the ram or the mb settings as this is the reading of the ram.
Motherboard ambient temperature: 59 C
Aux Fan Speed: 0900
Front Fan Speed: 0960
Rear Fan Speed: 0960
CPU Fan Speed: 3120
most of the newer sandy brodge and ib cpu use memory that 1.5v.. you can use 1.65v but it higher then the intel memory spec.
m
0
l
a c 75 V Motherboard
July 11, 2012 5:55:06 PM

That is really low voltage for RAM for the time that this system would have been built. With the Intel® Core™ processors you really want to run them with memory at 1.5v ±5% (1.425v to 1.575v) after that it is out of spec and may cause problems. I would guess that if you upped the voltage on your RAM you might run a little bit more stable.
m
0
l
July 12, 2012 8:20:20 PM

IntelEnthusiast said:
That is really low voltage for RAM for the time that this system would have been built. With the Intel® Core™ processors you really want to run them with memory at 1.5v ±5% (1.425v to 1.575v) after that it is out of spec and may cause problems. I would guess that if you upped the voltage on your RAM you might run a little bit more stable.


I upped the voltage in the Bios to 1.74 as the 1.66V setting was resulting in 1.35 according that what people say here. However, in doing so the system appeared more unstable (though could be other factors) so I reverted to the original but problems persisted.

I decided to reflash with the latest BIOS but now the board -after a successful flashing bios message- gives me a black screen and a sequence of three (3) long beeps.

Can´t do anything with my pc now.
m
0
l
a c 466 V Motherboard
July 12, 2012 8:30:09 PM

try clearing the cmos...look on intel web page to see if there a bios flash back or recovery for the mb..most time it the bios file on a usb or cd-rom. and power the system on. try removing all the ram but one stick. see if the error still there.
m
0
l
July 13, 2012 9:39:11 PM

As it turns out the 3 beeps means memory problems. I had them in channels A (slot 1), B & C. I removed A & B so only C was left (configuration for single channel) and same problem. I exchanged it for each of the other two and the system worked.

As it appears one of the DIMMs, the one at the lower range of memory (0-2GB) was exhibiting problems until it finally crapped out.

That would explain the BSODs and in particular why it was becoming sooooo difficult to boot the system, many times it would crash as soon as it started going into Windows OR Linux (Live).

Now I will try to run it with 4GB (two modules). These memories are now end of life and they are also not sold in my country (I moved). Since the memories must be the same for triple channel I am having a hard time finding that OCZ, they supposedly have "lifetime warranty", let's see if they live up to their word.

If not I will have to look for another triple pack, most likely order it from abroad because PC3-10666 are hard to find here.
m
0
l
a c 466 V Motherboard
July 14, 2012 2:51:03 AM

use the mb ram guild and email intel...if you cant find 10666 1333 should also work if it easer to find that in your country.
it only slightly faster then the older ram and should work...
m
0
l
September 5, 2012 4:44:11 PM

smorizio said:
use the mb ram guild and email intel...if you cant find 10666 1333 should also work if it easer to find that in your country.
it only slightly faster then the older ram and should work...


PC develoing Black Screen sudden death problems, fearing it was again a memory problem I removed the two remaining 2GB modules and bought a single 4GB module. Problem persisted and then simply got worse.

Now I even have problems getting into the BIOS, apparently my nVidia GTX285 card has now gone bad. I wonder if it is this motherboard damaging all my components! first memory and now my expensive video card or perhaps it is the motherboard (damaged), haven't figured out how to rule out one problem or the other. I don't want to buy another expensive video card only to find out it is the motherboard...
m
0
l
!