Official GIGABYTE Forum
Questions about GIGABYTE products => Motherboards with AMD processors => Topic started by: dirtydan on November 28, 2011, 04:03:15 am
-
Crash dumps reveal memory errors, 2x 2g sticks all pass memtest. I'm a total noob at bios settings, they are set to specs on sticker, 1.5v 9-9-9-24. PC health status in bios shows DDR3 1.5V / 1.488V. Do I need to increase RAM voltage? Thanks a bunch.
AMD Athlon X 3 450 3.2GHz
ARCTIC COOLING Freezer 7 Pro Rev.2
GIGABYTE GA-970A-UD3
Corsair 4G XMS3 TW3X4G1333C9A G
XFX Radeon HD5770
Corsair 650TX 650W
CREATIVE SB X-Fi Titanium
500GB WD Caviar Black
-
Hi and welcome.
The actual STOP codes for the BSODs would help a lot. Please post them here.
When you say that the modules pass Memtest have you run it in the way we advise on single sticks for ten loops each ?
-
Hi dirtydan, ;)
I agree Dark Mantis... of course. But I have a other things to suggest:
Just check if you have a dust or very small pieces of paper or something else, like hair, etc.. In the memory sockets.
Take a flash light if need.
Sometimes inappropriate connections can caused this situation with the memory.
Check carefully that your cables (IDE, SATA, for example) are good and connected properly.
Sometimes the pins or contacts of the sockets or cables are damaged.
So the transmission of DATA (or electric current) goes wrong or nothing.
Gloup_Gloup
-
Quite correct of course. All physical connections should be checked. ;)
-
I'm a noob at bios settings but fairly knowledgeable otherwise.
0x1000007e, 0x0000003b
Machine was built by local vendor in August. It's been back twice, they replaced Saphire 5770 with XFX 5770 first time. Replaced Asus M4A87TD-EVO with Gigabyte GA-970A-UD3 next time. They ran memtest multiple ways they claim.
Symptoms, random BSOD, multiple programs stop working. Crashes happen when idle or stressing, doesn't matter. Can go days without issue. I have exact identical software loaded on my Windows 7 32 bit machine, runs flawlessly.
Solutions tried:
Re-formated/ clean install 6 times.
Clean install of ATI drivers 10.2,10.11, 11.7(current) 11.8, 11.11
Microsoft hotfixes pertaining to this crash, wouldn't install don't apply to this system.
Installed latest direct x runtime.
All MS updates, fully patched
System file check, sfc /scannow.
Memtest 4 passes sticks checked individually
Repair install, blue screened right after.
System crashed with no software installed after a reformat.
If you want me to rerun memtest 8 passes I will but at this point I think I'm chasing my tail.
-
Hi there,
my advice would be to go into BIOS and load Optimised Defaults. Save and exit BIOS to set the system back to it's default settings.
Go back into BIOS and make the changes that you would normally make, if any butDO NOT make any changes to the RAM speed or timings at this stage.
Your RAM will probably drop to 1066MHz but don't worry about that at this stage as we are looking to get the system stable before moving forward.
When you have done this try your PC. Do you still get the BSOD's or does it seem better?
-
I'll try that, I'll try anything at this point. I will have to run for a few days BSOD seem to be running every few days.
-
Yes, I understand and if you can get the system to run without BSOD's for a few days then at least we will have a platform to move forward with.
If possible, check in BIOS after you have loaded Optimised Defaults and run the PC for a while, to check what the RAM is being set at, and make a note of this, as the information will help with further troubleshooting. Also check the default CPU Vcore setting as this too may help later on.
-
I would suggest following Absic's advice and see if you can get any sort of stable platform to start with. If you have any problem doing that at this stage try running with a single stick of memory for now.
-
I see you are using a X 3 CPU.
For some reason my X3 720 BE also BSOD-s with that error codes, in various scenarios that require internet access ,when using a specific firewall (Outpost) .
So my advice would be to make sure you are not using any other security software than whats included in the OS and also to check your system for rootkits.
It may be some software incompatibility ,as in my case that firewall works ok with a dual core and the X3 works ok with another version of the firewall.
Otherwise just swap another CPU in and sell yours.
A defective memory controller can give that as well.
You could try and rise the NB VID control voltage ,the IMC (Integrated Memory Controller) voltage with 0,05 or 0,1 Volts .
-
Hi all,
I find this about 0x1000007e :
http://support.microsoft.com/kb/983615/en-us?fr=1
Extract :
This problem only occurs when the GPU utilization is high.
CAUSE
This issue occurs because of a time-out issue within the GPU. This causes a reset that leaves the GPU in an inconsistent state.
EDIT01: Other cause possible :
Drivers not updated, like by exemple :
nvm62x64.sys 2008 minimum to have
nvlddmkm.sys 2009 minimum to have
ASACPI.sys 2009 or 2010 minimum to have
If this drivers is not over or equal this years , change it.
Check your drivers about video, graphics card , directx, etc...
By exemple in : C:\Windows\System32\drivers
EDIT02: Enabled ACPI in Bios
Have you Service Pack 1 of Windows?
Make update from microsoft update.
Gloup_Gloup
-
SP1 fully patched. I requested that hotfix it wouldn't install.
Thats interesting about the firewall comodo is the first thing I install.
-
being a newbie ,I recently had a memory problem ,solved now thanks to Absic , during my investigations I contacted Gigabyte support , told them the specs (corsair memory 16gb 4x4) and that i had set it up as written on the corsair modules , voltage was listed as 1.5v .they told me amongst other things that the voltage needed to be at least 1.6v to run stable ,for my board/memory combination .
I did encounter several stop error crashes and a couple of BSODs before finally solving it .
voltage was not my main problem ,but now I have sorted my rig out ,the voltage is set to auto and runs happily at 1,6v .
obviousely I am not as learned as senior members here ,but its just a suggestion based on my own experience.
-
SP1 fully patched. I requested that hotfix it wouldn't install.
Thats interesting about the firewall comodo is the first thing I install.
Remove it and enable the included one.If you are skilled you can enable the outbound protection as well in the Windows firewall and never look back to 3-rd party ones.
It can be very well a conflict between the firewall drivers and some other driver ,video driver included.But as this firewalls drivers have to bypass the kernel patch protection you have a instability candidate right there.
-
I set bios to optimized settings and still running at 1333mhz. Should I set it to 1066 and reset the timings and to what settings?
I'm looking at W7 firewall control, I'll set it up on my test machine.
(http://img.photobucket.com/albums/v190/greyeyezz/P1030416s.jpg)
-
Take a look in CPU-Z and check if on auto the Command Rate goes to 1T.
I have some Corsairs than when on Auto ,the 890 GPA sets them at 1 T for some reason.So i manually readjusted to 2T and manually inserted the rest of Auto settings.
Memtest does not always show if the memory is ok ,you can pass those and have instability.
And don t forget about the memory controller voltage tweak.
Try all the suggestions and report after checking what happened.You already have work to do :)
Make sure you do some repairs of the operating system as well ,from the original disc.
-
Installled W7 firewall control, comodo uninstaller actually did a good job of cleaning up. Yes its at 1T. Lets wait see what happens before I proceed with memory adjustments. I'll run sfc /scannow Thanks.
(http://img.photobucket.com/albums/v190/greyeyezz/cpuz.png)
-
You are welcome.
Most probable the 1T generates issues with some programs/drivers.
Usually, if the timings are too tight you will observe slower booting than normal.If after changing to 2T you will have faster boot then thats the issue 100%.
-
Another BSOD right on schedule. I was using snagit editor. 0x0000001a
2: kd> !analyze -v
*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************
MEMORY_MANAGEMENT (1a)
# Any other values for parameter 1 must be individually examined.
Arguments:
Arg1: 0000000000005003, The subtype of the bugcheck.
Arg2: fffff70001080000
Arg3: 0000000000008491
Arg4: 000000001517f049
Debugging Details:
------------------
BUGCHECK_STR: 0x1a_5003
CUSTOMER_CRASH_COUNT: 1
DEFAULT_BUCKET_ID: VISTA_DRIVER_FAULT
PROCESS_NAME: SnagitEditor.e
///////////////////////////////////////////////////
I'm really getting aggravated at this point. I have some Gskill ram on the way.
-
So you practically need to go for the 2 T now ?!
Try that too ,until your new ram gets installed.
Anyways you should do a repair of W7 from within the disc ,sfc scannow may not do to much.
Also run a check disc on the windows partition.
Many BSOD-s can lead to file system corruption.
I would do a clean install or a back-up restore after replacing RAM.
If the new ram doesn t fix the issue RMA the CPU , it may have a bad IMC ,as it already has a core locked ,maybe a bad chip.
Edit: Oh, it s Vista :)
You should get rid of that too :)
-
Its W7 pro 64bit.
I did a disk check.
I did a repair install, BS'd right after.
-
What was the STOP code on the BSOD last time ?
-
Another BSOD right on schedule. I was using snagit editor. 0x0000001a
-
Another BSOD right on schedule. I was using snagit editor. 0x0000001a
Wasted RAM error code.
One of the sticks is bad , but using dual channel masks it when you run Memtest.Run Memtest for each stick ,20 + passes for test 5 and 7.
Or maybe they are not matched sticks and you should use them in single channel slots.
-
Yes, you must run Memtest on single modules. It is no good running it on multiple sticks at once as you won't know which one is faulty and have to do it singly anyway.
-
Was running memtest Tuesday on individual modules, 4 passes each, for 8 hours, and one did have an error but i was switching between tests, that might have borked the test. i don't know. I know which module is bad, I'll run it again overnight. 5, 7 or default test?
I also set my speed today to 1066 to see if it would crash at that speed, should I set it back to 1333 for testing?
-
Run one full pass , then 20+ per test 5 and next 20+ per test 7.It should show something.
Keep it at auto clock as frequency.
You could alternatively test run the stick that seems bad ,in all slots, to remove the chance of a faulty slot or faulty memory controller.
-
Full pass, #5/20, #7/20 on questionable module. 11 hours running memtest last three days. I'm done running memtest.
(http://img.photobucket.com/albums/v190/greyeyezz/P1030428.jpg)
-
Keep in mind the test was not done at 1333 as the clock in the picture is not 666/667 Mhz.
If you have not already tested your Windows you should with this speed that i see in your picture.
Just installed my 970A-UD3 today with some 2 X2 Gigs Corsair RAM and it seems to be ok ,ram was properly detected ,BIOS F3 ,but in single channel mode.
Keep us informed about how it performs with the G-Skills.
If you still see this BSOD-s with a new set of RAM you should RMA the CPU as well.
-
Running fine now at 1066. If I can make it 2 whole days without a BSOD that will be a first with this mobo. RAM will be here Mon. Will keep updated.
-
It may be that your CPU is an incipient revision and it "doesn t take dual channel properly" at 1333 Mhz.
I m not sure , but i think i ve heard of issues with Athlon II CPU-s integrated memory controllers, that would not work properly with dual channel RAM at 1333 Mhz.
In this scenario it s also good to know if the sticks you have are single sided or double sided.As using double sided takes 4 banks of the integrated memory controller to work.
It could be a CPU(integrated memory controller) limitation in this case.
You could sell your CPU and get another one if issues arrive at 1333 Mhz if you want more.
I m interested too in how the G Skills perform on the 970A-UD3. :)
-
Interesting thread with my exact RAM although an Intel platform.
http://forum.corsair.com/v3/showthread.php?t=82751 (http://forum.corsair.com/v3/showthread.php?t=82751)
-
Gskill is in at 1333mhz. Corsair ran perfectly at 1066 for four days. We will see.
-
Another BSOD copying files to a external. This is with the new Gskill ram running at 1333.
0x0000000a, IRQL_NOT_LESS_OR_EQUAL
-
Your BSOD code often means that you have insufficient QPI/Vtt voltage so you might want to check that and maybe increase it a touch. It could also be memory or memory controller problems. In your case I would discount the modules themselves.
I would also suggest when testing try not to use external drives as that is adding another place for faults to creap in.
-
"IRQL_NOT_LESS_OR_EQUAL", haven't seen 1 of those in a long long time. I see you have an Xfi, is it an installed card? if so remove it.
It could be memory - but error IRQL_NOT_LESS_OR_EQUAL has, in tne past, for me, been a piece of installed hardware......... especialy sound cards & modems ( a modem today is as rare as carburetors on a car )
-
Uninstall creative drivers before i remove card? Enable onboard sound?
-
Hi dirtydan, ;)
Normaly in windows use a Device Manager...
But he exist others softwares...
Please completely remove older version driver first with :
Drivers sweeper : http://phyxion.net/Driver-Sweeper/Driver-Sweeper/Version-3-2-0/
and soft for Windows Registry :
Ccleaner http://www.piriform.com/ccleaner/download
jv16 PowerTools 2011 http://www.macecraft.com/downloads/jv16pt_setup_hb.exe
Wise Registry Cleaner Pro http://www.wisecleaner.com/soft/WRCPro.exe
Check on Hard Drive if you have a trace of deleted, removed software in folder.
Use Windows search function for find it.
and restart the computer.
So Remove all drivers about your sound card.
When the cleaning of computer is finish and ok, you have a possibility to Enabled Onboard sound in Bios.
================
But i am not sure if my infos are a solution for your problem. So I am glad to give this infos... (maybe util next time for you)...
I said this, because I find this infos (again :P) :
This issue occurs because Power Manager opens an Advanced local procedure call (ALPC) port. However, Power Manager closes another port instead of closing the ALPC port. Every time that a power request is made, a memory leak occurs. When the leaked memory usage accumulates to a certain level, the computer crashes.
Source : http://support.microsoft.com/kb/979444/en-us?fr=1
So maybe wait a new help from this forum before try this solution... Otherwise make more search on Internet about your problem...
Gloup_Gloup
-
I'm getting close to solving this, 1066mhz stable, 1333mhz not stable. Its clearly a memory setting issue I believe. To start pulling hardware might be counter productive.
-
I'm getting close to solving this, 1066mhz stable, 1333mhz not stable. Its clearly a memory setting issue I believe. To start pulling hardware might be counter productive.
Most probable i am right and your X3 is not able to run RAM at 1333 Mhz.
Athlon 2 at it s start up was with one RAM speed notch backwards compared to Phenom 2.
The limitation is in the CPU as the IMC-integrated memory controller is in it.
I would try to see what it does if you install all the sticks you have now ,G-Skill and Corsair.
Swap the CPU out and the enigma will solve most probable.
The motherboard by itself is ok as i see mine being ok stability wise.
Upgrade the CPU :)
-
My DRAM voltage was at 1.510 for some reason, I set it to 1.500, uninstalled Avira last Tuesday and haven't seen a blue screen or crash since, and am running at 1333mhz.
Out of curiosity Vezina what are your ram settings in the bios with your gigabyte board, our systems are very similar.