120 replies to this topic

[EuroFight]

Hi everyone
I've had my AMD FX rig for less than a year and from day one I was plagued by random crashes during gaming, sometimes it came up with a BSOD, sometimes a black screen then CTD. I have replaced the following components:
- GPU (Now HD7770 GHz Edition)
- RAM (BSODs and CTD occur with both sticks in all slots)
- SATA Cable*
- PSU (From El-Cheapo to Cooler Master)

And have made the following changes:
- Set memory ICs to Unganged
- Set LLC to 'Extreme'
- Reformatted to Windows 7
- Underclocked CPU*

For reference:
- BIOS is up-to-date with F6
- Graphics drivers up-to-date with Catalyst 12.10 drivers
- Memtest86 passed
- Ran OCCT a while ago with no errors, but can repeat if needed
- PC is generally stable for web browsing/Office etc. (unless underclock applied - see below)

* = Issue occurred less after that, then gradually came back with increasing frequency

At first, I thought the issue was graphics card related, so I spent £40 on a HD5570, then £80 on a HD7770 to try and resolve the issue, neither did.

I was then advised by a number of members on the forums that it could be down to the PSU, so I replaced it, but the issue still occurs. The power supply is more than required, at 100% CPU, GPU and RAM load my system should draw 400W, and I have a reputable 500W PSU.

Yesterday, I tried to sort out the issue by replacing the SATA cable (suggested by bp9801), and I thought the issue was solved, but alas no, this morning the issue recurred with increasing frequency. I then underclocked the CPU from stock and managed the system ran fine, but again, the issue came back after a while. Any attempt at a further overclock would cause Windows Explorer to encounter an error before even loading the desktop. I have cleared the BIOS to fail-safe after this.

I have no idea which component is causing the issue, and as you can see, I have made a number of purchases to attempt to resolve the issue. This is my only DDR3 and SATA rig, so I haven't got a spare HDD or RAM to test, nobody I know will have a spare AM3 or AM3+ CPU and I can think of nothing else that is causing the issue, except perhaps a PCI NIC, although I doubt that would cause this much of an issue so I haven't tried anything with it.

I can post a video if you think it will help, sometimes I get a BSOD, and have attached the minidumps, but if a BSOD does not occur and the system simply locks up and I have to manually power off the rig, or I get a CTD, there is no error. There do not appear to be abnormally high temperatures on any components before the crash, nor any strange voltages.

I am awaiting a reply from Gigabyte, my motherboard vendor with regards to an RMA, but I want to be sure it is faulty before I pay postage and waste time with a pointless RMA. I don't have any money to upgrade to Intel, hence why I went with AMD, although with hindsight, I probably should have gone on the blue team.


Any help anyone could give would be really REALLY helpful

Regards,

- Euro
OCC Pet Noob / General Helper

Update: Issue traced and resolved, despite Memtest86+ pass, issue was RAM

[El_Capitan]

What game are you playing? I crash on all my systems during Planetside 2. Buggy game that.

[EuroFight]

All games I have, Crysis, Battlefield 3, World of Tanks etc.
It doesn't crash at one specific point, it crashes at various points if that helps?

[Hyper]

This may sound funny but I've seen this happen because of a keyboard or mouse.... my room mate has a wireless keyboard that will randomly crash the system.. replaced it and the issue dissapeared.

if nothing else resolves it, try replacing or disconnecting any un-needed usb devices.

[EuroFight]

This may sound funny but I've seen this happen because of a keyboard or mouse.... my room mate has a wireless keyboard that will randomly crash the system.. replaced it and the issue dissapeared.

if nothing else resolves it, try replacing or disconnecting any un-needed usb devices.

Will try that now

[d6bmg]

It sounds like a completely wired issue to me. You can try to run the PC using minimum config (i.e. one RAM, one HDD, no USB devices except Mouse and keyboard.

Scratch that.
Your problem is happening due to atikmdag.sys which is part of catalyst.
See these threads:
http://www.overclock...ikmdag-sys-bsod
http://answers.microsoft.com/en-us/windows/forum/windows_7-hardware/atikmdagsys-windows-7-x64-ati-radeon-4850/94dbd5a5-4ba0-40cb-99d7-70cf14ffa5c3
http://www.sevenforums.com/crashes-debugging/152461-bsod-custom-error-3.html

The only solution I can find is the complete re-installation of catalyst.

Edited by d6bmg, 02 December 2012 - 01:48 AM.

[EuroFight]

It sounds like a completely wired issue to me. You can try to run the PC using minimum config (i.e. one RAM, one HDD, no USB devices except Mouse and keyboard.

Scratch that.
Your problem is happening due to atikmdag.sys which is part of catalyst.
See these threads:
http://www.overclock...ikmdag-sys-bsod
http://answers.micro...d7-70cf14ffa5c3
http://www.sevenforu...om-error-3.html

The only solution I can find is the complete re-installation of catalyst.

I've reformatted the computer multiple times, used multiple versions of Catalyst, installed driver-only versions of Catalyst. I've reinstalled Catalyst many times yet the issue recurs, so I'm pretty sure it isn't software-related.
Thanks for the suggestion though

Oh and a quick update: It isn't the WiFi card, keyboard, or mouse causing the issue, I've checked in different ports and using different equipment but the issue recurs unfortunately

[d6bmg]

^^ Ok, Then it is really wired.
After analyzing those dump files with whocrushed, it was pointing towards only one thing: atikmdag.sys

[EuroFight]

Yeah I noticed that aswell. That was my reason for initially suspecting the GPU. £120 later, I realize my mistake. Just because a specific driver is reported to have failed, doesn't mean nothing caused that driver to fail.
Although I'm glad I realized that, it means that any single component could have failed in my system leading to that driver failing. The atikmdag.sys or atikmpak.sys relate to no specific problem unfortunately.

Thank you for your time though

[EuroFight]

I'm thinking this could be heat problems, TMPIN0 and TMPIN1 are hitting 50C at idle. When I removed the CPU heatsink there appears to be very little thermal paste, so I'm assuming these temps are for the Northbridge or the CPU. I haven't seen these values so high before, so I'll try and get hold of some thermal paste to put on my CPU heatsink and see if it makes any difference.

Update: Moved the thermal paste onto the copper heatpipe on the CPU cooler, temperatures are now 37-38C at idle.

Edited by EuroFight, 02 December 2012 - 05:26 AM.

[ComputerEd]

Okay I notice you said you moved the RAM around but did you try a whole new set? Also concerning RAM are you setting the speeds or are you letting the motherboard auto set the speeds?

If the heat, RAM and PSU are all checking out and still having the issue you might have a motherboard itself that is bad.

[EuroFight]

I used a spare 4GB stick I borrowed from a friend to test the slots
I'm letting the motherboard detect my RAM speeds, although it is detecting a 1600MHz kit as 1333MHz if that makes any difference
The CPU is currently running AIDA64 at 26C ambient temperature, the CPU is hitting 57C max so I think cooling should be okay?
Cheers ComputerEd