Need Help: HD 5830 is crashing my system

[SlapHappy]

Yesterday I added a HD 5830 to the two HD 5770s I have in my mining rig.  The 5830 was doing fine for the first ~18 hours at mem=600/gpu=975.  I'm mining with cgminer on Windows 7 x64.  I stopped cgminer and adjusted the over clock on the 5830 to mem=300/gpu=1000 to see if I could get a few more MH/s out of the card.  All of the sudden every time I started cgminer my machine would hang and then BSOD.  I've since reset the clocks back to the factory defaults of mem=1000/gpu=800 but my system still hangs as soon as I start cgminer.

I've verified if I disable the 5830 I can run cgminer on the two 5770s just fine.

Is it possible I permanently damaged the 5830 by running it at mem=600/gpu=975 for ~18 hours?  I had cgminer set-up to stop using the card if it reached 95C which it never did, it hovered around 65-70C the whole time.

All I got from the BSOD was that a MACHINE_CHECK_EXCEPTION had occurred.  Are there any diagnostic tools available to gather more information from the card about what's wrong?  Any way the tell if the card to fried?

Thanks in advance.

[deslok]

What psu are you using, os and driver versions. have you tried the 5830 on it's own without the 5770's installed in the machine?

[jjiimm_64]

Yesterday I added a HD 5830 to the two HD 5770s I have in my mining rig.  The 5830 was doing fine for the first ~18 hours at mem=600/gpu=975.  I'm mining with cgminer on Windows 7 x64.  I stopped cgminer and adjusted the over clock on the 5830 to mem=300/gpu=1000 to see if I could get a few more MH/s out of the card.  All of the sudden every time I started cgminer my machine would hang and then BSOD.  I've since reset the clocks back to the factory defaults of mem=1000/gpu=800 but my system still hangs as soon as I start cgminer.

I've verified if I disable the 5830 I can run cgminer on the two 5770s just fine.

Is it possible I permanently damaged the 5830 by running it at mem=600/gpu=975 for ~18 hours?  I had cgminer set-up to stop using the card if it reached 95C which it never did, it hovered around 65-70C the whole time.

All I got from the BSOD was that a MACHINE_CHECK_EXCEPTION had occurred.  Are there any diagnostic tools available to gather more information from the card about what's wrong?  Any way the tell if the card to fried?

Thanks in advance.

I doubt if you hurt the card, although I dont clock anything over 950.

uninstall from windows and reboot...
reinstall the ati drivers with the card in if that doesn't work.

[vapourminer]

you set the clocks in cgminer manually via command line switches? or a overclock tool. uninstall any overclock utilities, run ccleaners registry cleaner and reinstall.

its faintly possible the 5830 drew so much additional current that the mobo 4/8 pin cpu (not PEG) connector or mobo traces are slightly damaged from heat or over current from the combined cards. unplug/reseat the card, 24pin and 4/8 pin CPU connector, as well as any PEG adapters you have.

[P4man]

I had cgminer set-up to stop using the card if it reached 95C which it never did, it hovered around 65-70C the whole time.

You should have checked VRM temps too, if your card supports it. Its  possible they blew out. Its not so hard to keep the GPU cool, its quite hard to keep the VRMs from melting.

Anyway, you may want to try with the 5830 alone, preferably in a slot that you know works (one where there is a 5770 now) and see if that works.

[SlapHappy]

Thanks for the suggestions.  So far I've tried reseating the card and its PEG connectors in the same slot as well as re-installing the drivers and still no joy 

I'll try moving the 5830 card to one of the 5770 slots, or to a different machine, when I get a chance in a day or two.  I'm sure hoping the card isn't damaged.

In the meantime here's a bit more info on my system config if it helps anyone make a better suggestion for my situation:

  - using cgminer 2.0.4
  - over-clocking with cgminer command line options, no separate tools being used
  - using AMD 11.8 drivers
  - Thermaltake 850W power supply