Jump to content
Great_Gig

What Have I Killed Following A 6800gt Overclock?

Recommended Posts

This is a continuation of this Topic:

Help Needed With A 6800 Gt Problem?

 

______________________________________________________________________________

 

Ok guys, I have posted about this before, but I thought it would be easier to start again and say exactly what I did and what has happened? I feel a bit stupid really, but I clearly went at it the wrong way and I think I made some fundamental errors.

 

System:

 

Asus A8V Deluxe | Athlon 64 X2 4200+ | 2GB Corsair TwinX | 6800 GT | ZEROtherm Cooling | 1 x 160GB - 1 x 500GB | 2 Pioneer DVD/RW's | Windows XP SP2 | Silverstone 460W PSU

 

This worked perfectly with all stock settings and coolers. I wanted a quieter PC and also better cooling to have a go at Overclocking. This is exactly what I did and the end result?

 

First I dismantled the system and cleaned all the dust out. I then installed a ZEROtherm GX820 HSF to the 6800GT. Next job was the install of the ZEROtherm NV120 HSF to the CPU, using Arctic Silver 5 for both. The components were reassembled with a better job on cable management, it was a real mess before! I also installed 2 Nexus fans, a rear 120 for exhaust and an 80 at the front for intake. Both were powered direct from the PSU.

 

Before

pc1WinCE.jpg

(the VGA card was removed before I took this)

 

After

pc2WinCE.jpg

 

The CPU HSF was connected to the MOBO CPU fan header and the GPU HSF was connected to a MOBO chassis fan header - this I think was possibly a mistake? Everything worked fine at this point and my temperatures were down considerably from the stock HSF's fitted temps.

 

Now the problem - I used ATITool and found the max core and memory clocks. ATI put the core OK at 439Mhz and the memory at 1112Mhz. Quite an increase from the stock 350/1000. I set these and then started to test the temps and stability of the system. Both the CPU & GPU fans were running at 100% from the MOBO headers and the hardware monitor was reporting cooler temperatures than before the new HSF, so I felt comfortable. After I guess about an hour the system froze and I had to do a reset. It then refused to boot to Windows and just after the Windows logo I would get the BSOD for a nano second, before it would try to restart. I couldn't go to safe mode to try and reduce the settings back to stock. I left it till today and tried removing the VGA card and rebooting without it, and reset the BIOS to default. Same result so I put a spare HDD in that was already setup and would have had the stock VGA settings. It all seemed fine for awhile with all levels back at default, I had also connected the HSF's to the PSU direct instead of the MOBO. Then it became unstable again and now I get BSOD everytime after a few seconds of Windows starting - the error message is the same each time.

 

bsodSmall.jpg

 

Has anyone experienced this before and know or suspect what I have fried? Is it the 6800 the MOBO or both, or could it be something else? Also, was I wrong to attach the new HSF's to the MOBO headers. Any suggestions solutions would be welcomed?

 

Many thanks from a sick and slightly bruised feeling newbie!

 

GG

Edited by Great Gig

Share this post


Link to post
Share on other sites

What do you have your ram settings at? IE: Main timings, voltage, etc. As far as i know that error is usually caused by bad ram.

 

First check all your settings and make sure they are set to the manufacturers recommended settings (the 4 timings and voltage should be on a sticker on the actual module). If they are set correctly for your ram, try both sticks individually to see if the problem persists. If the problem continues with 1 stick, but not the other, then i think you found your problem. You can test to see if they are working via a few hours of memtest (TMOD & Loggan's CD is bootable and has memtest on it) and if memtest passes, you can fire up windows and do a prime95 or Orthos blend to stress the memory within windows.

Share this post


Link to post
Share on other sites

When you reset the m/b bios, did you do a full cmos clear including removal of the m/b battery?

 

If not you might give that a try.

 

Unplug power supply from wall outlet and press power and reset switches on your case a few times

Unplug all power connections from the motherboard

Remove m/b battery and place CMOS jumper in clear position - leave it there for 2-3 minutes

 

Replace m/b battery then move CMOS jumper back to save position

Re-attach all power connectors to the m/b

Plug power supply back in to wall outlet, power up and boot to BIOS

Load Optimized Defaults, save and boot

 

See what happens

Share this post


Link to post
Share on other sites

Well first of all,you've probably put too much strain on your psu,having all those fans and drives has probably done it no favours. Second,just because ati tool says thats what u can get,it does'nt mean you should.Take out sticks of ram and see if it will boot that way.If not,try a different gfx card.Next time,be more careful with your overclocks.Increase in small increments,and if you start to see weird happenings,back your clocks down.

Share this post


Link to post
Share on other sites
When you reset the m/b bios, did you do a full cmos clear including removal of the m/b battery?

 

If not you might give that a try.

 

Unplug power supply from wall outlet and press power and reset switches on your case a few times

Unplug all power connections from the motherboard

Remove m/b battery and place CMOS jumper in clear position - leave it there for 2-3 minutes

 

Replace m/b battery then move CMOS jumper back to save position

Re-attach all power connectors to the m/b

Plug power supply back in to wall outlet, power up and boot to BIOS

Load Optimized Defaults, save and boot

 

See what happens

 

Many thanks for the replies guys and I have had a few messages too. Most suspect the PSU and maybe the RAM, but the PSU seems favourite? I have just ordered a new Corsair TX650, so will try that first after clearing the CMOS and will test each mem module too. Will report back with a result.

 

GG

Share this post


Link to post
Share on other sites

OK, so here's the deal so far?

 

My new PSU arrived (Corsair TX650W) which I have to say I am very impressed with. Excellent build quality and packaging, with a 5 year guarantee. I can see why they are well received, it's also noticeably quieter than the Silverstone.

 

Anyway, I removed the old PSU, removed all component parts from the motherboard (apart from the CPU), then removed the battery and cleared the CMOS. Everything was re-installed, including a new M/B battery checking carefully they were seated correctly. The new PSU was connected then powered up, I entered the BIOS and selected the default settings. The system rebooted and Windows started to load. As the desktop appeared I again get the BSOD with the same error message. A few reboots confirmed the same symptom and error message each time.

 

Next I removed all memory modules (I have 4 x 512MB) and tried a reboot with 1 module at a time - exactly the same for each module. I am now running Memtest86+, one module at a time for 6 complete passes each, so far all seem good. Next I intend to put all the modules in and leave Memtest running overnight, just to check all the DIMM sockets are fine too.

 

Next I intend to do a fresh install of Windows with just 1 HDD and optical drive. Following that if it's the same, I will try and source a graphics card to test, which may be tricky as it's AGP. I also have another Athlon 64 CPU which I can test, if all those changes don't work, then I can only assume it's the motherboard?

 

Now, as I have never been in this situation before, do you think this fault finding process is sound, or does anyone have another suggestion? Finally, is this type of thing common following an OC of the graphics card?

 

Thanks guys.

 

GG

Share this post


Link to post
Share on other sites

Your methodology is very sound. And in fact extremely well thought out I might add. Good luck and keep us posted along the way.

 

Regards,

Share this post


Link to post
Share on other sites

Update:

 

Right, all RAM chips tested individually and seem fine, then I have had all 2GB running in Memtest for the past 14 hours - no errors. A fresh install of Windows is currently working OK, but I have not asked anything of the graphics card yet. I did have a small glitch on one reboot, whilst individually testing the memory. Got this message just after POST - "Overclocking has failed! Hit F1 to enter setup". Thought it was the mem module, but rebooted and have not seen it again, so a little strange?

 

CPU has been stressed using Super PI and all temps and numbers are as before. So I reckon the RAM, CPU and PSU are fine. Next I'm going to stress the 6800GT with RTHIBL to see if it falls over. I also have another 6800GT on the way, which I believe has never been OC'd. I am determined to get this sorted before I retire the system, as I won't let it beat me! When I have found the fault and fixed everything, I will OC it again and do a proper job this time. It's taught me quite a lot and thanks to everybody for the help that you have given me. My gut feeling is it's still the 6800, as I pushed it way too hard I reckon. But it possibly still may be the M/B?

 

Also, when a system fails like this, is it common to take the operating system with it - as it would appear to have done that twice, before I changed the PSU?

 

GG

Share this post


Link to post
Share on other sites

Maybe "common" is too strong a word. But any crash "can" corrupt the OS. For example this is one reason to use Memtest from DOS before windows boots, convenience and time being other reasons.

 

Now that you have a clean install of the OS make an image with Acronis or Drive Snapshot. If it happens again just restore the image in a fraction of the time it takes to reinstall Windows.

Share this post


Link to post
Share on other sites
Now that you have a clean install of the OS make an image with Acronis or Drive Snapshot.

 

Ah, already done Obi-One! True Image is your friend ;) One of my favourite pieces of software and one I never mind paying for - how many times has that saved me?

 

GG

Share this post


Link to post
Share on other sites
Ah, already done Obi-One! True Image is your friend ;) One of my favourite pieces of software and one I never mind paying for - how many times has that saved me?

 

GG

 

Sounds like you are having better luck on your side now. :lol: After hearing of your findings of things that are not at fault reminded me of when i swapped the 3500 single core to a 3800X2 i had similar BSOD's. my question to you is have you ever flashed the BIOS? as this was what fixed my problem. I flashed it to ver.1014.after that everything ran fine with no BSOD.

Share this post


Link to post
Share on other sites

Yes I did flash the BIOS, my A8V is quite an old one (first REV 2.0), so when I put the new X2 4200 in, it didn't even recognise it! That had me swearing for a minute or two I can tell you, until I realised I hadn't bought a dud, but it was me that was the dud! ;) I've actually got BIOS Ver. 1018.002, so all's cool in that department.

 

And I have to say I reckon it's sorted now. I have spent all day throwing whatever at the GPU - Ntune, ATITool, 3DMark and that RTHDRIBL what's-a-ma-call it. Nothing fazes it, have consistent scores/marks/frame rates, temperatures and voltages on the PSU rails. I've also stressed the CPU and RAM. So I can confirm it was indeed the PSU and my loyal 6800GT is alive and well - mind you she gave me a bit of a scare and I reckon I better not upset the old girl again. I must admit to being a bit attached to it. In it's day it was a great card and can still give a lot a run for their money. What did surprise me is that this card still sells for top dollar on eBay. I guess there are still a few socket 939 | AGP users out there?

 

I can also say hand on heart that until the failed OC, my system and never once produced a BSOD. The only weak link in the chain was the PSU, but that only gave way after I asked too much of it, and exceeded it's rated capacity.

 

So guys, thanks for all the support, encouragement and advice - but I think this topic is done. I now have a healthy system, with a fresh XP Pro SP3 and Nvidea 175.16 drivers. Can I repeat as well, the Corsair TX650W is a superior product, so far during all the stress testing today it's remained near silent. I have to put my head right next to it, to hear the fan and that's what I want, a quiet PC. The quality and packaging is top notch, a highly recommended product and I will be using Corsair again for my upcoming build.

 

GG

 

PS. Anyone want to buy a Silverstone 460W PSU? :D

 

ALIVE & KICKING!

 

gx820.jpg

Edited by Great Gig

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×