Sign in with
Sign up | Sign in
Your question

Troubleshooting Nightmare

Last response: in Systems
Share
December 20, 2008 3:46:41 PM

Ok guys, this might be a little long. Let me give you some background first:

About a year ago I built my first system, which was the following:

ASUS M2N-SLI Deluxe
AMD Athlon X2 6400+
2GB (1 x 2GB) G. Skill DDR2 800
XFX GeForce 8600GT XXX
Seagate Barracuda 500GB SATA 3.0 HDD
Samsung Writemaster Optical Drive
Raidmax Smilodon Extreme Black Case
Raidmax 500W PSU (came with case)
Windows XP Pro (x86)

Probably a little much for a first build attempt, but it worked flawlessly for several months. About two months ago, I decided to upgrade to a Sapphire Radeon HD3850. I uninstalled the nvidia drivers properly (safe mode, driver cleaner pro, etc...) and everything was great for awhile. Then I decided to go dual channel on the RAM, so I took out the G. Skill and installed 2GB (2 x 1GB) OCZ SLI-Ready DDR2 800 RAM. This is when the problems began. As soon as I put that RAM in I started having lock ups and crashes. I made sure I had the voltage and timings right: 2.1V 4-4-4-15, and still got lockups and the system would even fail to boot occasionally. So I tried the RAM in two other computers with the same results. However, in all three systems, a single stick would work fine (didn't matter which one). But as soon as they were put together it caused problems. So I ditched that RAM and went back to the G. Skill. But then another problem arose. I would get occasional VPU recoverer errors in Far Cry 2. Then I purchased a 4GB dual channel kit of the same G. Skill RAM I was running. VPU errors persisted and became more frequent. I also began getting occasional BSOD's with the ati2dvag infinite loop error. I read that this could be related to RAM, so I tried going back to single channel. Same problems. As time went by, the errors, bsod's and system lockups became more and more frequent and began to affect all my games. It got to the point where I couldn't play more than two minutes of Call of Duty 4 and Far Cry 2 was by now totally out of the question. My next step was to go back to dual channel RAM, as that didn't seem to affect things much. Then I got a new HDD and did a clean install of XP. Still same problems. Then I changed my CPU to a 5000+ for unrelated reasons (heat). Finally, I just decided to RMA the card for a replacement. While waiting for my RMA to go through, I put an evga 8800gts 320MB in the system...and wouldn't you know it, I got the BSOD's and lockups with this card too! Infinite loop errors again! But they weren't nearly as frequent as with the HD3850. I managed to decrease their frequency even more by going into the nvidia settings and setting the max pre-render frames to 1 instead of 3. So yesterday I got my HD3850 replacement and just as I was expecting at this point, I got the same errors and crashes with more frequency than ever. So now I'm thinking PSU issue (probably should have been my first move). I just happened to have a 585W PSU lying around from an aborted build, so I put that in. Same results. So the very last thing it could possibly be is the motherboard. I'm pretty sure I confirmed this today. I didn't really want to do this, but I put the 3850 in the following system to test it out:

Biostar NF61S-M2 TE
AMD Athlon X2 6000+
2GB (1 x 2GB) G. Skill DDR2 800
Seagate Barracuda 500GB SATA 3.0 HDD
Memorex Optical Drive
Raidmax 420w PSU

I didn't want to put the card in this system because the power supply doesn't even come close to meeting the requirements. The 12V rail is only 13 amps. Plus, the PCI express link width is only x8. But I installed it and everything works like a charm! Ran CoD4 and FC2 with no problems on all high settings. The 8800gts also runs fine in this system.

So I guess after all that, I have three questions:

1) It has to be my M2N-SLI Deluxe mobo, right?
2) Could that OCZ Ram have triggered all of this and if so, how?
3) I'm on a very limited budget, but I need to get this system working. Would this motherboard be a decent choice: http://www.newegg.com/Product/Product.aspx?Item=N82E168... and if so, would it also be ok for an HD4850 later?

Thanks for any help.
December 20, 2008 4:17:22 PM

Well before you rule out a motherboard, how about actually using a real PSU?

Everyone you tried is crap.
December 20, 2008 4:22:49 PM

Sounds like the motherboard for sure.

Doubt the ram would have done it. Perhaps aggrivated an underlying problem in the motherboard. But the ram isnt going to cause the problem in the first place.

I've had the worst luck with Foxconn mobo's recently. The last batch of 6 I had in all had problems. They're usually a fantastic OEM option, but lately: Ehh.
I'd go for this one instead.
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
And with the mail in rebate, the price is the same.
Related resources
December 20, 2008 4:37:46 PM

roadrunner197069 said:
Well before you rule out a motherboard, how about actually using a real PSU?

Everyone you tried is crap.


Agreed, they are all crap. However, it seems odd that the the card and everything else runs the best on the crappiest of the three PSU's. I wish I could afford a real PSU, but I am very strapped right now and can only really afford one thing.
December 20, 2008 4:42:40 PM

curnel_D said:
Sounds like the motherboard for sure.

Doubt the ram would have done it. Perhaps aggrivated an underlying problem in the motherboard. But the ram isnt going to cause the problem in the first place.

I've had the worst luck with Foxconn mobo's recently. The last batch of 6 I had in all had problems. They're usually a fantastic OEM option, but lately: Ehh.
I'd go for this one instead.
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
And with the mail in rebate, the price is the same.


Yeah, I haven't heard the best things about Foxconn (never actually used one), but the one I was looking at has extremely good reviews by a lot of buyers. I had actually looked at the one you mentioned, but it was only reviewed by nine people, which isn't really enough for me to feel secure...haha. I do like that it's AM3 compatible though, as I plan to go AM3 in probably late 2009/early 2010. Any idea how reliable Gigabyte rebates are?
a b B Homebuilt system
December 20, 2008 4:45:27 PM

Before you go randomly replacing any more hardware I'd run memtest86+ overnight to test for RAM errors. Make sure the RAM timings and voltage are set manually in the BIOS to the manufacturers specs. That's the first thing I would have done. A good majority of BSOD's are caused by RAM problems. I'd RMA the RAM if you even get one error in memtest.
December 20, 2008 4:46:07 PM

curnel_D said:
Oh, and grab a better power supply. No one wants another damaged mobo because of a cheap power source.
Here are a few good examples.

http://www.newegg.com/Product/Product.aspx?Item=N82E168...
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
http://www.newegg.com/Product/Product.aspx?Item=N82E168...

Personally, I like the coolermaster power supplies the most. But they're all quality eq.


I would like to get a decent PSU at some point. I just don't have $80 to spend on a PSU right now. Definitely in the future, but for right now I just want to get the mobo situation taken care of. I will definitely keep these choices in mind though.
December 20, 2008 4:48:45 PM

shortstuff_mt said:
Before you go randomly replacing any more hardware I'd run memtest86+ overnight to test for RAM errors. Make sure the RAM timings and voltage are set manually in the BIOS to the manufacturers specs. That's the first thing I would have done. A good majority of BSOD's are caused by RAM problems. I'd RMA the RAM if you even get one error in memtest.


Forgot to mention that I ran memtest86 for a couple hours with no errors. Perhaps I'll do it again but overnight this time, just to make sure. And I have all the voltages and timings set to the manufacturer's spec.
a b B Homebuilt system
December 20, 2008 4:53:37 PM

I've seen a system pass memtest fine for a few passes then start throwing errors after multiple hours.
December 20, 2008 5:07:27 PM

Also check and see if there is any Bios' update for your board, which might have had some changes to work with your ram and graphic card. :ouch: 
December 20, 2008 5:08:14 PM

curnel_D said:
Sounds like the motherboard for sure.

Doubt the ram would have done it. Perhaps aggrivated an underlying problem in the motherboard. But the ram isnt going to cause the problem in the first place.

I've had the worst luck with Foxconn mobo's recently. The last batch of 6 I had in all had problems. They're usually a fantastic OEM option, but lately: Ehh.
I'd go for this one instead.
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
And with the mail in rebate, the price is the same.


You know the more I look at this Gigabyte board, the closer I am to being sold on it! Only thing is, it looks like that chipset heatsink is dangerously close to the PCI express x16 slot. A big card would obviously fit, it just looks like a tight squeeze.
December 20, 2008 5:09:54 PM

Yoosty said:
Also check and see if there is any Bios' update for your board, which might have had some changes to work with your ram and graphic card. :ouch: 


Already did that. I updated to the latest BIOS revision on ASUS' website.
December 20, 2008 5:09:55 PM

Memtest is usually much more extensive than a game or two that's run for a few mins. But... I suppose it wouldnt hurt.
Funny thing is, I've seen this problem before. (Ironically the two times I've seen it, it's been on an asus mobo.)
December 20, 2008 5:11:41 PM

matic3060 said:
You know the more I look at this Gigabyte board, the closer I am to being sold on it! Only thing is, it looks like that chipset heatsink is dangerously close to the PCI express x16 slot. A big card would obviously fit, it just looks like a tight squeeze.


The only way this would be a problem is if you're running one of the card coolers that wrap around card itself. Aside from that, there'd be no problem.
December 20, 2008 5:12:27 PM

curnel_D said:
Memtest is usually much more extensive than a game or two that's run for a few mins. But... I suppose it wouldnt hurt.
Funny thing is, I've seen this problem before. (Ironically the two times I've seen it, it's been on an asus mobo.)


Do you happen to know if replacing the board solved the issue?
December 20, 2008 5:19:17 PM

The first time, it was on an Asus board that was a couple years old at the time. I went through everything, and then finally got bored and tested the power supply which was overvoltaging extreme. I placed the blame on that. But yes, just replacing the motherboard and PSU fixed the problem obviously.

The second, I was building a new rig for a customer. Cant remember the exact mobo, but had a 6000+ and Crucial 800mhz 4x1gb ram. Took forever to figure out the problem, and finally just got a cross ship on an MSI model because that's what our Chicago distro had ready for same day shipping.
December 20, 2008 5:38:28 PM

curnel_D said:
The first time, it was on an Asus board that was a couple years old at the time. I went through everything, and then finally got bored and tested the power supply which was overvoltaging extreme. I placed the blame on that. But yes, just replacing the motherboard and PSU fixed the problem obviously.

The second, I was building a new rig for a customer. Cant remember the exact mobo, but had a 6000+ and Crucial 800mhz 4x1gb ram. Took forever to figure out the problem, and finally just got a cross ship on an MSI model because that's what our Chicago distro had ready for same day shipping.


Well, I think that pretty much seals it for me. This is what I think I'm going to do:

1) Run memtest all night tonight just to be sure. I haven't seen or heard of a bad stick of G. Skill yet, so I doubt that's the problem.
2) Order a new mobo tomorrow (more than likely it'll be that Gigabyte board).
3) If the problem persists, I guess I'll be forced to save up and get a new PSU ASAP. If the problem is resolved than I guess I should still save up for a new PSU!

I'll keep you guys updated.

Thanks for all the quick responses and guidance!
December 20, 2008 5:39:18 PM

Glad I could help.
December 20, 2008 5:45:16 PM

Have you tried reinstalling windows?
December 23, 2008 1:39:09 PM

To the above poster: Yes I have reinstalled windows...twice.

Update: I ran memtest86 overnight the other night. It completed 7 passes with no errors. So that should definitely rule out the RAM. Yesterday I ordered the Gigabyte motherboard discussed above. Looks like a really good board. If it works half as good as it looks, I should be back in business! I'll let you guys know what the final outcome is after christmas when I get the board. Happy holidays!
December 23, 2008 3:28:14 PM

7 is usually enough to find ram issues that are system stopping. (Infact, in my experience 99% of ram that's actually bad trips by the 4th pass)
December 30, 2008 1:18:50 AM

UPDATE:

So I got my Gigabyte board today. I took out my old ASUS board and installed the new one. Installed two hard drives, one for XP and one for Vista x64. Decided to put the 8800gts in instead of the hd3850, as it seems to run Grand Theft Auto 4 better and I've been on a GTA4 kick lately. So I installed Call of Duty 4 to the Vista x64 drive just to see if things ran smoothely before putting GTA4 on. As soon as I installed it and started it up, I got an error saying that my display driver stopped working. So I started the game again, and this time it says that directx had an unrecoverable error. Things are not looking good. So now I'm installing CoD4 to the XP drive to see how it fairs on there. If that has undesirable results, I guess I'll throw the hd3850 back in there and see how it does. This is so frustrating and its getting very old at this point.
December 30, 2008 4:53:04 PM

Yoosty said:
Download the latest DX9 Runtimes, even if you are running Vista 64-bit OS. It should help out with COD4. :hello: 
http://www.microsoft.com/downloads/Browse.aspx?displayl...


Thanks for the link. Unfortunately, Vista x64 has decided that it no longer wants to connect to the internet now. I so wish I didn't have to deal with Vista at all, but I want to be able to use my entire 4GB of ram and Directx 10. I think I'm just going to wipe Vista from my system and stick with XP. Maybe I'll give XP x64 a shot and see what happens. I really, really hope that Windows 7 doesn't have all of these issues.

Also, I installed Call of Duty 4 on my XP drive and I haven't had any crashes or problems yet. So at least something has gone right.
December 30, 2008 6:30:13 PM

UPDATE:

Well, I figured out why I couldn't connect to the internet. Vista, in its infinite wisdom, thought it prudent to download and install its own driver update for my Realtek Gigabit LAN. I simply rolled back the driver to the originals from the installation disc, and all is well now. So I was able to download the DirectX 9 runtime update (thanks, Yoosty!). It seemed to help a little bit. After installing it, I tried CoD4 and it didn't crash immediately. I actually got a good 1-2 minutes of gameplay in before the display driver would stop working. So then I decided to try rolling my graphics drivers back. I downloaded the Vista x64 drivers for an 8600GT, which is also an older driver for the 8800gts (WHQL 178.24), and it actually seems to be working. I just got through an entire round of CoD4 multiplayer with no problems! It did stutter a few times and I thought the display drivers were going to crash, but I'm thinking that might have been due to my ping being a lot higher than usual. So I think I'll try playing it for a bit longer and also try some single player and see how that works out.
January 3, 2009 2:50:20 AM

Well, I hope people are still reading this thread, because I have a new problem with this setup now (guess I picked a good title for this thread). I abandoned the whole Vista idea altogether. In Vista x64 and x86 I get constant nvlddmkm driver errors, which as I've learned is a long standing issue with Vista. So, I got everything working properly on XP and haven't had a crash yet, but now I've got a major stuttering problem in all games. I'm thinking it pretty much has to be the Gigabyte motherboard, because I've tried three different graphics cards all with the same results. The cards I've tried are:

XFX GeForce 8600GT 256MB XXX Edition
eVga Geforce 8800GTS 320MB
Sapphire Radeon HD3850 512MB

With each card I get the exact same microstutter. The main game I have been testing it on is Call of Duty 4. I will get very smooth graphics for a second or two and then stutter, smooth then stutter, and so on. My framerates are steady with the respective cards. I don't understand what is going on. Is it at all possible that since the board has a pcie 2.0 slot and none of these cards are 2.0 that somehow that is causing problems? It doesn't seem likely since pcie 2.0 is supposed to be backwards compatible with 1.0 or 1.1 . That's the only thing I can think of, though. I have a Sapphire Radeon HD4850 ordered and in the mail. It's definitely pcie 2.0, so I guess I'll know for sure if that is the issue sometime next week. The thing that really sucks is that I can't RMA the board because I already removed the UPC and sent it out for a rebate. Anybody have any thoughts on this?
January 5, 2009 5:29:24 AM

I doubt it's the motherboard. (Unless you're using very outdated or faulty chipset drivers. Then you might see those problems, but then again It'd be more likely you wouldnt be able to run games at all.)

http://downloads.guru3d.com/download.php?det=745
Download this, and make sure that all previous drivers are uninstalled before you install the new 4850. This will ensure a clean slate, so that you can be sure video drivers wont be a problem. And while you're at it, I would update your other important drivers such as sound (I say that first because the problem sounds like a problem a clan friend of mine just experienced with his sound drivers), chipset, ect...

If the problem is still there, send me a PM, since I wont likely look at this thread agian.
!