Hardware/Software error after BSOD

My PC (AMD Athlon64 X2 3800+ Dual Core, nVidia 256 MB GeForce 6800 GT PCI-E, 1024 MB DDR (one plaque), S-939 ASUS A8N SLI, SATA Disk (Maxtor 250 GB), Windows XP Professional) had a blue screen of death (BAD_POOL_HEADER, STOP: 0x00000019). Something corrupted a RAM file apparently, causing Windows XP to shut down immediately and a RAM dump took place.

Ever since that first happened, my PC gave the following message at startup: IDE Channel 1 no 80 conductor cable installed.

Now, that’s a crock of sh!te, because I think my SATA disk has 40 pin cables and it used to work just fine before the first BSOD. I however do think to have isolated the program that causes the BSOD in the first place. It happened a lot more times, always when Azureus (Java based, RAM devouring) bit torrent client was active.

However, since it first happened, a lot of other programs are mildly to heavily influenced by whatever change happened inside of my system. Games, such as Knights of the Old Republic were very heavily influenced; pathing stuck, movement became close to impossible, textures and shaders seemed to disappear or corrupted in some way. Even the menus flickered and blurred in a way they shouldn’t. Serious Sam II and F.E.A.R. were affected so heavily that mouse panning became damn near impossible (the scope kept switching from looking straight up to looking straight down within the slightest movement). Most other games boasted smaller, but similar troubles. Normal applications generally seemed unaffected.

I might be mistaken, but I would say that rules out hardware troubles. If the cable doesn’t link Mobo and HDD correctly, it shouldn’t work correctly to begin with.

I’d like some Advice on this matter. Is there chance that a simple reinstallation of Windows XP on my system can resolve the matter, or do I have to take more drastic actions?

I had something similar happen in that the Bios seemed to get corrupted & the only solution I found was restoring from the on-board Bios backup. I then had to load the updated Bios firmware I was using.

My symptom involved the IDE controller & unrecognised devices.

BTW. have had no problem with Azureus although the Java process does get quite large ( currently 116MB )

I’m not an expert on hardware problems, so please sorry if my will be a little dumb questions.

I see in your config that you have only a s-ata harddrive, with no IDE harddrive, right?

For what I know, s-ata controllers on this mainboard are all RAID controllers, and during first windows installation they require specific drivers installed before than windows itself. Have you installed these drivers before than windows installation?

Another thing: if I’m not wrong, in this mainboard there are two s-ata controllers, one of them can create RAID sets in conjunction with IDE channels. If your harddrive is connected to this second controller, maybe your bios is searching for an IDE drive that in your system is absent, and this can create some sort of conflict. Can you try to connect your harddrive to the other controller? Maybe this can solve your problem.

Start by downloading Memtest86, put it on a floppy and run it overnight to check your RAM. Windows may be telling you in it’s round-about way that you’ve got some RAM that’s stuffed.

That’d be strange, because RAM seems to be dumped when it happens. Unfortunately, my PC has no floppy drive. Didn’t really feel I needed one honestly.

I had something similar happen in that the Bios seemed to get corrupted & the only solution I found was restoring from the on-board Bios backup. I then had to load the updated Bios firmware I was using.

How do I do that? I’ve never done such a thing :eek:. Can’t I simply reinstall Windows?

Often the motherboard comes with software to update the Bios, some actually have backup as my Gigabyte does.

Initially check the manual as it’ll probably have something in there about this. Or the manufacturers website.

If this is the problem, and to be honest I don’t know if it is, then a Windows install won’t fix it.

You can usually flash from any bootable device - you can make a bootable CD to flash from, or a thumb drive.

But seriously - check the RAM. The handy little program memtestx86 comes in a version you can burn to CD and boot from. I’m going out on a limb and guessing that your RAM is hosed.

“I however do think to have isolated the program that causes the BSOD in the first place. It happened a lot more times, always when Azureus (Java based, RAM devouring) bit torrent client was active.” :confused:

Hmmmmm Maybe we caught our selfs a bug off of P2P that is playing BSOD for you.
2+2 still =4 :iagree:

“SATA disk has 40 pin cables” This is new :eek:

Lets be honest: You ain’t got Azureus running to download every version of Linux on the planet :disagree:

… or BitComet.

Watch out for trojans in this case.

This all is adding to my confusion a bit :stuck_out_tongue:

I cannot emphasise enough how much I would like ‘stroking it makes it better’ to apply to computers right now.

PS: I have given Far Cry another try. It cocked up as well, but it told me that it ‘couldn’t read from memory correctly’. My heart gets heavy here. So yeah, I’ll make a memtest CDROM and try it out. Question though.

Can memtest x86 only scan for trouble or resolve some as well? Like, if it finds a bad block, can it like, cement it to avoid programs calling it?

unfortunately no it cant, if you dont get any errors then try a program called prime95 (ran from windows) which will stress your cpu and basically test for stability

This is getting dirty Yeaaaaaaaaaaaaaaaaaa :bigsmile:

Apparently you have only the one stick of RAM? Swapping in a different stick would be a quick way to rule in or out bad RAM. But nearly every KB article on that stop error points to hardware. You could even just try removing and re-seating the RAM.

Video card problems are not ruled out either. You wouldn’t by any chance be overclocking anything?

What I’m wondering… as you get a msg that no 80-pin conductor cable is installed… could it perhaps be that your harddrive cable is failing? What if a faulty cable caused the crash in the first place? May be a stupid thought, but it should be easy to check out.

That sounds rather awkward, since a conductor cable failure would result in the whole damn thing not working, right? I thought first that perhaps it could be a driver failture, but I’m not sure anymore.

About the overclocking question: I’m not overclocking anything. Overclocking is for idiots who aren’t happy with what they have and blow up what they do have in the process. I have a nephew who overclocked his CPU and put an arm and a leg into heavy water cooling nitro vac nonsense. Why not simply buy a bigger and better CPU for like 20 % of the costs :p. Poetic justice was delivered when he was going out at night and left his computer to download; his water cooling failed and his CPU blew up. I kinda feel bad for him though, but everyone advised him against it for months. But then again, what can you expect from a guy whose favorite movie is ‘the Fast and the Furious’? :slight_smile:

And yeah, bkf. I’d stroke anything to make it feel better. :wink:

It’s easy to try if the cable is causing the problems. Imagine that you find out, after a lot of hassle, that the cable was causing problems. Thwat would really suck eh? :slight_smile:

Overclocking is fun btw. Esp when you can get more performance from your system without buying expensive cooling techniques etc…

Yeah, perhaps…

Anyway, I’m running memtest now on my PC, and I’m seeing something awkward. RsvdMem is 276 MB. Now I might be crazy but isn’t that a little much memory that’s not being tested for some reason?

The MemMap is e280 (-Std), which meant, I believe, that Reserved Memory is also tested, but I’m not sure. Is it?

Oh yeah, an additional question: My PC has ran a full pass scan three times now. How many… things… blocks, or whatever, need to be tested fully? Or rather, howmany of these blocks tested are there in my Memory?

Run Memtest86 for 6-8 hours, or at least 20 full passes.

What does one pass mean? One chip on the dumm? Or like, the whole thing?

Um, well it tells you on the screen the number of passes. One pass is all the memory addresses. With 1024, a pass can take 30-60 min depending on the memory and controller.

What you’re looking for is “0” errors, no errors are acceptable. Often they don’t appear till after some passes.