Always restarts in middle of boot-up

It looks like you have gotten some very good advice and have your work cut out for you.

As usual, I look for the simple solutions since they are usually the ones I overlook myself. You could try reverting to the earlier BIOS version the you used before the problem appeared. The mentioned sensitivity of the Highpoint setup makes it worth a try.

Why the problem would not appear until a long time after your BIOS update is a mystery, but not an unheard-of occurrence.

 

Wow, it's been a long time. Sorry for the delay.

Well, I think I may have finally pinpointed the problem: it was my memory.

I never did do the kernel debug, mainly because I couldn't get my hands on a null modem serial cable. What I did instead was re-install Windows. I didn't want to, but I was desperate to have a semi-functional computer again. In the process of re-installing Windows, several very interesting things happened:

1. During the very early stages of Windows XP setup, while it was asking me which hard drive partition I wanted to put Windows on, I got a blue screen error. At this point I was certain my problem was hardware-related since there was really no OS to point a finger at in this early state.

2. After finally completing a successful clean install of Windows XP, the first thing I did was enable boot-logging to determine once and for all what driver it was that followed "volsnap.sys" that had been causing all of my problems. Turns out, it was "atapi.sys." But the cause of the problem was not due to anything ATAPI-related, I found out later.

3. Next I started installing my essential appliations, but after installing just one, I tried to run it and it gave me all sorts of errors about corrupted DLL's that it needed, and failed to initialize properly.

3. Also around this point I noticed that every single file I downloaded was corrupted. I had downloaded some executables and all of them gave some sort of error about corruption when I tried to run it. Now my hard drive was currently on the Highpoint controller so at this point I thought the problem was either my memory or the Highpoint controller. So I moved my hard drive off the Highpoint controller and onto one of the IDE controllers. But still, every file I downloaded was corrupted and I was getting errors trying to run any executable.

5. Then, I got an "IRQL_NOT_LESS_OR_EQUAL" blue screen exception. The blue screen itself gave no insight into the cause of the error. Thankfully, I had enabled complete memory dumping beforehand, so after restarting I had a full image of my memory on disk. I inspected the memory dump using the "WinDbg.exe" debugging tool from Microsoft, and ran the "!analyze" command. It told me that the crash was most likely caused by memory corruption. Bingo.

So I took out one of my two memory sticks and everything has been clear sailing since then. No corruption in any files or any blue screens or boot-up problems or anything. Whew!

I guess the problem I was having with the restarting during boot-up was a result of my memory corrupting the "atapi.sys" driver, and so long as that corrupted driver was left on the hard drive, my computer would never be able to get through boot-up.

What's really weird to me, though, is why MemTest86 never once reported a single error with my memory in all the countless times that I ran it.

 

very odd. i'd be interested to hear if this works out long term.

The short answer is that you cannot test hardware with software. Sometimes, if you are really lucky, the problem is so grossly obvious that even software can detect the problem. This is not usually the case. Of the hundreds of 'bad ram' problems i've worked over the years, only 5-10% of them were software detectable. Another interesting fact is that another 5-10% never failed in the hardware tester, but were resolved by replacement.

 

Was starting to wonder what happened to you. Glad to hear you got it.

Haven't tried Joe's recommendations yet myself as I've been too busy, but I sure tucked them away for rainy day experimentation.

Congratulations.

 

Damnit! It appears I spoke too soon in saying that this issue has been resolved. Now, a new behavior has surfaced, but I'm sure it is related to the same problem.

What's happening now is that my computer will spontaneously restart with absolutely no warning, no blue screen, nothing. I'll just be working on something, and BOOM!, restart. There is nothing written to the Event Viewer in regards to it, and no memory dump is performed.

At first I thought this problem could be something software-related, but now I know it's lower than that, because the random restarts started to happen during the middle of BIOS loading, at the very beginning of bootup, absolutely anywhere! It was really crazy. One time my computer restarted, and then restarted again before the BIOS screen could even appear...I just kept hearing my computer make its little restart "click" noise, which is how I could tell that it was restarting.

So now I don't even know if the problem is memory-related anymore. I don't know anything! I'm starting to think it might be power-related. But still, the cooling fans and everything remain spinning during these random restarts. And I've checked that the power cable is securely plugged into both the computer and the wall outlet. I was just wondering if there might be a problem with the amount of power being supplied by the wall outlet? This is an old house and it's very possible that the power being supplied is insufficient or unstable. Can I check this in some way? What else should I check?

 

You didn't actually reply to my post.
Did you check your capacitors?

 

And btw.....
Absolutely electricity could be part of your problems. I had a freind that had a problem in his fuse box many years ago shortly after he moved into his house... insulation was coming off one of the wires out there. Suddenly the light would flicker and TV pics 'turn' - computers would switch off.

Quickest way to test it...would be to take the box to a friend's house with better wiring than yours.

 

Jugdish, you can or have someone else test it with a voltage meter, if not steady, that's the problem.

The symptoms are identical to when I get one of those "micro" power failures, luckily, very rare for me.

Regards - Charles

 

The more likely suspect may be your power Supply or one of the cables connecting it to the MOBO. If you're up to it, pop the cover off and exercise those connectors by unplugging and replugging them (power turned off of course). When finished, leave the cover off and power the computer up. Using a wooden dowel or pencil, probe gently on all the wiring and connectors. Press gently on the MOBO and any cards that are installed. Watch all the time for any indication on the monitor or for the reboot taking place. Often an intermittent circuit can be pinpointed that way.

Then to satisfy Daisy, inspect the capacitors. However, the newer caps don't buldge or show corrosion when they fail so there's not much chance that you'll see anything that would indicate a failed one. Unless you have bench experience down to the component level, that's about all you can do for capacitor checking.

Post back your findings and any progress or whatnot.

 

I originally bought a bunch of these boards on a $39 closeout, including both RAID and non-RAID versions. Straight KG7's were simple to setup, not so the KG7 RAIDS. The behavior you are describing doesn't surprise me - it was a somewhat common problem and you should be able to find a lot of others who shared the same fate. The whole secret to this board is in the initial setup. Throughout this thread, I've tried to give you my best advice.

My very first post here ............
Couple of suggestions -
Use BIOS Kg7ds
Use Highpoint 2.34 drivers
Assuming success with an installation, load only the AGP mini-port drivers and filter using AMD DRVPK 1.30 - reference your manual (read more about it) but download this stuff from ABIT.

I don't profess to know exactly whats causing your problem - there have been a lot of valid suggestions offered by several very knowledgeable people and it may all boil down to a household electrical problem. I hope thats the case for your sake. At the same time, I am quite familiar with this board and its limitations - I'll stick with what I've already posted including some very detailed memory settings, but my first post is perhaps the most important one for achieving stability with this MoBo.

The quickest way to set this up is to leave 3&4 empty (still throw in the highpoint drivers following F6 prompt), and put your system drive on #1 master so as to avoid the Highpoint/ATAPI conflict all together. Once the OS is installed, move your system drive to #3 Master and put the rest of your HDD's on 3&4. - I'd recommend doing it in a two step process and booting with only the system drive on #3 so it retains drive C status. Then you can re-letter your CDROM drives so the rest of your HDD's can pick up sequential lettering when they are attached and fired up. (faster than re-lettering after the fact)

Good Luck - I hope its just household electricity, but I think there's more to it than that.

 

How very interesting. I've not come across anything written about this. They bulge because of the chemical reaction inside of them and I find it hard to understand how this would be different in newer components.
Not that I doubt you in the least. Could you link me to some articles so that I can read up on this?

 

I was speaking of the molded chip tantalum that is currently being used to replace electrolytics. I think it's mostly market shortage driven and not to imply any breakthrough in technology. You could say it's "Newer old technology revisited." My computer isn't new enough to have them but I suspect they'll get more common as supply conditions warrant.

Frankly, I've always appreciated the eyeball method for testing capacitors since they are normally short lead hard wired and paralleled with other components to the extent that it's nearly destructive to test them accurately. When the corrosion and bulging is eliminated, diagnostics will be much harder. We can live in hope that the newer units will last so long that capacitor failure will not happen before the computer becomes obsolete. The challenge is to get better components but not so much so that you have the best components in the obsolete computer bin.

Since the big fiasco during the 2000 ~ 2002 period when several manufacturers opted to use cheaper capacitors to meet production quotas, there has been renewed emphasis on developing better capacitors that use materials that aren't in such short supply. It's a dynamic situation and you can get more information on it than you want to know from a Google search.

Best regards.

Apologies to Jugdish, I didn't mean to hijack the thread.