Isolating BSoD to IDE HD and/or AVG

[infiniti25]

Change the northbridge to +0.1v and reboot. The error you are getting occurs during bootup (xp scrolling screen) when any overclock is bad and it is usually down to the voltage being low. Since you are not overclocking but have used all 4 ram slots and both storage controllers, you might be putting a strain on the chipset, as occurs during overclocking. The remedy is usually a single increase in voltage to stabilise things.

As for the ram I would still be adding +0.2v because your ram is rated as 2.0-2.1v and the board runs the standard 1.8v with the normal setting. You could be causing data corruption to occur by undervolting the ram, I'm surprised your sysem makes it into windows and doesn't bluescreen during the boot sequence.

Also to prevent the automatic restart so you can see the full stop error, check this http://www.microsoft.com/windowsxp/using/h...el_02may13.mspx

Edited December 22, 2008 by infiniti25

[Groonx]

YOU GUYS. . . are SUPERLATIVE!

Okay, it is evening, and I'm back from a long tiring walk downtown through fairly rare Vancouver snow. I was not planning on doing this until tomorrow, but, well. Anyway, please bear in mind that I have not done everything yet. I haven't recorded the crash dumps. Er—if there ARE to be any more crash dumps to record.

Now, let's see. First I scored a point myself. When my computer starts up from a cold power on, it has—or had—the definite tendency to make up to a dozen quiet, monotone groans. More or less the way your stomach might in anticipation of starvation, but very much on a single tone. Each groan lasted a second or so; then there was a silence; then another groan. Meantime, the monitor waited on power-saving mode. It was a bit disturbing, but eventually the system would come up.

With so many pieces of hardware in a small space, I didn't think I would be able to get my ear close enough to the offending piece. However, I tried, and I succeeded. I realized it was one of my two optical drives. So I opened the drive bay door in the front and started the system again. This time I could SEE that the green LED for—okay, c'mon, Guys: for WHICH optical drive??—t-h-i-n-k. . . for the CD-RW which is the Master on the same IDE cable on which the problematic HD is the Slave!!!!—I could see that that CD-RW's green LED was coming on repeatedly, or staying on (in truth I forget which, but it sure was NOT doing what it should have been doing).

So I done good, huh? Okay, I felt my next move would be to replace that cable. It is a brand new, rolled-type IDE cable.

However, I also felt that poor old infiniti, despite his name, would not be able to wait forever for me to make a move on my BIOS. So, merely out of kindness to his infinite patience, you understand, I entered the BIOS and I did exactly as I had been instructed, like this:

1.

System Voltage Control: Change from [Auto] to [Manual]

Northbridge: Change from Normal to +0.1V

— Now the Northbridge showed 1.200V beside it.

Make no other changes, and reboot into Windows. I thought the system came up a bit slowly—it seemed to—but it came up okay. The PATA hard drive remained disconnected.

2.

Back to the BIOS:

DDR2 Voltage Control: Change from [Normal] to +0.2V

— Now it shows:

DDR2 Voltage Control = 2.000V

Northbridge Voltage Control = 1.200V

CPU Voltage Control = Normal

Normal CPU Vcore = 1.3500V

Make no other changes. Save and reboot into Windows. It seemed to come up slowly again, but it came up okay. The PATA hard drive remained disconnected.

3.

Power down.

Attach the PATA hard drive.

Power up.

Strangely, for the first time ever, since I built this system several weeks ago, I did not see a verbose screen preceding the bootscreen. I had intended to look to see where I could eventually turn off the verbose display, but this time the machine went directly and very quickly into the bootscreen. To disambiguate, by verbose pre-bootscreen I mean a black screen covered from head to toe with lots and lots of white-lettered information about a ton of things I know nothing about. By bootscreen I mean a silhouette of a young lady swimming up toward the surface of a tropical sea.

— Very quickly, faster than I'd ever seen things happen before, the system came up into Windows. With the PATA hard drive connected. WITHOUT a subsequent I'm-Outta-Here Boot Dump.

This is very early in the game. I was so excited that I came here and wrote this right away. Maybe next time the whole thing will blow up.

But maybe it won't.

So I'm sending each one of you who helped me, a free Rolls Royce. I think it is the least—oh, PHOOEY. I forgot that I had promised the dog I'd buy some new dogfood for him. Well, gee, I'm sure you all realize that it is the THOUGHT that counts. After all, what would you have wanted with new Rolls Royces anyway? They are good for little more than trading for a house.

Infiniti told me to spend more time on my system, and less time saying thank you.

Thank you thank you thank you thank you, every one of you!!

(I don't always follow advice.)

_________________________________

[EDIT:] Following the above, I did another restart. Then I did a full power off/power on startup.

— The verbose screen still precedes the silhouette of the lady swimming in the tropical sea. (Well, I only assume it is tropical, as she would appear to need not an overabundance of attire.) Maybe I simply didn't see the verbose screen that one time.

— I get the strange feeling that some startups are rather slow, and some are very fast. I shall look into that, but honestly, I'm so happy I couldn't care less!

— The third hard drive is attached and working flawlessly.

— The empty-stomach monotone groans of the CD-RW are gone. All is quiet and serene.

You know, I had always THOUGHT it seemed as though the system was lacking juice. I mean, "I knew that." I was just testing you. No; not so. It really did seem to be lacking power somehow, but our friend's infinite wisdom was what made it work. And also I learned a lot from all of you other guys, because next time I get a crash dump, I'll know what to do about it.

Poor Infiniti. He kept on saying, over and over, in his quiet way: "Uh, say, uh, you really should adjust a few things in the BIOS, y'know."

Infiniti, you deserve the right to say at least a dozen "I told you so"s, perhaps one a month all through 2009.

But I am so very impressed with all of you. I wonder if you can imagine how much this means to me. It's my first self-built computer, and it is by far the nicest computer I've ever owned. Doggone it, I am very very very happy.

P.S. — I'm 62.

Edited December 23, 2008 by Groonx

[infiniti25]

i know from experience the northbridge voltage would require upping for all that connected stuff to work stable.

Hope you don't experience any problems.

I think it might be wise to monitor temperatures on the northbridge using speedfan because we increased the voltage. A general temperature guide for an intel p-35 based board would be to not go over 65degrees, but your amd based board could be more or less.

I would aim for no more than 60 degrees under load (heavy use) to maintain stability and data integrity. I mean we increased the voltage to prevent data corruption due to northbridge instability in the first place right

Your welcome anyway. Glad I/we could help.

MATT

[Groonx]

Well, I installed SpeedFan okay, but when I pressed on the icon to try using it, I immediately got a blue screen again which went (very fast) into an automatic restart of the computer.

I haven't had the guts to retry SpeedFan. Maybe tomorrow.

I've never had that happen before, and I play with hundreds of software programs. Strange.

At the moment, without making any changes, I still have SpeedFan installed but not in use, and I'm running a full AVG scan, a TuneUp Utilities maintenance scan, CCleaner, the Windows Disk Cleaner, and Firefox. All together, and they are doing fine.

Maybe SpeedFan has some constraints. I'll read some more.

I'm still happy anyway.

__________________________

[EDIT:] So I tried a download of Lavalys EVEREST Home Edition 2.20. Same thing happened, exactly.

I downloaded it and installed it; no problem. I got into it with no problem, either.

But when I pressed on Computer/Summary to get EVEREST to check things out, I got another very fast blue screen and automatic reboot. Just like with SpeedFan, the system is working fine as long as I don't ask for an analysis. Something in the analysis, whether performed by SpeedFan or by EVEREST, is crashing the system.

I've done zillions of analyses before, on all my systems. I'm very used to using analysis software.

I wonder what is being tested that has such a violent reaction. It is as though the tests trigger something that can't tolerate being tested.

By the way, CPU-Z seems to run fine.

More tomorrow.

Edited December 23, 2008 by Groonx

[cluberti]

I'll ask again.

Would you be able to get a complete memory dump from the box the next time it occurs, and then compress/upload it somewhere so we could look? It'd remove some of the guesswork.

[Groonx]

Cluberti, yes, I'll do that tomorrow.

It looked for awhile as though the blue screens had vanished, as those associated with the third hard drive do seem to be a thing of the past.

Now I have found that all 3 of SpeedFan, Everest, and Alex Nolan's Drive Manager—all of which monitor S.M.A.R.T. information—will cause the blue screens and automatic restarts. I'll use your technique to get a report.

I have System SMART Fan Control disabled, as I like my rear case fan to stay at high speed all the time. When SMART was enabled, that fan would not even start with the computer. I guess it was too cool. The side panel of the case is still off, while I finish working with components inside, and the BIOS reports a system temperature of 37º and a CPU temperature of 30º.

I tried enabling the SMART Fan Control, but Drive Manager wiped out the system anyway. I would not have expected it NOT to, as it would be a pretty stupid analysis app that crashed a system just because System SMART Fan Control was disabled. So there is something more to it than that.

In the BIOS:

Vcore = 1.392V

DDR2 1.8V = 2.048V

+3.3V = 3.392V

+12V = 12.239V

System Temperature = 37º C

CPU Temperature = 30º C

CPU Fan Speed = 771 RPM [This is a 120mm Scythe I use on the Xigmatex cooler, instead of the stock Xig fan.] CPU SMART Fan Control enabled.

System Fan Speed= 1439 RPM [This is the Xigmatek fan, which I use as the rear case fan.] System SMART Fan Control disabled.

In the case of Lavalys Everest, you can actually see that the app is reading SMART information at the moment the blue screen crash occurs.

All of these recent crashes have been during "analysis scans". No other programs seem to be having any trouble at all, and the computer seems to be behaving very well. The third hard drive remains connected and functioning properly. The CD-RW on the same IDE cable is starting properly. If it were not for all 3 of these "analysis" apps causing crashes, I would not have any reason to think anything was wrong.

So, more sophisticated and complete crash info tomorrow morning.

___________________________

[EDIT]

I have done the registry work, and this is all set for a test in the morning.

Two questions:

— Following the instructions exactly, I've set up the pagefile for my 8GB of RAM as 8192 + 50 = total 8242 MB. This could be bigger. No problem with 8242, right?

— Does anyone have a good idea for a place where I can upload the results? I'm thinking of doing it as a 7-Zip file.

I would do it tonight, but it is past midnight, and I'm very tired. I'm better to do it when I'm wide awake.

The crash itself is all too easily replicated.

— Any preference? SpeedFan, Everest Home Edition, or Alex Nolan's Drive Manager? I'm thinking of Everest, because it takes the longest time to crash (by a couple of seconds), as it goes through a few things successfully.

Edited December 23, 2008 by Groonx

[infiniti25]

I know it seems that I keep banging on about correcting your bios settings, but read here first before bashing me.

I would try EITHER upping the memory voltage to 2.1v (1.8vStandard+0.3vOvervolt) OR the mchv +0.2v.

The PATA drive started to work after upping either of those so its still a stability issue, and that link backs me up regarding putting a strain on the chipset so I suggest the +0.2v on the mchv(northbridge voltage).

...But I can tell you right away what everybody there will tell you: The more sockets you use, the less stability you get....

...It does not need rocket science to understand the powering 4 sockets not to mention the data management between them and the controller is more demanding than when populating only 2 sockets....

Still give the dump a go, but 8gb will take lots of time to upload! (at least it would for me with a 10mbps down and only 384kbps up )

Those bluescreens are the sign of a bad overclock or a strain on the northbridge. Hope someone else could back me up here apart from just that link.

The fact that the system is crashing while performing the analysis part suggests that the system hasn't had a real load placed on it since you got the PATA drive working. I guess just web browsing?

[jaclaz]

TWO things:

1)

Well, I installed SpeedFan okay, but when I pressed on the icon to try using it, I immediately got a blue screen again which went (very fast) into an automatic restart of the computer.

First thing ANYONE should do is to fix the "stupidest" (IMHO - there may be much stupider ones ) default setting of XP:

http://www.pctools.com/guides/registry/detail/229/

Knowing WHICH exact STOP ERROR comes on the BSOD is usually half to 3/4 of the solution.

2) do what cluberti says, though not "specific" to this board, and although you are responding very nicely even without probably having read it, do read the "common sense advice" here:

http://www.boot-land.net/forums/index.php?act=boardrules

with particualr attention to point #f and expecially point #f.3

By looking a a full dump cluberti can tell you more things about your system that you will learn in months of studying.

......

Ok, I lied , they were three:

3) NEVER introduce more than one change/fix at the time, if you introduce two or three of them together you will NEVER be able to understand WHICH one is the actual fix (and it's needed) and WHICH one(s) is irrelevant to the problem at hand.

jaclaz

[Groonx]

All advice well received. I find this morning that I must be away for much of the day; will recommence as soon as I return.

[Groonx]

I've just returned, and in a few minutes I'll see about getting that complete memory dump. However, I'm not sure exactly where I'll upload it, assuming it is a big file.

At present:

— Both times I started the computer by a cold power-on today (this morning and now), I heard 10 or more of those single monotone "groans" from the CD-RW, as it was doing before the voltage was adjusted yesterday.

Eventually I did a Ctrl-Alt-Del to try again, but this time I heard a single very long beep—something new. After a moment or two, it repeated: a single long beep.

— I didn't like that, so I shut the power off.

Power ON again: everything okay.

— I had read that some BIOSes would change power settings they didn't like back to default, so I checked the BIOS just now. But it still has the adjusted voltage settings:

DDR2 +0.2V = 2.000V

Northbridge +0.1V = 1.200V

Okay: F10 to single short beep POST (good) and into Windows. Seemed a bit slow.

However, the PATA hard drive remains connected and working, so I am really in an improved position over the way things were originally.

Having checked this thread, I am now about to try for the (guaranteed) crash and complete memory dump, which I hope I'll be able to duly report back to you. Here I go.

==============================

— First is set the Registry value to 0, as jaclaz indicated, to disable the automatic restart from the blue screen.

I used Everest. When I received the blue screen, I held down the Right Ctrl key and pressed Scroll Lock twice.

— "Beginning dump of physical memory"

— "Dumping physical memory to disk"

It took maybe 2 minutes.

THEN it did an automatic reboot into Windows. (So I assumed the registry value of 0 was not intended to hold it beyond the dump.)

— But when Windows was on, there was a second blue screen. This surprised me.

It counted to 100 again, all by itself.

This time, when it reached 100, I held down the Right Ctrl key while I pressed Scroll Lock twice. I thought maybe I had done something wrong the first time.

— Nothing happened.

I pressed Ctrl Alt Del. Nothing happened.

— Hard power off at PSU.

— Waited 20 seconds

— Power On

I heard a "two-tone" du-dip POST beep.

— Windows booted okay

— I waited; absolutely hands off

Another blue screen; like always, a Stop 50. This was the 3rd blue screen since I purposely had initiated what I thought would be just one.

— "Dumping physical memory to disk" to a count of 100

Power Off

Disconnect the IDE hard drive.

— Why? Because it seemed very likely that there was no other way to get into Windows.

Power On

du-doop 2-tone POST beep.

System booted up to Windows. I waited for a time, hands off. When all seemed okay, I looked for a memory.dmp file.

No such file. I asked Agent Ransack to check the whole computer. It found no such file anywhere on the system.

— BUT, there IS an i2api_debug.log in the C: \ root of some 11,446 kb.

Oh. And C:\adplus is empty.

I await further instructions. Yes, I have the 11,446 kb log, but as I said last night and earlier today, I'm unsure how to upload it—if indeed that file is any good to you.

jaclaz, I'm sorry that I have fallen off the tracks again. Actually, your earlier advice about following instructions missed the point.

I'm trying; I'm trying.

But as the advice given itself suggests, if these things were that straightforward to people like myself, we would not be here asking for help.

I'm quite willing to accept that I must have done something wrong, as these results seem to me to be far, far away from what the instructions suggested.

____________________________

Edit # 2

At 3am, I am adding this ASIDE: When I attempted to use ASCI on this XP-64 install, I had a lot of trouble. Nothing seemed to work, until I made an nLite disk. In nLite, I was successful. However, there was one setting which confused me: just where you go to add in ASCI, nLite contains a warning against using multiple drivers. There were two drivers available to me there: one for 32-bit and one for 64-bit. I would have selected both, but because of the warning, I considered the matter for a time and selected only the 64-bit driver.

— This is a Gigabyte motherboard. For XP, that file is in Gigabyte's download under SBDrv >RAID7xx, and I think it is the AHCIx64.inf file there. However, nLite definitely showed two files in any case; a 32- and a 64-bit file, and (as I said) I chose just the one because of nLite's warning at the very same place.

— I wasn't sure I was right about that. I am wondering now whether I should set the BIOS for all IDE-mode and see if that works. (By the way, for those who question my not just using IDE-mode anyway, well, I built the entire computer because I wanted to learn how to do it. I could have stayed with my old computer, which was terribly slow and outdated but which worked. Instead I made this my project for 2008, and I was darned if I was going to stop at ASCI, at the very very end of the project, because I didn't know how to do it.)

— However, I realize that such wild, unguided, shot-in-the-dark experimentation as re-trying IDE-mode might be offensive to jaclaz and cluberti, who must be wondering why I cannot come up with something as simple as a memory.dmp file.

I won't even mention that it is COLD in here, because the building manager thinks it should be—"put towels by the windows"—as the outside temperature has gone down to zero Celsius. So much for an $800 a month rental suite in Vancouver. Merry Christmas.

Edited December 24, 2008 by Groonx

[Groonx]

Update

Christmas Day

_______________

For those following this thread, at this point infiniti connected with me via chat, and he walked me through various possibilities related to his expertise in voltage and his familiarity with Gigabyte, high RAM, and other details.

His instructions were invaluable, because I was having trouble proceeding when I felt I might do real physical damage to my new system.

At this moment, I have all 3 hard drive on and working just fine. The various analysis programs are also working perfectly. The "groans" I described at startup, which were like this:

grrrr . . . . grrrr . . . . grrrr . . . . grrrr . . . .

have ceased. Basically the system is working ALMOST exactly right.

What is missing from the system is the CD-RW, which was Master (and later Slave) on the IDE cable shared by the PATA hard drive.

I had used the CD-RW to play and to burn disks (which it did flawlessly) in just the past few days.

But it was producing those startup "groans". We did not know whether the CD-RW was faulty, or whether the system was faulty and unable to accommodate the CD-RW. I'll be testing it out further on an older PC.

Infiniti had me adjust the voltage in the BIOS so that the System is running at [Normal] voltage, but the DDR2 voltage control is +0.3 volts to be 2.100, equal to this RAM's specified requirements.

I'm expecting there will be more for me to find out, still. But it's a new system, and the first one I have ever built. Maybe I'll be fine-tuning it for the next six months. My only real complaint is about myself—I find that my moods go WAY up or down according to computer failures and successes. I'm not bipolar, but the computer must be an extension of some part of my personality, because I just can't distance myself from the latest bad or good results.

Fortunately, with infinite help, I'm doing all right. You know, as the song says: "with a little help from my friends".

[jaclaz]

Knowing WHICH exact STOP ERROR comes on the BSOD is usually half to 3/4 of the solution.

— First is set the Registry value to 0, as jaclaz indicated, to disable the automatic restart from the blue screen.

I used Everest. When I received the blue screen, bla-bla, bla-bla, ...... bla-bla.

The original idea was that we knowing the STOP ERROR number might help.

If you now know it - thanks to the disbled automatic reboot - but don't actually post it, we are at the same point of the other day....

jaclaz

[Groonx]

Hi, Jaclaz,

As I said before, apart from knowing that all of them being Stop 50, I was not able to carry out the complete crash dump when I attempted it. I posted that above.

Yes, I did manage to stop the automatic reboot, so the blue screen was readable. Maybe you are telling me that the information I needed was displayed right there on that screen. But I wouldn't know. If I did, I would not have been asking. I think you steered me to that the other day: people ask questions because they don't know the answers. I can teach you how to recognize tanagers, but if you don't know what a tanager IS, then I am not being very helpful to you, am I?

In any case, there are no blue screens now. The voltage setting for the RAM was increased to fit the RAM specifications.

However, as you said, there is more to it than that. On a cold start, there is a 2-minute (silent) delay after the power button is pressed. The monitor waits in a power-save mode. After about two minutes, there is a 2-tone be-BEEP! and the system comes up with everything seeming to work just fine. I can use all the analysis apps, and none are showing any errors. But the 2-minute wait is wrong, and so is that double-beep. I have looked on Gigabyte's POST error codes, but not found it yet.

[jaclaz]

Well, there is no such thing as "Stop 50".

A BSOD error looks like this:

It is composed of the main error code (0x00000050) plus a number of added parameters of subcodes:

0xFD3094C2,0x00000001,0xFBFE7617,0x00000000

in the example. plus, in some cases, additional info:

*** SPCMDCON.SYS - Address FBFE7617 base at FBFE5000, Datestamp 3d6dd67c

in the example.

Though this might seem "Chinese" to you (and to me ) the fact that in their contorted minds the MS programmers wrote a routine to actually display those apparently senseless numbers, should mean that someone can actually understand something more than just "error 50".

I was merely trying to help you in getting helped hopefully one or two steps beyond the "I'm ill doctor, help" stage:

http://homepages.tesco.net/J.deBoynePollar...ard-litany.html

jaclaz

[Groonx]

Okay, thanks. Cluberti said the other day:

All the bugchecks are Stop 0x50's

and I recalled that as a Stop 50, instead of a Stop 0x50.

Well, back to my tanager analogy, if you referred to Tanagra, I would still be able to make that leap. Not so in the computer world, I suppose.

By the way, I know you love that "litany", and I know you are trying to be helpful, and so is it. But the litany badly fails some of the primary rules of communications theory. It assumes, for just one example, that the novice explaining the problem will not "helpfully" write three pages of details in trying to conform to what the litany asks. My own background, my friend, is pure science; the university courses I instructed included both communications and applied logic.

— If I knew enough to ask you questions precisely as you would like, I would have to know as much as you; and I would not be here asking you

— The First Rule of Communications must always be: What the listener/student fails to comprehend is the communicator/instructor's fault—not the listener/student's fault

Yes, I photographed blue screens like that one. I thought you required the complete dump record, which I failed to get.

Technical Information:

STOP: 0x00000050 (0xFFFFFAE01C772839, 0x0000000000000000, 0xFFFFFADF2F5B1290, 0x0000000000000005)

disk.sys -- Address FFFFFADF2F5B1290 base at FFFFFADF2F5AD000, DateStamp 45d69708

Beginning dump of physical memory. . . .

Edited December 26, 2008 by Groonx