XP stuck right before Welcome screen

[taz420nj]

This one is really pissing me off. I had this machine come in exhibiting random blue screens. First thing I did was run Memtest, which revealed a bad stick of RAM. Installed a new stick, ran Memtest for 12 hours to rule out bad memory controller with no red. Went to boot it normally, and it got to the blue logo screen and won't go any further. The numlock light does not respond to presses and mouse does not move. There is zero hard drive activity about 2 minutes after the screen loads.

Booted to ERD and did CHKDSK - which revealed some file issues but no bad sectors. Then did SFC and found a bunch of corrupted files - repaired and rebooted.. Still does it. Same thing when booting to safe mode. Sticks in same place. Only difference is something is driving the CPU temp up in safe mode, because after a minute or two the fan starts ramping up.

Any ideas? Searching only turned up people who ran into this after cloning drives, but those didn't lock up solid (numlock still responded). Nothing was changed other than the bad RAM stick.

 

[qubit]

I think it might be time to start over with a quick format and clean install.

I've had repeated hardware failures before and over time it causes so much damage to the system files and registry that there's just no saving it, even with a repair install using the original setup disc. On top of that, XP is a lot more prone to this corruption than later Windows versions. Any chance of installing W7 or W10?

Make sure your data is backed up first.

 

[taz420nj]

I think it might be time to start over with a quick format and clean install.

I've had repeated hardware failures before and over time it causes so much damage to the system files and registry that there's just no saving it, even with a repair install using the original setup disc. On top of that, XP is a lot more prone to this corruption than than later Windows versions. Any chance of installing W7 or W10?

Make sure your data is backed up first.

It's a customer's machine and I've never come across this particular issue before. I hate doing clean installs unless it's absolutely completely borked - too many shops wipe/reinstall for any problem because they're too lazy to actually solve the problem. I was going to try a repair install if nobody had any other suggestions. And I always take an image of the drive before I do anything.. Cheap insurance lol!

No chance whatsoever of W10, and I honestly wouldn't feel right charging someone for a W7 license to put on an 11 year old P4 machine.

 

[GoldenX]

Linux it is then xD. Or the latin-american/cheap way...

It seems normal damage to boot system files, if the CHKDSK and CD repair can't fix it, the best solution is a clean install.

 

[qubit]

It's a customer's machine and I've never come across this particular issue before. I hate doing clean installs unless it's absolutely completely borked - too many shops wipe/reinstall for any problem because they're too lazy to actually solve the problem. I was going to try a repair install if nobody had any other suggestions. And I always take an image of the drive before I do anything.. Cheap insurance lol!

No chance whatsoever of W10, and I honestly wouldn't feel right charging someone for a W7 license to put on an 11 year old P4 machine.

Ok, so it's a customer's machine, I see what you're saying.

Unfortunately, I can't think of anything else to recommend but a clean install and I've been in this situation many times, especially with XP. Do you have the product key and disc for XP to do a clean install? If not, then you can get this utility which may be able to read it off the disc.

www.magicaljellybean.com/keyfinder

Good one on that image backup.

 

[micropage7]

This one is really pissing me off. I had this machine come in exhibiting random blue screens. First thing I did was run Memtest, which revealed a bad stick of RAM. Installed a new stick, ran Memtest for 12 hours to rule out bad memory controller with no red. Went to boot it normally, and it got to the blue logo screen and won't go any further. The numlock light does not respond to presses and mouse does not move. There is zero hard drive activity about 2 minutes after the screen loads.

Booted to ERD and did CHKDSK - which revealed some file issues but no bad sectors. Then did SFC and found a bunch of corrupted files - repaired and rebooted.. Still does it. Same thing when booting to safe mode. Sticks in same place. Only difference is something is driving the CPU temp up in safe mode, because after a minute or two the fan starts ramping up.

Any ideas? Searching only turned up people who ran into this after cloning drives, but those didn't lock up solid (numlock still responded). Nothing was changed other than the bad RAM stick.

have you tried changing the sata cable or another port, is there any clicking that unusual. could be from cable or hdd itself
you may hook the hdd to another pc and back up the data first

maybe the best way it tell your customer that it has a problem so they know it

 

[MIRTAZAPINE]

I would take the precaution to backup any data from the hard drive now. You don't know if this can be a HDD failing or the motherboard sata controller(south bridge) is dying out corrupting the windows install. I had something similar on my pc but with slightly different symptoms where windows xp keep loading and loading but could not get to desktop. That was on an overclock motherboard. Strangely it went away a few months later when I come back to it. I have the PC with that problem now that I reassembled again and shown on the nostalgia thread here because I have no pc now. Just like your systems hang and blue-screen despite at stock. Strong suspicion the motherboard is borked in a way. I tested my HDD on another pc is is healthy with smart test and files are transfered.

You would be lucky if it is just the HDD that need replacement or just needing a clean install. The motherboard problem is much harder to fix. It would be huge pain to solder those surface mount chips. Try re-pasting the chip to cool it down more to see if it helps

 

[Jetster]

Unhook any hardware not essential to run. If it still doesn't boot then the system is corrupt or bad sata cable. Clean install

 

[taz420nj]

I've already imaged the drive. I use Reflect, which halts if it finds bad sectors. All of the errors that CHKDSK found were NTFS problems, presumably from when it crashed. I'm fairly certain this isn't a hardware issue because hardware issues are random - this is locking up at the exact same spot every time, which indicates a software glitch.

 

[Ferrum Master]

Maybe a poor SATA cable?

 

[taz420nj]

I would take the precaution to backup any data from the hard drive now. You don't know if this can be a HDD failing or the motherboard sata controller(south bridge) is dying out corrupting the windows install. I had something similar on my pc but with slightly different symptoms where windows xp keep loading and loading but could not get to desktop. That was on an overclock motherboard. Strangely it went away a few months later when I come back to it. I have the PC with that problem now that I reassembled again and shown on the nostalgia thread here because I have no pc now. Just like your systems hang and blue-screen despite at stock. Strong suspicion the motherboard is borked in a way. I tested my HDD on another pc is is healthy with smart test and files are transfered.

You would be lucky if it is just the HDD that need replacement or just needing a clean install. The motherboard problem is much harder to fix. It would be huge pain to solder those surface mount chips. Try re-pasting the chip to cool it down more to see if it helps

It's not blue-screening, it's just hanging at the logo screen (with the blue background) that comes up right before the welcome screen. The last mini-dump is from the last crash the customer had before he brought it to me. It's not seeing that it didn't start correctly, nor is it seeing that it didn't shut down correctly (not getting an automatic safe mode prompt at boot). It is not creating new mini-dumps, as I have to hard power-off. He did not mention that it hung at boot, so I don't want to tell him until I have exhausted all other possibilities. It certainly wouldn't be the first time someone neglected to mention something but for the time being I have to treat this like it was from something I did.

And I don't repair motherboards. I don't have a reflow station or the equipment to test for faulty components. If I determine a motherboard is borked, they buy a new or used one.

 

[taz420nj]

Maybe a poor SATA cable?

Like I said, it's hanging in the exact same spot every time. Pretty much rules out any hardware.

 

[Ferrum Master]

Like I said, it's hanging in the exact same spot every time. Pretty much rules out any hardware.

Well NTFS errors should have a source... if those are not bad sectors... usually it is bad ram or SATA cable in my books... it gets corrupted same way on each boot...

 

[MIRTAZAPINE]

Like I said, it's hanging in the exact same spot every time. Pretty much rules out any hardware.

I would not rule hardware out that quickly. Hardware problems can happen without any blue screen. And they can be silent.

 

[taz420nj]

Well NTFS errors should have a source... if those are not bad sectors... usually it is bad ram or SATA cable in my books...

It DID have a bad stick of RAM - that was the initial diagnosis. I did not try to boot to Windows at all before I started. First thing I did was boot Memtest86 because random BSODs are usually from faulty RAM. I replaced the bad stick and THEN I tried to boot it and now have this issue.

it gets corrupted same way on each boot...

I would not rule hardware out that quickly. Hardware problems can happen without any blue screen. And they can be silent.

It's impossible for a bad cable or bad RAM to cause it to hang in the exact same spot every single time. Hardware problems are very rarely reproducible on demand (bad sectors would be one of the exceptions, but that has also already been ruled out). As I said, I suspect whatever this is was a result of the crashes caused by the bad RAM, but as of now it is no longer related to hardware.

 

[Ferrum Master]

I remember XP often got screwed up like that while OC'ing, it could not recover itself, way too dumb... That's why I gladly ditched it. It is a faulty driver/system DLL that is inevitably with bad crc and causes BSOD by doing haywire...try to do sfc /scannow, that's for system core... if it is a driver dll, you are screwed. NUKE the OS...

 

[taz420nj]

I remember XP often got screwed up like that while OC'ing, it could not recover itself, way too dumb... That's why I gladly ditched it. It is a faulty driver/system DLL that is inevitably with bad crc and causes BSOD by doing haywire...try to do sfc /scannow, that's for system core... if it is a driver dll, you are screwed. NUKE the OS...

No offense dude but did you read my OP at all? You're telling me everything I already did. And this is a Dell Optiplex with a stock BIOS. It isn't overclocked.

This one is really pissing me off. I had this machine come in exhibiting random blue screens. First thing I did was run Memtest, which revealed a bad stick of RAM. Installed a new stick, ran Memtest for 12 hours to rule out bad memory controller with no red. Went to boot it normally, and it got to the blue logo screen and won't go any further. The numlock light does not respond to presses and mouse does not move. There is zero hard drive activity about 2 minutes after the screen loads.

Booted to ERD and did CHKDSK - which revealed some file issues but no bad sectors. Then did SFC and found a bunch of corrupted files - repaired and rebooted.. Still does it. Same thing when booting to safe mode. Sticks in same place. Only difference is something is driving the CPU temp up in safe mode, because after a minute or two the fan starts ramping up.

Any ideas? Searching only turned up people who ran into this after cloning drives, but those didn't lock up solid (numlock still responded). Nothing was changed other than the bad RAM stick.

 

[Ferrum Master]

No offense dude but did you read my OP at all? You're telling me everything I already did. And this is a Dell Optiplex with a stock BIOS. It isn't overclocked.

It happens from getting old, bad contact in old DIMM, cold joints etc... so the corruption occurs... due to that ECC RAM exists actually. You sure the RAM stick was really BAD? Or you just bent the board while changing sticks, thus fixing the cold joint for a while? As I said SFC fixes only system dll's not the 3rd parity, like GPU, Sound etc driver that are not accessed all the time... if such corrupted DLL loads, it writes into wrong memory space and thus again it can screw things up... it is XP.

 

[taz420nj]

It happens from getting old, bad contact in old DIMM, cold joints etc... so the corruption occurs... due to that ECC RAM exists actually. You sure the RAM stick was really BAD? Or you just bent the board while changing sticks, thus fixing the cold joint for a while? As I said SFC fixes only system dll's not the 3rd parity, like GPU, Sound etc driver that are not accessed all the time... if such corrupted DLL loads, it writes into wrong memory space and thus again it can screw things up... it is XP.

It doesn't take ECC, it's standard unbuffered PC2-4200. I've tried removing all the RAM and switching it around, and running only one stick. It's NOT hardware related. Please stop trying to convince me that it is.

 

[Ferrum Master]

It doesn't take ECC, it's standard unbuffered PC2-4200. I've tried removing all the RAM and switching it around, and running only one stick. It's NOT hardware related. Please stop trying to convince me that it is.

Sorry I am putting a curse upon you... anyways you had a hardware error RAM and have a random black screen and a corruption afterwards, and you suggest it is still software? Nada won't happen.... The ECC ram was just a hint why they exist as such... for these rare matters once in a year.

 

[qubit]

@taz420nj Reading your OP again, I think you should try to rule out more bad hardware, even though you're convinced it's not.

Simply grab anoher HDD, install XP on it and see if it boots properly. If not, then you can take it from there. Won't even take very long. Then, load the the chipset, sound and video drivers and see if it still works properly.

Also, before you do this, it's a good idea to clear the CMOS RAM (as in remove the battery and load defaults) since this can sometimes get corrupted in subtle ways which bork the PC without any of the settings looking wrong. Then load sensible defaults and manually configure it. You'll still likely have to do that clean install to fix the software side of the problem.

I'm sorry that there are no shortcuts here with this type of problem and it's gonna take time. If the motherboard is bad then your customer might as well chuck the computer and get a new one as it's long obsolete. Certainly it doesn't appear to be worth all that time to work on it, from your perspective as you're charging for the work.

 

[taz420nj]

Sorry I am putting a curse upon you... anyways you had a hardware error RAM and have a random black screen and a corruption afterwards, and you suggest it is still software? Nada won't happen.... The ECC ram was just a hint why they exist as such... for these rare matters once in a year.

You have it totally backwards. The hardware problem HAS BEEN SOLVED. It is NOT giving random BSODs anymore, it is getting to the EXACT SAME SPOT and hanging every single time. There is absolutely NO WAY that an unsolved hardware problem (again aside from bad sectors which again have already been ruled out) can cause it to hang in the EXACT same spot EVERY SINGLE TIME. This corruption WAS caused by the ORIGINAL hardware problem. The ONLY way a problem can be repeatable with 100% certainty in the EXACT SAME SPOT EVERY SINGLE TIME is if it is a glitch in software.

And the fact that it happens in Safe Mode too rules out any 3rd party drivers because those don't load in safe mode.

 

[taz420nj]

@taz420nj Reading your OP again, I think you should try to rule out more bad hardware, even though you're convinced it's not.

Simply grab anoher HDD, install XP on it and see if it boots properly. If not, then you can take it from there. Won't even take very long. Then, load the the chipset, sound and video drivers and see if it still works properly.

Also, before you do this, it's a good idea to clear the CMOS RAM (as in remove the battery and load defaults) since this can sometimes get corrupted in subtle ways which bork the PC without any of the settings looking wrong. Then load sensible defaults and manually configure it. You'll still likely have to do that clean install to fix the software side of the problem.

I'm sorry that there are no shortcuts here with this type of problem and it's gonna take time. If the motherboard is bad then your customer might as well chuck the computer and get a new one as it's long obsolete. Certainly it doesn't appear to be worth all that time to work on it, from your perspective as you're charging for the work.

It loads alternative boots just fine. It loaded Hiren's via PXE to do the initial diagnostics (edit: I just tried, it also loads MiniXP just fine), it loaded Reflect via PXE to do the drive image, it loaded ERD via live USB to run the check utilities.. It is NOT hardware.

 

[qubit]

Did it boot MiniXP off the original HDD? Did you try loading the drivers?

Sorry, I know that you're convinced that the hardware isn't faulty - and it may well not be - but you haven't ruled out everything. Those other things you've tried do suggest it's ok, but you have to totally eliminate it as a suspect before you move on. You do that by following the advice that I gave you in post 21. As you've already taken an image of the HDD it will be trivial for you to try this out and eliminate the suspect. This is especially important as it's an ancient machine so hardware faults are quite likely; you already found bad memory, so there could be more faults.

Also, you must remember that we're not sitting here with the machine in front of us, so we can't get a sense of it from the subtleties of its behaviour, unlike yourself, so you've got to work with us if you want us to help you succesfully.

 

[taz420nj]

I'm not going to waste an hour RE-ruling out something that I know is not the problem because I've already ruled it out. It is a problem in WINDOWS. Why can't someone just help me figure out what it is? And I realize the machine is not in front of you... Thats why I described exactly what it is doing: It gets to EXACTLY the same spot (the blue logo screen right before it says "Windows is starting up") then hangs every single time. Not a BSOD, not a restart, it just hangs. A hardware fault can NOT hang the system in the same exact spot every single time. Bad sectors have been scanned for and not detected on the hard drive, the SMART test passed. I have rescanned the drive with CHKDSK /f /r, and re-run SFC, and no further corruption was found.

It is a problem with Windows. Period. I have tried enabling boot logging to figure out what is hanging it, but the log is not written to disk until the registry finishes loading - which means it doesn't get written since it is hanging sometime DURING the registry load. Catch-22.