Blue Screen View Output

[Mike McCall]

As I said in a previous thread here, the customer insists on fixing rather than replacing the system. It's been Blue Screening for awhile she said, though it didn't for me until last night. My suspicion is that since it's over 4-years without updates everything's fairly well knackered, as the Brits say, and a nuke & pave will be required to get it sorted. The problem is I'm not sure it's strictly a software problem and I'm not equipped to do board level testing.

When the system crashes I don't always get the blue screen. It often just goes black, USB ports are dead, and the PS is still on. It's an HP Touchsmart 300-1120 AIO.

I'm uploading the BSV output text file Can I get some more eyes on this?

 

[Porthos]

I would get the drive replaced with a clean install and also run memtest overnight. Of course while you have it open for the drive replacement blow out the dust bunnies.

When I question a software or hardware issue I use a spare SSD to install Windows on with updates and drivers updated and run it for a couple of days. If ok I image it to the drive I purchase for the client and move on. If that does not fix the issue it is time for the replacement computer.

I also try a different power supply before the towel is thrown in.

 

[Markverhyden]

Analyzing BSD's from a machine like that is like trying to determine how a human died 1000 years ago by analyzing the skeleton. I've spent more hours than I care to remember trying different tools to parse dump files only to realize that they are mainly for devs troubleshooting app issues.

Like @Porthos said pop in another drive, preferable an SSD, install clean, update and leave running and testing. I will say that those machines are problematic. Back in the day, when under warranty HP used to authorize onsite repairs for desktops, which included AIO's. They quickly found out that was a loose/loose proposition so they switched to ship to depot only.

 

[Mike McCall]

Analyzing BSD's from a machine like that is like trying to determine how a human died 1000 years ago by analyzing the skeleton. I've spent more hours than I care to remember trying different tools to parse dump files only to realize that they are mainly for devs troubleshooting app issues.

Like @Porthos said pop in another drive, preferable an SSD, install clean, update and leave running and testing. I will say that those machines are problematic. Back in the day, when under warranty HP used to authorize onsite repairs for desktops, which included AIO's. They quickly found out that was a loose/loose proposition so they switched to ship to depot only.

Being a bit obsessive, I looked up each one. A clear majority point to driver problems and file corruptions. A couple point to the hard drive, and a couple point to a possible memory problem. Though not definitive, I don't see anything pointing to the hardware itself. Just looking to avoid doing the work only to fine the patient is dead.

I don't have a spare SSD, but I do have an extra drive set aside for such things. The device is rather clean inside with both fans and heat sink fins clean and shiny.

 

[mr m]

That HP All In One should have onboard diags if the partition was left unmolested over the years. You can run the short or quick tests first but if they pass run the extensive tests for multiple passes overnight.

I just found a bad HDD after 2 hours of stress testing an HP AIO. I suspected a failing HDD after task manager showed 100% disk usage. I usually fix that easily with Process Explorer but not this time. None of my diags (HDDScan, GSmartcontrol) picked up the HDD failure, only the HP embedded drive stress test found it. SMART data showed "good".

If it passes I would suspect a driver issue but like Mark alluded to, analyzing those dump files are a lesson in futility.

 

[ComputerRepairTech]

It's an HP Touchsmart 300-1120 AIO.

Processor: AMD Athlon II X2 235e

...they are going to end up paying way more than that system is worth.

 

[Porthos]

Just looking to avoid doing the work only to fine the patient is dead.

No one likes to do extra work. But that is what a diag fee is for. If the paitentis dead and they do not want to get a replacement from me or at least the setup and data transfer from me I charge $59 for diag before they can pick it up.

Another reason I install Win 10 exclusively now is that loading 10 on my spare SSD (updated image) and loading drivers with SDI can be done to completion in about 15 minutes 30 max. This $25 ssd is affordable to have for just this issue. https://www.amazon.com/gp/product/B076XMH2JT/ref=oh_aui_detailpage_o00_s00?ie=UTF8&th=1

Many times I will use it to prep a computer while I wait for a drive to come in for the customer.

 

[glennd]

My suspicion is that since it's over 4-years without updates everything's fairly well knackered, as the Brits say, and a nuke & pave will be required to get it sorted.

I would have thought the contrary. No updates means the driver set is in the same state it was four years ago. If it was working four years ago it should be working today. So you can discount drivers as the problem. I would be leaning towards hardware, most likely a failing hard drive possibly causing file corruption or read errors in the HAL resulting in the blue screens. If this is true, a N&P may well kill the drive. Time for an image backup!

When the system crashes I don't always get the blue screen. It often just goes black

Memory corruption can do that. Could be ram modules or corrupted OS files. I'm betting on the HDD.

 

[Porthos]

I'm betting on the HDD.

Already said the drive has issues.
So a drive and a clean install to eliminate corruption could be all it needs. That is why I usually test with the spare before I order a drive just in case it is a bigger issue.

CrystalDisk says the drive is cautionary for several issues.

 

[Mike McCall]

That HP All In One should have onboard diags if the partition was left unmolested over the years. You can run the short or quick tests first but if they pass run the extensive tests for multiple passes overnight.

I just found a bad HDD after 2 hours of stress testing an HP AIO. I suspected a failing HDD after task manager showed 100% disk usage. I usually fix that easily with Process Explorer but not this time. None of my diags (HDDScan, GSmartcontrol) picked up the HDD failure, only the HP embedded drive stress test found it. SMART data showed "good".

If it passes I would suspect a driver issue but like Mark alluded to, analyzing those dump files are a lesson in futility.

It passed all the onboard tests without issue.

Processor: AMD Athlon II X2 235e

...they are going to end up paying way more than that system is worth.

Yeah, tried to talk them out of it but...no. Goes that way sometimes.

No one likes to do extra work. But that is what a diag fee is for. If the paitentis dead and they do not want to get a replacement from me or at least the setup and data transfer from me I charge $59 for diag before they can pick it up.

Another reason I install Win 10 exclusively now is that loading 10 on my spare SSD (updated image) and loading drivers with SDI can be done to completion in about 15 minutes 30 max. This $25 ssd is affordable to have for just this issue. https://www.amazon.com/gp/product/B076XMH2JT/ref=oh_aui_detailpage_o00_s00?ie=UTF8&th=1

Many times I will use it to prep a computer while I wait for a drive to come in for the customer.

Oh, I charge a diagnostic fee ($45), then waive it if they have me do the work.

Uh, that's really cheap for an SSD. They any good? Who really cares at that price I guess.

I would have thought the contrary. No updates means the driver set is in the same state it was four years ago. If it was working four years ago it should be working today. So you can discount drivers as the problem. I would be leaning towards hardware, most likely a failing hard drive possibly causing file corruption or read errors in the HAL resulting in the blue screens. If this is true, a N&P may well kill the drive. Time for an image backup!

Memory corruption can do that. Could be ram modules or corrupted OS files. I'm betting on the HDD.

The drive has issues and is being replaced as stated above. I have no doubt the system files are corrupted. But you're right about the memory. I stuck another drive in it and started loading the OS when everything just went black and quit again. Tossed in Memtest86 and it ran for a bit, then just lit up with red failures. So, the client gets a memory upgrade (4GB - 8GB), a new hard drive (refused SSD), and a bunch of labor from me. Haven't seen memory fail with no beep codes and passing the onboard diagnostics, but it does explain much of what I'm seeing. Lesson learned. Nevertheless, it doesn't change anything except add a memory upgrade.

 

[Mike McCall]

BTW, the client wrote me a letter telling me the problems she was having. She also took pictures of the screens with her camera, had them printed (3 1/2 x 6 full color), and glued them in place on two pages.

 

[Porthos]

Uh, that's really cheap for an SSD.

I used to use sandisk plus for my clients and moved to these after a post here on the forums. No issues so far. The price has gone down since I first started using them. I only use 240 and the 480 gig ones for clients though. 120 is more than enough for testing like I do.

 

[Markverhyden]

Being a bit obsessive, I looked up each one. A clear majority point to driver problems and file corruptions. A couple point to the hard drive, and a couple point to a possible memory problem. Though not definitive, I don't see anything pointing to the hardware itself. Just looking to avoid doing the work only to fine the patient is dead.

I don't have a spare SSD, but I do have an extra drive set aside for such things. The device is rather clean inside with both fans and heat sink fins clean and shiny.

One needs to understand the sequence of events in dump files. Those entries occur because the offending item(s) occurred within the OS. That means the OS must be stable see the process(es) to log them. About the only hardware items reported I regularly see are HD's, usually bad blocks or delay read/writes. An underlying hardware component problem does not usually produce a log entry. It just tanks the machine with no notice.

 

[Mike McCall]

One needs to understand the sequence of events in dump files. Those entries occur because the offending item(s) occurred within the OS. That means the OS must be stable see the process(es) to log them. About the only hardware items reported I regularly see are HD's, usually bad blocks or delay read/writes. An underlying hardware component problem does not usually produce a log entry. It just tanks the machine with no notice.

Thanks, I get it. I don't see a lot of this so I don't fool with it much. I was trying analyze where to look for the problem from the information in the dump files. While it didn't send me in the wrong direction, it didn't get me any closer either. In fact, it wasn't until it crashed when installing the OS on a different drive that I thought either memory or board level issue. I'm sure for those who see memory failures more frequently than I do they would have found it long before I did. Got to learn something in this and I'm thankful for that.

 

[Porthos]

t wasn't until it crashed when installing the OS on a different drive that I thought either memory or board level issue.

That is when I run Memory tests.

I'm sure for those who see memory failures more frequently than I do they would have found it long before I did.

Memory is not first guess till a clean install fails. This is where I hate AIO's because the next step before I cal it a board issue is swapping the AC adapter and those I don't have spares of.

 

[Mike McCall]

Memory is not first guess till a clean install fails. This is where I hate AIO's because the next step before I cal it a board issue is swapping the AC adapter and those I don't have spares of.

No adapter, regular cord and socket.

 

[Porthos]

No adapter, regular cord and socket.

That sucks even more on an AIO. Good luck, for your sake and the customers hope there is a memory issue.

 

[Mike McCall]

That sucks even more on an AIO. Good luck, for your sake and the customers hope there is a memory issue.

I'm certain of three things with this machine:

1. The memory is bad and needs to be replaced.
2. The hard drive needs to be replaced.
3. The OS should be wiped & reinstalled.

I don't think the PS is an issue at this point because when it crashes the power remains on, hard drive spins, and I think the fans do too (hard to hear). Hoping there's no board level issues, but I don't think there are at this point.

The biggest reason I don't like this project is because the money spent to repair this is twice what one could replace it for. I've done my best to convey that to the customer, but they insist on fixing this one. Her comment was something like; "I spent $700 and it's not even 10-years old yet." Ugh.

 

[Porthos]

The biggest reason I don't like this project is because the money spent to repair this is twice what one could replace it for. I've done my best to convey that to the customer, but they insist on fixing this one. Her comment was something like; "I spent $700 and it's not even 10-years old yet." Ugh.

A fool and thier money are soon parted.

 

[Mike McCall]

A fool and thier money are soon parted.

I don't know about foolish, lots of people make choices I wouldn't for their own reasons and are not fools. Older people truly don't understand technology or the rate at which it becomes obsolete. Their decisions are often based on reference points having nothing to do with technology, like a car or a house. I've learned to be less judgmental about customer's sensibilities. I certainly used to be.

I gave the customer honest, accurate, and straight forward information so they could make an informed decision. That's all I can do. They determined that spending over $300 to repair their machine was good value compared to the cost of replacing it. I wouldn't have made that decision in this case. What I will do from here is ensure the work is done properly and the customer is as happy as I can make them.