Is This Even A Software Problem?

Hi! A friend recommended I make a post here to ping you guys and see if anyone could help me. Apologies for any potential ignorance up front, though I do think I am at least a little versed in computer maintenance.

So yesterday morning I had a very, very bad time. It follows a sequence of bad times over the last two weeks, give or take, though really, at this point, it's honestly felt like ever since I built my first PC in 2007 (as opposed to buying them pre-assembled from the store), I've been kind of cursed.

I will warn you now: This is going to be a long read. Some of it may seem redundant, but I figure all possible information will help identify what the problem is going forward and figure out possible solutions. Before we begin, here is my system configuration:

Motherboard: ASRock Z97 Extreme6
CPU: Intel Core i5-4690K (Devil's Canyon)
RAM: G.Skill Ripjaws DDR3 16gb (8gb x 2)
GPU: GeForce GTX 1060 6GB
PSU: EVGA SuperNova 750w G2
OS: Windows 10 Pro
Optional PCIe Expansion: Elgato HD60 Pro (I have a Youtube channel that's on the low end of popularity)

We'll start by rolling the clock back two weeks ago, since that's when the relevant issues began:

• 9/29/2017: I hook my computer tower up in the spare bedroom at my brother’s place where I’m living until we can move in to our new apartment. It has been shipped from Colorado to Nevada in a moving trailer with as much padding as I could find. It probably still wasn’t enough. The system boots up fine, but very slowly, and software (especially games) stutters. I shut it down and touch a few connections inside the system, making sure the RAM, graphics card, and HDD connectors are firm. It’s not very thorough as far as checks go, but I start the system back up anyway. Stuttering is still there. The system doesn’t have internet yet, because even though I bought a $15 TP-Link USB Wifi adapter back in February or March in anticipation of this moment, I absentmindedly left it in storage. Storage doesn’t have lighting inside of the units, so finding the Wifi adapter will have to wait until the sun comes back out. At this point I think maybe the stuttering will go away once I connect it to the internet, since some games act funny (or don't work at all) when they can’t connect to the web.

• 9/30/2017: With the TP-Link USB Wifi adapter installed, the system updates everything, but the stuttering problem persists. Friends urge me to open the system and do a full check of all the connections. Some suggest taking all removable components out and reconnecting them. I don’t do that, but I do find that the CPU cooler is loose and that the backplate on my GTX 1060 is also loose (this is me pushing down on the metal plate, which has separated from the 1060′s chassis a little bit, allowing the loose screws to raise up). I take the 1060 out and gently tighten all the screws so the backplate doesn’t shift around anymore, and I manage to snap 3 out of 4 of the CPU cooler pins back in to place (it’s one of those generic, stock Intel coolers). The 4th one kind of snaps in, but it’s a little mushy and these coolers are fragile so I figure it’s good enough. Amazingly enough, the computer starts up much faster and all stuttering in games is gone.

• 10/1/2017: A new problem has emerged: when the system sits overnight while powered off, on first boot it will show the BIOS logo, POST, and then instead of loading Windows, it gets stuck on a black screen. While on this black screen, if I hit the reset button, the system boots normally. One friend suggests a power supply issue. I contact EVGA, my PSU manufacturer, who claim to have a 10 year warranty. EVGA says that doesn’t sound like a power supply issue to them, and we start talking about what it could be.

• 10/2/2017: I have this strange paranoia about the power cable I’m using for the tower and whether or not the surge protector is too old. I don't have any alternative options for a power cable (I'd have to buy one), but I grabbed a newer surge protector from storage thinking that may also fix the stuttering problem. Since it didn’t fix that nor the black screen problem, I plug the computer directly in to the wall outlet to see if that changes anything and joke to myself, “hopefully this isn’t a bad idea.” Is it a bad idea? I can’t quite remember.

• 10/3/2017: Out of options, EVGA suggests I just disassemble the entire PC besides the CPU and see if the black screen still happens. If it doesn’t, I'm supposed to reconnect the entire computer component-by-component until it happens. That will tell me what's causing the issue and I simply replace that component. That sounds like a lot of scary work (I hate opening my system up), and given how small this spare bedroom is, I don’t have a lot of room for that kind of surgery. I’ll have to psyche myself up for it.

• 10/4/2017: Suddenly I realize: the first time the system started up here in Nevada, it didn’t have the black screen problem. It was only after I connected the TP-Link USB Wifi adapter. I pull it out of the system before I start it up, and sure enough, it boots straight in to Windows. A quick Google search reveals others, with nearly identical models of TP-Link adapters, suffering the EXACT same problem. It’s a long standing hardware conflict with a Windows 10 USB 3.0 Controller and TP-Link devices that neither are interested in fixing (and at least for some, seems to cause a variety of USB problems until they got rid of the device). This guy says his TP-Link device actually damaged a USB 3.0 port of his because of this problem. A friend suggests an internal PCIe Wifi card. It’s $60, which is a lot for someone who is still technically homeless, but you get what you pay for, and the last time I cheaped out with the TP-Link adapter, it bit me in the butt – so I go for it.

• 10/7/2017: The ASUS PCIe Wifi card arrives, I put it in, and it works with zero problems. The day is saved. Or is it?

• 10/11/2017 (12:45am): Around midnight I go to the bathroom and return to find my TV is turned off. The way things are set up isn't normal -- usually have an ASUS monitor, but the base is too big for the table I'm using, so I've supplemented it with a Vizio 1080p TV over HDMI. Roughly the same size as the monitor, but since it's just got this wedge in the back it's propped up on, it takes up less room. What this means is that when my desktop tells my TV to enter standby mode due to inactivity, the TV just flips over to “No Signal” and eventually turns off after 30 seconds. Absentmindedly, I move the mouse to wake the display, but forget to to turn the TV on at first, and by the time I get the TV on and past the Vizio logo, the computer is already mid-reboot for some reason. Seems like coming out of standby with no display may have crashed the video card. Surely it’ll come right back on.

• 1:05am: Windows has been stuck on a loading spinner for close to 20 minutes. I’m getting worried and looking up stories from people who left their system sit on this same spinner for hours, even days, with no progress. The HDD activity light hasn’t blinked in a while. The system is just sitting there, idling, doing as far as I can tell, nothing. No messages about installing a Windows update or anything like that. Some solutions for this eternal loading screen say to just shut it off and try again, even though that may be risky. So, I throw caution in to the wind and go for it.

• 1:15am: It takes at least another five minutes of loading spinners, but the system finally boots. Seems like Windows may have rolled back to a restore point as some icons have been moved around on the desktop to where they used to be earlier in the night, but Chrome automatically pops back open on the page it was on when I went to the bathroom. The Event Viewer offers no insight on what caused the system to reboot, nor what it was doing when it was stuck on the loading screen. Not only that, but my ASUS PCIe Wifi Card is gone. Windows is complaining about there being no ethernet connection, which it shouldn’t be doing. Checking the device manager, the ASUS wifi card is there, but it’s saying there “aren’t enough free resources” for it to function. Code 12. A Google search on my tablet says this means it’s run out of IRQ slots. What? It wants me to disable other devices on my system to make room. Does that mean something with my Elgato or the 1060? Given this was apparently a video error, I’d say something happened to the 1060. The day before, GeForce Experience had notified me of new drivers and I ignored it because I was in the middle of something. Maybe I stumbled upon an IRQ bug they patched?

• 1:35am: I uninstall, reinstall, disable, and renable the PCIe wifi card repeatedly. I get out the CDROM that came with the PCIe card and install the “official” ASUS drivers instead of whatever Windows thinks it needs (Windows says it’s a Broadcom device; it’s not). Nothing changes: every time I reboot, it’ll say wifi connections are available, but when I connect to my brother's router, I get wifi for a split second and then the device disappears and stops functioning. I’m considering downloading clean Nvidia drivers on a USB stick using my laptop to see what that does.

• 1:50am: I’ve run the Windows Hardware Troubleshooter. It states the obvious: hey, your wifi card’s not working. It claims to do some magic behind the scenes but nothing works. The problem evolves and the Hardware Troubleshooter starts saying the wifi card’s drivers might be faulty, even though five minutes ago they were fine. I'm thinking that all I need is those Nvidia drivers. I plug in the TP-Link USB Wifi adapter, knowing that’ll give me internet long enough to download the driver update. This was a bad idea – When I got the ASUS card, I’d uninstalled the TP-Link device completely, and I think it needed the drivers disc before you plugged it in to the USB port. Windows seems to summon drivers from somewhere, for something, and instantly the whole system is brought to its knees and eventually BSODs with a DPC_WATCHDOG_VIOLATION. Looking that up, it apparently happens when a piece of software leaks memory so violently that Windows kills the system before it can do serious overflow damage.

• 2:13am: The system boots up after the DPC_WATCHDOG_VIOLATION blue screen and I load up Display Driver Uninstaller (DDU), ready to, at the very least, clear out what I'm hoping is just a bad Nvidia driver. DDU says I should run it in Safe Mode, and I idly wonder if maybe I can just turn my antivirus off. I disable Malwarebytes, and for some reason pop in to the Device Manager to have one last look to see if the Wifi Card is still broken. Instead of being broken, it’s merely disabled. That’s… new? I re-enable it and suddenly Wifi just… works. I have full internet again. It’s like nothing was wrong. What? I didn’t actually fix anything. Why is it working now? Maybe it was me turning off Malwarebytes? Was it conflicting with Windows Defender? (I’d had Malwarebytes installed for a while but all of its premium real-time protections were turned off – until the night before, where it updated to a free trial of the paid version and turned all the real-time protections on). Just to make sure no conflicts happen in the future, I uninstall Malwarebytes.

• 2:41am: Sure enough, I tell the Nvidia Tray Icon to update my 1060′s drivers and an entry appears in the Event Viewer saying the GeForce Experience is either missing or corrupt. The tray icon downloads and installs a fresh copy of the GeForce Experience to replace the corrupted stuff.

• 2:54am: Suddenly, the wifi disappears again. Same problem: there aren’t enough IRQ slots. When the GeForce Experience reinstalled and updated the drivers, the Nvidia Tray Icon went away and never came back. Maybe something’s still corrupted in there. I boot in to safe mode and use DDU to clean out the all traces of the Nvidia driver.

• 3:02am: When I restart with no graphics drivers, wifi is instantly working again. This looks promising. Fresh install of the drivers and everything’s looking like it’s back to normal.

• 3:40am: Wifi goes out yet again, because once again, it’s run out of IRQ slots. What do I gotta do to make sure this stays fixed? Well, since restarting last time fixed it, maybe restarting again this time will fix it.

• 3:45am: The system hangs on the “Restarting…” screen. Not sure what to do, I leave it there for a few minutes until eventually it cuts to a bluescreen. Our pal DPC_WATCHDOG_VIOLATION is back.

• 4:02am: The system seems to have steady wifi for about 30-45 minutes before it runs in to that IRQ error and dies. Now, the IRQ error precedes a guaranteed DPC_WATCHDOG_VIOLATION blue screen, usually by only 60 seconds or less. I’ve also noticed that when the wifi runs out of IRQ slots, the ASMedia eXtensible USB 3.0 Host Controller in the device manager also fails along with it (no error about system resources, it just stops working, this time with “Code 24.”).

• 4:45am: Skimming through hidden devices in the Device Manager reveals 2-3 different wifi devices listed as "Not connected." Is this from when I connected the TP-Link device a few hours ago? I uninstall all of them. This causes something weird to happen. Wifi will dip out for just a second, but then come back and the system will be “fine.” It’ll still eventually BSOD with the Watchdog Violation, but it’ll keep the internet up until that moment.

• 5:10am: The event viewer starts spitting out weird warnings about “Reset to device, \Device\RaidPort0.” followed by messages about retrying “IO operations” on the Disk at “logical block addresses.” I have no raid set up, it's just two SATA drives (one is 500gb, the other 2tb) Checkdisk seems to not care, says all drives are healthy. Earlier in the night a friend mentioned checking out software called “WhoCrashed?” that analyzes BSOD crash dumps and can help tell you what’s going on. I have to clear some HDD space for it, but eventually it’s just a matter of time of waiting for the next BSOD. I also install a system resource monitor called WhySoSlow from the same place just to see what’s going on under the hood. At this point, it’s been over an hour, and I start to think maybe I won’t have another BSOD.

• 5:20am: Idly, I run a system integrity check ("sfc /scannow"). It also says everything is fine.

• 5:50am: Within five or ten minutes of the integrity check finishing, “Application Responsiveness” and “Kernel Responsiveness” in WhySoSlow spike HARD out of nowhere. They go from 0ms to peaks of 200ms or more. It’s like when I first plugged in the TP-Link device a few hours earlier. The system is incredibly sluggish, but the Event Viewer isn’t reporting anything out of the ordinary, the USB 3.0 controller’s still working, and so is the wifi. Even the system temperatures are normal (40C and below), so this isn’t a loose CPU cooler again. Regardless, seconds later, boom: we have our Watchdog Violation BSOD to analyze.

• 6:08am: Windows is LETHARGIC to start up. It takes forever just to get to the desktop, and even longer to show icons. I manage to get it to load WhoCrashed and it analyzes five dumps made by Windows. Unfortunately, I can only read one and half of another. The system won’t shake this sluggishness and I have a bad feeling. I snap photos of the two entries so I can look at them in detail later, with the other three impossible to read.

• 6:10am: Windows eventually BSODs again with another Watchdog Violation. WhoCrashed said the Watchdog “detected a prolonged runtime at an IRQL of DISPATCH_LEVEL or above” and that this was “typical of a software driver bug” and not a hardware issue. Of the second memory dump it read, all I could make out was that the error happened in asstor64.sys – aka the ASMedia eXtensible USB 3.0 Host Controller. I try to get Windows 10 to boot in to safe mode so I can have a look at the rest of the dump analysis, but Microsoft removed the ability to boot you system in to safe mode by holding F8. Now you have to actually get in to Windows and pick “Safe Mode” from a menu option. Hard to do that when Windows starts BSODing before you can get to the Safe Mode menu.

• 6:16am: Windows is still starting up INCREDIBLY slowly. I can’t even get the start menu to come up. And, before I know it, once again, boom: Watchdog Violation. I think they’re getting closer together.

• 6:20am: As Windows 10 once again lurches back to life, I try and get it to shut down, but the start menu still won’t come up. Instead I hit the power button to force a shut down, but it gets stuck on the “Shutting down…” screen for several minutes before also getting a Watchdog Violation Bluescreen. Instead, while it’s on the BIOS screen, I just power the system off entirely, frustrated.

So, MajorGeeks, what does this sound like to you? I have my own theories, and I've talked to a few tech-savvy friends who have different theories, but I don't have a lot of money to throw around and I don't want to make a move just yet.

My theory: It's possible it's the TP-Link adapter causing havoc. I shouldn't have plugged it in without the drivers CD, but I forgot I uninstalled it so thoroughly (I removed it from the Device Manager and uninstalled the software it came with). The failing USB 3 Controller kinda-sorta relates to the Microsoft Community article I linked above about what the TP-Link adapter can do to a system. But you'd think these drivers wouldn't be activating unless the USB adapter was actually connected, which it wasn't. Nor does this explain what happened to my GTX 1060's drivers or the hang on the loading screen. But it's possible those were unrelated and by plugging in the TP-Link adapter I just added a new, different, much worse problem to all of this.

One Group of Friends: just want me to reinstall all of my motherboard's drivers through Safe Mode. I know there are ways to get in to Safe Mode without using F8 (either a USB stick with a Windows Repair thing on it, which I don't know where to get, or just simply interrupting the Windows logo three times by pushing the reset button). But I only have one USB flash drive, which means I probably can't put the mobo drivers AND the Windows Repair tool on it at the same time. Some of them say to just reinstall Windows 10 completely. I'd definitely want to avoid reinstalling Windows, as it takes me a while to get all my stuff set back up. Plus, I have files that are hard to replace, and if Windows is going to wipe my whole C:\ drive for the install, I stupidly kept all the installers for that stuff in my User/Downloads folder. Plus...

One Friend: seems to think maybe it's a motherboard issue. Between the IRQ problem, the graphics card problem, and the USB controller problem, he thinks it's a hardware issue. I mentioned plugging the system directly in to the wall outlet because we did have a power outage the same day all of these problems cropped up. The system was off, but I know (now) that doesn't matter when it comes to a surge. But wouldn't the power supply take the brunt of that and mostly spare the mobo? EVGA's ten year warranty means the power supply can probably be replaced, but Newegg doesn't sell my mobo anymore. There's a nearly identical one they sell (just with Fatal1ty branding and fewer SATA and USB 3.0 ports) but that's $100 I'd rather not spend.

So what am I looking at, Geeks?

 

Welcome to Major Geeks

You can save any important data by first hooking the bad system hard drive up to another working system, then save that data.
Now use the windows Media Creation Tool on another working system and save a copy of Windows 10 (the exact version and edition you have on the broken system) to a USB Pen Drive with at least 6gb of free space. If you use a DVD you will then need to burn it to make it bootable. USB is already bootable so much easier.
Now reinstall Windows afresh. As it had Win10 on it originally. Microsoft servers will recognise the COA Product Key.
http://www.thewindowsclub.com/windows-10-media-creation-tool-create-installation-media-upgrade

 

I would check your screws on your mother board if they are not the correct tensing.The screws could throw sparks and short out the board every now and than.

See if that is the case.

I know my desktop has a problem every once in a while where some how the screws on the board be come lose.And it causes the mother board to short out.

They have to be slightly lose to tighten where it's grounded.

Edit: and make sure you Re check all your wires and cables to make sure they are not in the way of any fans or air flow.

 

What exactly do you mean by "the correct tensing"? I guess just check to see if they're tight? That's not a bad idea, I guess, but I've always been fuzzy on how tight is too tight or how loose is too loose. (I think I bring that up in this video I did where I built the PC originally, about 18 months ago)

 

Using terminology that 99% of people wouldn't have a clue about is not helpful. The tightness of screws should be just a little bit more than snug. That's all that 'tensing' means. If the system still doesn't work without the prior issues after checking the screws are a snug tightness, then you are left with a reinstall.

 

I want to ask. SSD?
s.

 

No, no SSD. SSD was on the list of things I wanted to look in to in the future, but for now it's just been regular old platter drives. One's 500gb (Windows 10) and about 18 months old, the other's 2tb (Steam games) and is about 2-3 years old.

 

My advice is to invest in a cheap one.
125GB should be fine...

Good luck,
s.

 

I wouldn't look at is an upgrade. Moore like working better. For £50.
s.

 

I use this KINGSTON SUV400S37240G (SSD).
https://www.amazon.com/Kingston-Digital-SSDNow-SV300S37A-240G/dp/B00A1ZTZNM

Pretty cheap for a speedy drive.

 

I would simply try to get the system to work first, with the existing drives - that's the priority. Upgrading to an SSD later is a secondary issue.

 

Please forgive my intervention. Bringing SSD's into this thread.

I am currently getting my knowledge around IDE to SD via sata connection.
The op has some great equipment.
I thought the motherboard might love SSD.

I really was trying to offer help.
Best wishes.
s.

 

You said "I mentioned plugging the system directly in to the wall outlet because we did have a power outage the same day all of these problems cropped up. The system was off, but I know (now) that doesn't matter when it comes to a surge."
I would save as much data as I could , perhaps some in the "cloud", and run tests on the motherboard , and psu before buying/replacing, or, re-installing anything.
The loose 4th screw on the heatsink bothers me.........too.

 

I'd be worried about the loose screw on the heatsink too, but the temps were all pretty normal even as the repeated BSODs were getting to be their worst. I've broken those Intel coolers before trying to get the pins to snap in, I'd rather not have that happen again by pushing too hard. It's one of those things where the pin seemed to lock down but if I tugged hard enough it would come back up.

I haven't done anything with the desktop since posting in this thread because I was trying to gather as much information from as many sources as possible. The largest consensus seems to be to just try reinstalling Windows, and if that doesn't work, go for hardware stuff.

The main issue now is how I get files off the desktop before I do so; I figured I was a real smarty-pants by keeping and organizing all of my software installers when I built the system 18 months ago, but recently I realized that folder is on C:\ and will get wiped when I reinstall. There's also things like Firefox bookmarks and such that haven't had fresh backups in a while that I need to rescue. But I only have one USB flash drive. I'll have to figure out how to get in to the system (interrupting Windows' boot enough times to trigger automatic safe mode, maybe), get the files on to the flash drive, and then maybe transfer them to my laptop so I can throw them on Google Drive or something.

This would be so much easier if I wasn't in the middle of moving. I have copies of The Ultimate Bootdisk on DVD, but they're locked away in a box in a storage unit right now.

I'm curious, though, how exactly would I run tests on the mobo and PSU?

 

If you got any Back up 1 TB drives that i would recommend on saving any thing that is important to you.

Like game files if you save all the game files like a copy and past it will save hours of re downloading every thing and updating it.

This is what i do when i fresh install game files.If they are large files this saves a lot of time and data.All you would need to is recopy the files to the current fresh install of games.Works wonders.

This even works with online mmos just copy and past in a different hard drive.

 

The MB comes with a support cd with utilities on it. See what the utilities are for.

 

I used to use motherboard monitor, but it seems outdated now- the only thing I can find with motherboard testing is this, but on a new board.
http://www.computerhope.com/issues/ch001089.htm

The psu test- something like this
http://www.frozencpu.com/products/4081/psu-208/Coolmax_PS-124_Power_Supply_Tester.html?tl=g11c28s88
You could check the +/- 3.3v, 5v, and 12v lines

 

Hello again,
I would love to save your OS disk but my solution still holds some weight. A different Bootable disk.
You could then link the two drives on the motherboard and files recovered.
It would be an extra drive.

Of course I would like to try Admin Command prompt sfc /scannow

Good luck
s.

 

I did sfc /scannow, it found no problems (mentioned it in the big long OP).

I dug through storage today, where most of our stuff is (I mentioned we're in the middle of moving to a new place; we move in to our new apartment on November 18th) and managed to dig out a copy of The Ultimate Boot Disk on DVD. What should hopefully work is using the "Parted Magic" Linux distro on The Ultimate Boot Disk to get at my files on the HDD and transfer the important ones to my 32gb USB flash drive. From there I can toss them up on Google Drive or whatever.

I'll probably try that soon, but tomorrow I'll be out of the house all day taking care of getting a new State ID among other things.

Who knows, if Parted Magic exhibits problems maybe this will confirm it's a hardware issue. I also have a pair of USB HDD enclosures in storage (one for IDE and one for SATA) so if it's the motherboard there's a good chance I can pull the HDDs out of the system and just access the data over USB.

Then I guess I'll try reinstalling Windows, given that seems to be the most common path forward from everyone I've spoken to.

 

Looks like I can't edit my posts after they've been up for long enough, but I actually didn't see baklogic's post until now. I'll probably give that a shot before I reinstall, but I gotta trying to rescue those files, first.