Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Author Topic: My PC's death, and ressurection (Adventures in Troubleshooting)  (Read 1972 times)

nenjin

  • Bay Watcher
  • Inscrubtable Exhortations of the Soul
    • View Profile
My PC's death, and ressurection (Adventures in Troubleshooting)
« on: December 10, 2017, 04:44:24 am »

Posting this because, maybe, somewhere down the line this will be useful to some forumite.

Synopsis 1: My PC turned off one day and wouldn't turn back on. Cue around 8 to 10 hours of troubleshooting to figure out why, and resolve the problem. Did I get off scott-free? No. But I managed to isolate the problem and get back to internet capable functionality.

Maybe I'm just forum-deprived for the last two days and need to type some shit, or maybe I'm just elated that I feel like I've navigated this web of fuckery and want to talk about it.

This post will cover several main points related to sudden PC calamity and what you can do/might do and what to look for. In my case it will cover:

1. WTF just happened.
2. How did I work through the issues?
3. How did I make it harder/more stressful on my myself?
4. What are the long term consequences?

Chapter I: Trouble In Paradise.

Be me, playing 7 Days To Die with friends, late on a school night. Blood Moon is coming. Gun be gud.

And then *poof*, darkness. My whole PC shuts down with barely a whisper.

I've been building my own PCs for near 20 years now. I work Tech Support as a job. My brain goes to work.

Assumption #1: Power surge. Even though none of the other lights in my room did anything, it was my gut reaction.

Solution #1: Wait a moment, power on PC.

Result #1: PC doesn't turn back on. Something in my case clicks when I hit the button, but nothing starts spinning.

Assumption #2: Short/Power supply has died. Very feasible explanation for why I get zero reaction to powering on.

Observation #1: Strangely though, an LED power switch directly on my motherboard is still lit up after a minute or two of looking at the situation and thinking. Couldn't be residual juice left in the board or battery because it drains off pretty quickly.....

Solution #2: Power off Power Supply, unplug it from the wall, wait 30 seconds. Let whatever electrical/magical wizardry dissipate. Reconnect, flip power switch back on, try to power on.

Result #2: Nada. Zero. Zilch.

Conclusion #1: Power supply could be having real problems.

Double-checking #1: Flip everything off again. Check Surge Protector. Yep, it's on, so the socket is good as well. Swap plugs on the surge protector. Yep, still not turning on.

Conclusion #2: PSU definitely fucked.

Consideration #1: Suddenly loss of voltage under load can have severe consequences for any component of your PC. Sure my power supply might be dead but who knows what else might be toast....

But it's way too late to start digging into the real guts of troubleshooting. So I go to bed and dream of a terrible world without internet and video games.

Chapter II: Panic, Mistakes, Discoveries and Troubleshooting 101.

The next day arrives and I do several things: I ask a co-worker if I can borrow his Antec 750w Power Supply to test in my rig. And I start checking PSU prices on Newegg. There are some nice deals on some good PSU's and they're going to be running out soon....

I order a PSU and take my loaner PSU home.

Troubleshooting Step #1: I disconnect my PC from wall and peripherals and take it into the living room to start disconnecting the PSU from everything.

Discovery #1: As I'm looking in and getting ready to disconnect stuff I notice....the supplementary connector that is part of the 24 pin ATX connector is dangling loose while the rest of the connector is still plugged......ruh-roh. That should not be like that!

Realization #1: It might not be the PSU. And the damage could be way, way more severe than I thought. The mobo losing power before the other components could fuck up all sorts of things.

Troubleshooting Step #1A. Still, I can't confirm it didn't uh...fall out while I was moving it. So I proceed on with replacing my current PSU. I get it all hooked up, hold my breath and......

Result #3: Fuck all, beyond *click* again. Well, now I'm in deep water. Could the motherboard be shot? Luckily I still have a phone and Google so I ask "computer won't turn on, what do?"

Lesson #1: Motherboards have this thing called the CMOS. It's like the core settings for the motherboard (*not an expert*). And the thing that helps it remember these settings is a battery about the size of a watch battery that sits on the motherboard itself. If you switch a certain jumper on the motherboard (to break the electrical circuit and "purge" the connection between battery and CMOS), you can pop that battery out and reset the CMOS. After a severe, unexpected shutdown due to something like, say, a random chunk of power suddenly disappearing, weird shit can happen to how computers store information. It gets corrupted, unreadable. It can't be processed and in the big chain of things that need to work for your PC to turn on, it runs in to a show stopper.

Troubleshooting Step #2: Still using the replacement power supply, I move the jumper over on the motherboard, move it back after 10 to 15 seconds, pop the CMOS battery out, wait 30 seconds, put it back in, hook everything back up, flip the power on and.....

Result #4: POOOOOOWWWWWWAAAAAHHHHHH! Things are spinning. Fans are turning in the case and on the CPU and that's good! But uh, wait a second.....

I've got power but nothing is happening. The screen doesn't show anything.

Panic #1: PPPPPPPPPPPPFFFFFFFFFFFFFFFFTTTTTTTTTTTTTTTT.

Chapter III: The Plot Sickens

Troubleshooting Step #3: Seek help. Not from the internet this time.

Lesson #2: There are a million paper manuals you get for the shit you own these days, and I keep almost none of them. But one manual I always hold on to are my motherboard manuals, because if you know what to read for there's a lot of diagnostic information in it. And if you buy a fancy enough motherboard with good diagnostic features.....My motherboard (Asus z97 Pro) has several LED lights to indicate problem areas of the board, and another LED to tell you (by way of codes you look up in the manual) what the motherboard is currently "doing." There's lots of small subroutines and stuff your motherboard has to do before Windows even starts up. 99% of that is invisible to you and happens in milliseconds. But when just one of those steps is fucking up....you got a seemingly dead PC on your hands.

Troubleshooting Step #4: Look up the codes.

Chapter IV: Chaos.

My recollection here gets a little hazy because, well, to be honest, I did a lot of stuff without a ton of rigor.

Essentially, between the codes being shown on the motherboard display, and the LEDs that point to specific areas and components of the motherboard as having troubles, it was telling me that it couldn't read my "IDE devices" (basically my hard drives and/or CD/DVD drive) or that there was a problem with my VGA slot (i.e my video card.)

Troubleshooting Step #???: These were dark times. I spent a lot of it alternatively unplugging hard drives trying to resolve why the motherboard couldn't see them.

Result ???: Every time I'd get the IDE detection problem to go away, the motherboard would complain about the VGA slot. It was frustratingly inconsistent.

Hindsight #1: Getting confused about what drive is plugged in to what SATA port, and what the drive label names are, on top of all the other inconsistent weirdness going on, added a lot of time and confusion to the troubleshooting process. I'll expound on this more in a bit.

Troubleshooting Step #5: Since things are so fucking weird and inconsistent between reboots and device swapping, I decide to reset the CMOS battery again (after switching the jumpers) and let it drain out for a good 30 mins while I collected myself. I also removed the video card because I couldn't seem to resolve any problems going on around it, and plan to just hook my monitor to the default display port on the motherboard itself.

Consideration #2: Electricity is a strange beast. It doesn't always operate cleanly the way we'd like it to with our devices. Charges get held on to by pieces of hardware, and this charge represents what it thinks it's doing. Maybe that's what's going on with all this inconsistent weirdness, or maybe my motherboard has been damaged by all this shit.....

Troubleshooting Step #6: I replace the CMOS battery, pray to the Dark Gods and....

Result #5: WE HAVE POST. Immediately after POST it goes to the BIOS. After everything that's been going on that's not unexpected I'd guess but it's indicative of other weird shit going on, that it didn't just immediately boot to the primary drive. The system clock has been reset (because the CMOS battery was reset) and that's expected.

Hindsight #3: Your device boot priority also gets reset when you pop the CMOS battery. If your drives were arranged to boot in a specific order, well, that order is now lost.

Observation #2: In the BIOS it does not recognize any of my drives. Wat. That....that's not right. Nothing on them has changed.

Troubleshooting Step #4: Set the right clock time, reboot, see what happens.

Result #6: System reboots, goes back to the BIOS immediately, and the system time is not what I set it to. Ruh-roh raggy. Devices still aren't visible in the BIOS boot priority list.

Troubleshooting Step #5: Breathe deeply, and calmly reset the system one more time.

Result #7: System goes immediately to the BIOS on boot. The system time I set previously is now being displayed. I have two devices showing in the boot list: my CD/DVD drive and one hard drive. (I have three total in the system.)

Observation #3: As I ponder all this, I happen to notice that, between various restarts and what not, that my GPU fans aren't turning. Not especially abnormal when the system isn't even really doing anything, but it's weird that they kick a little bit every 45 seconds or so.

Troubleshooting Step #6: Since I've got at least one drive visible, I should try to boot to that.

Result #8: The system POSTs and....I see a black screen and text saying BOOTMGR NOT FOUND.

Chapter V: The Tech Priest goes to work
If we were in to it before, we're knuckle deep now.

Assumption #3: BOOTMGR is basically a core file on the harddrive that needs to be there before Windows will start. If it's not there, my hard drive is probably corrupted or unreadable.

Hindsight #3: Remember what I was saying about 3 drives and not remembering what they were and how they're physically plugged in? Two of them (my main and my backup) both have Windows installs. The third is like 12 years old and is only there because is still runs, pretty much.

Stupidity #1: Guess which drive I was booting to, in my failure to read which one the BIOS was saying it saw. The shitty 12 year old drive with an utterly broken set of data on it. No wonder it didn't boot. But I don't realize this at the time, so I....

Troubleshooting Step #7: Try to repair the drive via the Windows disk. Worth a shot right?

Observation #4: Most troubling, during some of my reboots the BIOS screen locks up....uuuuuuhhhhhhh............

Result #9: Windows can't fix my shit ass old drive (because there's nothing to fix.) Moving on I....

Troubleshooting Step #8: Work my way up the chain. I unplug all the drives and plug them in one at a time as the primary and only drive next to my CD/DVD drive, to see if I can get one to boot.

Result #10. After trying to boot my shittiest, oldest drive, I boot my second oldest drive, which has a Windows install on it. It boots! I see the usual "Windows didn't shut down right" message and am like, yeah, ok, let's try to fix windows so I can get ANYTHING going. I'm concerned my primary drive might be fucked too and what to see if I can read, and maybe copy, data off it.

Troubleshooting Step #9: Let Windows do its thing.

Result #11: After a few minutes of Windows trying to repair the installation on disk, it reports back that it can't. What a shock.

Troubleshooting Step #10: I hook up the last drive in the stack as the primary drive.

Realization #2: I am a dumbass.

Result #12: I'm greeted with another "Windows did not shut down properly" message. But this time it's for the right Windows install. I tell Windows, you know what, go ahead and boot and voila....I'm looking at a glorious 800x600 rendition of my desktop.

Synopsis 2:
PSU: Probably not dead.
Mobo: Works....sorta? Some quirky shit going on.
GPU: The jury is out still.
CPU: Good, otherwise none of this shit would have worked.
Memory: Memory problems are often a show stopper. To be on the safe side I unplugged and reseated all the memory anyways. Even blew on them too.
Hard drives: My two older drives are intact but need reformatting. But my primary drive seems ok!

Smart Things #1: First off, I power it all down. I replace my loaner PSU with my own (just to make sure it really is OK) and hook up my older drives again as secondary drives. I do not yet hook up the GPU.

Result #13: System powers on and boots normally. (Sorry Rosewill, I maligned you for nothing) And in Windows I can see my other two drives. Things are starting to look up!

Smart Things #2: I format my other two drives and do fresh backups of documents, music and pictures. At least I can breathe a little easier knowing that's done.

Troubleshooting Step #11: All that's left now is to hook the GPU back up. Hell I might even get some gaming done tonight!

Result #14: The system boots, I see the POST messages....then it goes to black screen as soon as Windows tries to boot. FML.

Chapter VI: Or Is It Merely A Trick Of The Light?

Observation #5: The GPU fans are still not spinning correctly. Which is not encouraging. The system is clearly booting, Windows is clearly loading, why do I have no display...hrm.....

Troubleshooting Step #12: I have two monitors plugged in to the GPU. (When I was plugged directly in to the board I had just one monitor.)

Supposition #1: Perhaps Windows, after all this, and the program I run to use dual monitors non-natively in Windows, are very confused by having two inputs.

Troubleshooting Step #12a: Simplify the problem. Go down to one monitor.

Result #15: After POST, I see the Windows logo loading!...and then it goes immediately in to Windows repair. Oops. I guess all those reboots midstream when I don't actually know what the system was doing is really pissing Windows off. I decide it's best to just let Windows do Windows and so I wait.....

Observation #6: After about 10 minutes of waiting for the Windows repair to run, my display suddenly shuts off.

Assumption #4: My graphics card crashed.

I'm now in a state where Windows is doing repairs, after too many sudden restarts causing problems, and I can't see anything it's doing. Like when it's done or when it says it can't fix the problem. Which means I'm going to have to do another blind restart.

Panic #2:AAAAAAAAAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH!!!!!!!!!!

Troubleshooting Step #13: Do not make the problem worse. Luckily almost all computer cases have a little light to let you know there's hard drive activity. Resetting your PC while there's hard drive activity enough that light is constantly on is generally a bad thing. So I use it as my guide. I wait about 20 minutes until I stop hearing the hard drives run and that light is no longer constantly flashing. Still staring at a black screen wondering what's going to happen next, I hold the power button and my breath and.....

Result #16: It reboots and goes straight into a normal Windows start up. PHEW. Ok.

Smart Things #3: After THAT little scare I decide to back up some more shit, because at this point I'm taking nothing for granted.

Observation #7: I'm about to start copying files over to my back up drives when my screen just goes white, and I can't see my mouse cursor.

Assumption #5: My graphics card died again. It's the only real explanation for why my display is turning on and off seemingly at random and not coming back on until a restart.

Troubleshooting Step #14: All I can really do is a) reboot the system or b) pull the graphics card. I decide to give it one more shot. Again I'm told Windows failed to shut down properly so, I decided to boot it in safe mode, so it doesn't have to load the graphics card drivers, so I can just get some more files copied before something else goes wrong.

Result #17: Windows begins loading core files for booting up safemode. It hits CLASS.PNP I believe when......the screen goes solid lime green and freezes.

Conclusion #3: My graphics card, if not completely fucked, is more or less fucked because it's unstable and is faulting while not even under load. The color is kind of a dead give away. If the graphics card is crashing just trying to be initialized sometimes, there's no way it's stable anymore.

Troubleshooting Step #15: Unplug and remove graphics card, plug single monitor directly into the board's onboard display adapter.

Result #17: So far I've been running for about 2 hours with no restarts, hangs or colored screens of any kind. I've got internet, my data is backed up, my primary drive has been error checked and come back clean, and things are looking ok.

Epilogue: What Price Victory?

So here's where I stand:

-PSU: OK. So easy to assume it's a PSU problem sometimes.
-CPU: OK. Never gave me a problem.
-HDD: OK. After all I put them through....
-MEMORY: OK. Like the processor, there was never anything to indicate a problem here, either in start up issues or LED codes and such.
-MOBO: ....It's hard to say. The BIOS locking up could be due to a lot of things. The failure to save BIOS changes on occasion could be due to a lot of things. If my display adapter on the board started giving me trouble, I'd say the board was damaged at the very least. It's also possible that my video card isn't the thing that's damaged, it's the PCI-E port that's messed up. But right now it's doing it's job so I'm inclined to say it's ok.
-GPU: The real casualty of this event (other than my weekend and peace of mind) It works for a little bit but it's damaged and so I'm going to have to get a new one. And now is a particularly pricey time to buy a video card. This card is a Geforce 970 and I hear people rave about their 1070s....but I'm not super enthusiastic about spending that kind of money when I was happy with my 970. As I said above, it's possible the video card isn't damaged and it's the PCI-E slot it's going in that is unstable.... and that's something I could test using one of my other PCI-E slots. It wouldn't be the first time I've heard of the expansion slots themselves getting damaged. I might as well try it before I order a new video card entirely.

So that's muh story. Let it be a cautionary tale about thinking you know what's going without testing, re-testing and verifying. Computers are legos with voltage, and voltage can make some weird shit happen to hardware. And much like software, there's a stack of processes that need to happen for things to work correctly, and the problem you're seeing might actually just be a symptom of something else.

I've already cancelled my order from Newegg for a new PSU and will be placing one for a new GPU soon. Hopefully when I install it everything goes smoothly and when the system goes under load I don't find out that, like, my memory is shot or some capacitor on the board is actually fried.

THE FINAL LESSON: So basically the random chaos of a connector becoming unplugged at the worst possible moment cost me a weekend and probably a couple hundred bucks. But maybe it wasn't so random. See I just moved a couple months ago to this new place. And while my PC booted up fine and everything when I moved in....I never did stop and double check all the connections. Hell I've gone to LAN parties over the years and never bothered to check the internals of my systems after I moved the computer around. I still don't have any evidence the loose power connector to the mobo actually popped out when my system died, I only saw it out after I'd been doing things. But it seems the likely culprit.

So the next time you move your PC, double check your critical connections like your CPU power, your CPU heat sink, your HDD power and SATA connections and of course your MOBO power. Any one of these things coming loose while your system is under load can set off a chain reaction of tech-fuckery that, obviously, if you don't know what you're doing can result in "welp, there goes my computer" or "better go pay a computer tech to figure it out." And if you're a builder yourself, just remember that when shit goes wrong, simplify the problem you're looking at. Take it one piece at a time, isolate the problem, fix it and go on to the next problem.
« Last Edit: December 10, 2017, 02:37:03 pm by nenjin »
Logged
Cautivo del Milagro seamos, Penitente.
Quote from: Viktor Frankl
When we are no longer able to change a situation, we are challenged to change ourselves.
Quote from: Sindain
Its kinda silly to complain that a friendly NPC isn't a well designed boss fight.
Quote from: Eric Blank
How will I cheese now assholes?
Quote from: MrRoboto75
Always spaghetti, never forghetti

wierd

  • Bay Watcher
  • I like to eat small children.
    • View Profile
Re: My PC's death, and ressurection (Adventures in Troubleshooting)
« Reply #1 on: December 10, 2017, 08:47:31 am »

Personally, I would have done these steps:

System not booting? Suspect psu--
Unplug PSU cable, flip interrupt switch on back of PSU, wait 5 seconds. Replug, flip switch back, try power.
No juice to motherboard-> suspect PSU
Juice to motherboard-> suspect hardware fault for lack of post.

Juice to motherboard[true]:
Unplug everything (except power and power on button) from the motherboard, and remove all cards. Ensure that a speaker is connected to the speaker pins. Attempt power on.
Post = YES -> Troubleshoot peripherals
Post = NO ->Listen for beep codes
Beep Codes = YES -> Identify Beep codes, take action accordingly
Beep Codes = NO -> Suspect Motherboard failure, Begin motherboard diagnostic process

Peripheral troubleshooting:
[obsolete section, here just for nostalgia]
Floppy drive light stays on, does not pass POST test-> Ribbon cable on upside down.
Both floppy drive lights come on at the same time, fails POST -> One of the drives is incorrectly cabled on the wrong side of the cable twist.
System hangs when detecting IDE devices -> Improper jumper settings, ribbon cable upside down, or defunct IDE device/IDE device not powered
[/obsolete]

System does not produce POST screen with GPU connected -> Check GPU for function in known good motherboard to rule out component failure
GPU in test rig == GOOD -> check motherboard slot for damage, video cable for damage, check monitor function.
GPU in test rig == BAD -> Inspect card edge, if not very shiny then clean with white vinyl eraser. Recheck function.
GPU in test rig powers on, shows colorful snow -> Video memory on GPU is bad / GPU is toast. Replace.
GPU in test rig works properly, causes shutdowns on production system --> Check GPU heatsink for ancient thermal pad; replace with proper compound, check GPU getting proper power (for cards with power cable connectors), check GPU fan turns (clean/replace if faulty)


System hangs with peripheral card installed:
[Obsolete, for nostalgia only]
Card is ISA/EISA --> verify jumper settings of card, verify card fully inserted, verify card edges clean, verify slot not damaged
Card is MCA --> verify that device descriptor data is present on MCA architecture BIOS, load appropriate descriptor floppy if missing, verify slot is activated, verify card configuration/jumpering
Card is VLB --> verify card FULLY SEATED, Try card in different slot, verify card jumpering
[/obsolete]
Card is PCI --> Verify card PCI-IRQ not conflicting with other devices (onboard devices, such as USB controllers, sound hardware, and video hardware, still are attached via a 'phantom' PCI slot. Each slot has a hardware line that is the "PCI-IRQ" of that slot. It is often physically wired to an actual physical slot, which shares that IRQ. A device inserted in that slot that cannot share its IRQ will cause serious problems!) Verify card not older PCI spec/Too slow for system bus speed, verify card edge clean, verify slot not damaged
Card is AGP --> Verify card edge is clean, verify slot not damaged, verify card fully/properly inserted
Card is PCIe --> Verify card edge is clean, verify slot not damaged, verify card fully/properly inserted

System experiences thermal shutdowns:
Check for ancient thermal pad on CPU; replace with proper compound
Check heatsink appropriate for CPU/clockspeed; replace if needed
Check CPU fan turns freely/spins when system on
Check case fan operation
Check for dust bunny infestation

System reports non system disk, missing bootmgr, missing ntldr, etc:
drive = SATA --> Check drive port connection, check drive is system disk, check for incorrect translation mode (legacy modes), check for damaged partition table, check proper partition marked active
Drive = IDE --> check drive is on proper channel, check drive is system disk, check incorrect geometry detected/translation mode, check damaged partition table, check proper partition marked active
Drive = SCSI --> Check drive LUN ID, check adapter ID, check drive is system disk, check system boot order is SCSI first, check incorrect geometry detected, check damaged partition table, check proper partition active
[Obsolete]
Drive = MFM/RLL/ESDI --> check drive properly paired with controller (may need low level format if drive attached to foreign controller and not properly set up), check geometry in BIOS correct, check is system disk, check damaged partition table, check proper partition marked active
[/obsolete]

etc...


In your case, the GPU is probably overheating, causing the system to hang/crash.  I would investigate it for a dried out thermal pad, and would make sure the fan spins when attached to a known good voltage source (disconnect the fan cable from the GPU, then run ~5v through it, paying attention to polarity.) If it spins freely/easily, then the fan is good. If it does not spin, or only spins after you poke it a little, then it is either burned out or all gummed up with crap (and not long for the world anyway.) If it is all gummed up, you can TRY to clean the fan using pure rubbing alcohol (90+%!! None of that 50 or 75% shit) in a bowl, submerging the fan, and working it manually back and forth while submerged, but the damage to the bearings is probably already done. You might be able to revive it for another year or two of use though. If the fan is just straight up dead, (or you dont feel like trying to revive a gummed up one) you can try replacement with a same-voltage, same CFM(and dia) fan that will mount to the heatsink. Otherwise, you are just stuck replacing the GPU.

Logged

nenjin

  • Bay Watcher
  • Inscrubtable Exhortations of the Soul
    • View Profile
Re: My PC's death, and ressurection (Adventures in Troubleshooting)
« Reply #2 on: December 10, 2017, 01:16:35 pm »

The GPU fans were turning full tilt for a while during my testing in my rig. So the bearings are good I believe. No tension or drag when spinning them by hand. I suspect there's something else faulting in the card, although overheating is a distinct possibility. (When I pulled the card from the board the last time, the back was noticably warm even though I didn't put it under load.)

I'll probably try a different PCI slot today, watch what the fans are doing. I do really hope I'm not in for another video card. The GeForce 1070s are going for a cool $500 and that is like 2x what I paid for my current card just a couple years ago, which is a little ridiculous.
« Last Edit: December 10, 2017, 02:01:22 pm by nenjin »
Logged
Cautivo del Milagro seamos, Penitente.
Quote from: Viktor Frankl
When we are no longer able to change a situation, we are challenged to change ourselves.
Quote from: Sindain
Its kinda silly to complain that a friendly NPC isn't a well designed boss fight.
Quote from: Eric Blank
How will I cheese now assholes?
Quote from: MrRoboto75
Always spaghetti, never forghetti

nenjin

  • Bay Watcher
  • Inscrubtable Exhortations of the Soul
    • View Profile
Re: My PC's death, and ressurection (Adventures in Troubleshooting)
« Reply #3 on: December 10, 2017, 09:30:30 pm »

Posting from my phone, so as you can imagine the situation has worsened.

I attempted to test the video card on last time on a different PCI-E port. The display didn’t start on the first boot (fans were spinning good though) but did on the second and I got in to Windows. (Noticed on the 2nd run that the fans were not spinning.) I tried going to the display properties to see if I could change resolutions, the screen started to get snowy and then it went white and my hard drive started grinding away.

So the card is pretty clearly messed up.

But now I have my old problem back: I can’t get the system to power on again.

After hard powering down the system and removing the video card and switching input back to the onboard display adapter, the system wouldn’t power on. I reset the cmos again, making sure the jumper was on the right position before rebooting. Still no luck. I’ve pulled all the drives and all the memory sticks but one, and reseated that one and still no power. I’ve reseated all the major power connectors and still no luck. About the only things I haven’t tried are reseating the processor and all the case fan connections.

Kind of demoralized. I feel like I made the situation worse but I don’t know how.

The board is still receiving power from the PSU. I know this because an led on the board lights up when it gets power. So maybe there’s a short in one of the connections I haven’t touched that is preventing power on?

Tomorrow I guess I’m pulling the CPU and reseating, and basically rebuilding it from square one, because I’m running out of ideas to try. Could it be another weird problem I’m overlooking, or is the board truly fucked up now too?

If anyone has any suggestions I’m all ears.
« Last Edit: December 10, 2017, 10:18:54 pm by nenjin »
Logged
Cautivo del Milagro seamos, Penitente.
Quote from: Viktor Frankl
When we are no longer able to change a situation, we are challenged to change ourselves.
Quote from: Sindain
Its kinda silly to complain that a friendly NPC isn't a well designed boss fight.
Quote from: Eric Blank
How will I cheese now assholes?
Quote from: MrRoboto75
Always spaghetti, never forghetti

nenjin

  • Bay Watcher
  • Inscrubtable Exhortations of the Soul
    • View Profile
Re: My PC's death, and ressurection (Adventures in Troubleshooting)
« Reply #4 on: December 16, 2017, 05:26:00 pm »

Posting from my PC again.

I took a week off of even thinking about it, it was stressing me out that bad. Spent it watching TV instead.

Came back today, and within the first 5 minutes figured out what was wrong. Sometimes stepping away from a problem is the best thing for you.

Turns out when I removed my video card the last time, the 20-pin mobo combination connector for the speaker, power, reset and HDD LED had become disconnected from the board. Which isn't too encouraging given how many important things that shouldn't get disconnected keeping becoming disconnected from the board. But hey, it starts, and nothing else appears to be damaged.

So I'd say the absolute final lesson in troubleshooting is.....take a deep breath, take your time, and check. Fucking. Everything. The problem is almost always something small, easy and dumb.

The video card is still boned, but now at least I'm only looking at $200 to $500 replacement cost instead of $700+, and no huge pain in the ass retrofitting parts. So relieved. I was seriously starting to question my own competence, just like every time I have to fuck with PCs and something goes wrong, even though I've been doing it for two decades now.

So glad I took one more crack at it before giving up and spending money.
« Last Edit: December 18, 2017, 08:21:02 pm by nenjin »
Logged
Cautivo del Milagro seamos, Penitente.
Quote from: Viktor Frankl
When we are no longer able to change a situation, we are challenged to change ourselves.
Quote from: Sindain
Its kinda silly to complain that a friendly NPC isn't a well designed boss fight.
Quote from: Eric Blank
How will I cheese now assholes?
Quote from: MrRoboto75
Always spaghetti, never forghetti

nenjin

  • Bay Watcher
  • Inscrubtable Exhortations of the Soul
    • View Profile
Re: My PC's death, and ressurection (Adventures in Troubleshooting)
« Reply #5 on: December 30, 2017, 12:02:44 am »

My baby, she lives again. Got a MSI GTX 1070 8gig. Looks like a damn space ship and costs about as much. But it's a big upgrade over the 970, more horse power, less energy usage, has stupid LEDs on it, and looks badass in my case.

Logged
Cautivo del Milagro seamos, Penitente.
Quote from: Viktor Frankl
When we are no longer able to change a situation, we are challenged to change ourselves.
Quote from: Sindain
Its kinda silly to complain that a friendly NPC isn't a well designed boss fight.
Quote from: Eric Blank
How will I cheese now assholes?
Quote from: MrRoboto75
Always spaghetti, never forghetti