add arrow-down arrow-left arrow-right arrow-up authorcheckmark clipboard combo comment delete discord dots drag-handle dropdown-arrow errorfacebook history inbox instagram issuelink lock markup-bbcode markup-html markup-pcpp markup-cyclingbuilder markup-plain-text markup-reddit menu pin radio-button save search settings share star-empty star-full star-half switch successtag twitch twitter user warningwattage weight youtube

Resurfacing Freezing/Reboot Issues

RStovetop

16 months ago

Hello all,

This is my first post so I will try to explain as well as possible, but I'm sure I will forget some things. I had built a PC in January of this year, mostly for gaming purposes. It worked fine for about 3 months with no issues whatsoever. About the end of March, whenever I would play any sort of GPU/CPU demanding game, I would get a sort of buzzing sound loop and a total PC freeze that could only be resolved with a hard shut down and reboot. This occurred pretty regularly and at anywhere from 10 mins to an hour into playing. This would also cause difficulty restarting the machine as I would normally have to attempt to power it on multiple times after a crash before I could actually get a signal to the monitor and a normal boot. This was also coupled with a DRAM led lit on my motherboard during the times when it would NOT boot correctly.

I started to dig into the issue and tried multiple things to no avail: memtest86+ turned up no issues, hard drive scan gave no errors, hardware monitors showed temps were fine, tried different combinations of RAM seating and testing, turned on XMP in bios, updated BIOS, updated grpahics drivers, maybe more but these are the things I remember.

Kind of limited in the amount of time I had to mess around with the thing, this ended up going on for a couple months. Eventually I came across a suggestion to check my PCI-e configuration on my mobo. Turns out, I had my GPU in a slot that was being reduced to 1x speed (4x slot that got dropped to 1x because of another card I had in the first slot, some kind of weird MSI board thing). I reseated my GPU into the top PCI-e slot that is supposed to run at 16x and my problems vanished. From most things I read, this shouldn't have really been the problem or caused crashing, but i figured maybe this board gave different voltages through different PCI-e slots and who cares anyway, it's working.

It did work absolutely perfectly for about another 3 months and now last night the problem has suddenly returned, albeit with slightly different symptoms. Now when it crashes, it instantly attempts to reboot itself and if it is successful I get a message that "Over clocking has failed" and options to enter setup, load default values, or restore to last known working settings, none of which fixed the issue. Otherwise it is the same issue that occurs in about the same time frame of running a game and has the same issues booting up on occasion with a lit DRAM led when it cannot boot.

A lot of things I have read suggest it may be a PSU issue or RAM esp. with the led that is on, but I hate to just start buying stuff without some input. Aside from obviously looking for a solution, I'm trying to learn a little bit about what is happening here or if anyone has seen something similar. Is there anything I can rule out at this point? Can I be confident it is a hardware issue? Did reseating the GPU actually DO anything or was it a happy coincidence? Why/how could this happen in intervals where there is no crashes and then constant crashing?

Any help would be greatly appreciated and I can try to perform whatever testing would help arrive at some kind of solution if needed.

Here is the build:

Intel i5 6600k Hyper 212 EVO CPU Cooler MSI Gaming B150 M3 ATX LGA1151 Motherboard Kingston HyperX Fury 8GB (2x4GB) MSI GeForce GTX 1050ti EVGA 500W 80+ Bronze Power Supply

Gigabyte - GC-WB867D-I PCI-Express x1 802.11a/b/g/n/ac Wi-Fi Adapter ^ This is my wireless card that is the only other thing in my PCI slots in case that matters.

https://pcpartpicker.com/user/RStovetop/saved/NnHPsY

Comments

  • 16 months ago
  • 1 point

As far as the PCIe voltages. The board should deliver the same voltages to the same pins no matter what it is with the 16x being longer it has more pins, but the shared pins between 16x and 1x should be the same.

You mentioned it gave you an error about OCing. Are you? If yes, then dial it back a bit. if no, then we will move on.

Based on what info you provided I would say it is a hardware issue. RAM or PSU is a good bet. I would start with RAM since it is freezing/ crashing. If it was PSU then it would just shut off. I did have a problem with RAM that memtest did not pick up. I swapped the RAM and have had no problems. I would swap the RAM and re-install the OS. what is probably happening here is that the RAM has a bad sector or chip and a certain part of the OS is running through that which is corrupting it. The rest of the OS is fine so you get a fine boot, but once it needs to access that section that is corrupt then it can't and crashes.

Borrow or order some new RAM and then ideally use a new drive to install the OS (if you don't have a new drive then the one you were using is fine, just be sure to format it completely). See if that fixes your issue.

Let me know and I can help you further!

  • 16 months ago
  • 1 point

Thanks so much for the quick reply. It will probably take some time to try those things out since it looks like I'll be ordering some replacements, but I'll let you know how it goes. Also, you are correct, not doing any OC despite the message, other than XMP which appears to use 2133 @ 1.2 regardless of if I have it on or not.

As far as the problem going away for months after switching PCI-e slots, was it really just a coincidence? Or did I potentially do something to circumvent the issue for a period? In the first round of troubleshooting, I tried to be careful to try one thing at a time so I could pin down the issue but I could have done something and not realized. Should the problem have stayed indefinitely if it is the RAM?

  • 16 months ago
  • 1 point

One other quick question before I leave you alone ;) What is the purpose of using a new drive? Do I risk anything by trying the new RAM on the same drive, or should I just replace both to be safe?

  • 16 months ago
  • 1 point

I would say that it is purely coincidental. The problem that I had with RAM that I mentioned only happened about once a month. I installed Windows, it was fine for about a month and then on a boot it crashed and went into self recovery and was unable to recover. Reinstall Windows, wait a month, do it over again. RAM issues are weird. It SHOULD have stayed as a problem, but with RAM you never know. I had another issue that turned out to be RAM, but it was only a problem when you went to print. The entire OS worked fine, but when you printed it would only print dots. I spent a lot of time smashing my face into the keyboard trying to figure that one out. You might have re-seated the RAM which changed which sectors the OS went to, but that is about it.

  • 16 months ago
  • 1 point

When your checking your temps, leave the temperature program running in the back ground and play a game. After awhile, alt+tab out of the game and check your temps. Make sure your checking your temps from a load and not idle.

If you are over-clocking, perhaps try resetting your BIOS to defaults by turning off and unplugging the system and removing the battery.

  • 16 months ago
  • 1 point

Overclocking with a B150 motherboard..............

Sort

add arrow-down arrow-left arrow-right arrow-up authorcheckmark clipboard combo comment delete discord dots drag-handle dropdown-arrow errorfacebook history inbox instagram issuelink lock markup-bbcode markup-html markup-pcpp markup-cyclingbuilder markup-plain-text markup-reddit menu pin radio-button save search settings share star-empty star-full star-half switch successtag twitch twitter user warningwattage weight youtube