0

A friend of mine gave me a Dell XPS 8930 tower because it was freezing intermittently with no obvious cause; he didn't have time to figure it out himself and was past the point of being able to return it. I figured since I do this sort of thing for a living, I would be able to easily fix this computer and use it as a nice gaming PC. Well, I have now exhausted my usual toolbox and I'm out of ideas on what could be wrong.

The problem: Windows intermittently freezes during tasks without a clear cause. I have been unable to identify anything in particular that will make it freeze. Sometimes it freezes shortly after booting, other times the computer can run for a while without issue. Because of this, I haven't had much success in isolating the issue. For example, I can play GTA (story or online) for a few hours and there might only be a short freezes; I tried Halo 4 today and it was unplayable. Sometimes the computer will lock up just browsing the web.

Hardware:

  • 460W power supply
  • NVIDIA GeForce GTX 1050Ti 4GB GDDR5 Graphics
  • 8th Generation Intel i7-8700 6-Core 3.20 GHz Processor
  • Dell stock motherboard (not sure on the model) with Qualcomm WiFi/Bluetooth module
  • ADATA 512 GB NVMe SSD
  • Toshiba 2 TB 7200RPM HDD
  • 32 GB DDR4 Memory
  • Windows 10 Pro (64bit)

Symptoms:

  • When a freeze happens, the computer does not always freeze completely immediately. Often, performance quickly degrades until the computer is completely frozen.
  • At the beginning of a freeze, animated buttons will still work, such as a change in gradient when hovering over a button.
  • In one scenario: I attempt to navigate to a web page, the browser gets stuck at "waiting for cache", I press the Windows key to open the start menu, the menu opens but is then frozen, the mouse cursor still moves, the mouse cursor eventually freezes.
  • Another scenario: I'll be playing a game, the screen freezes and the computer no longer accepts input.
  • Once while I was playing a game, the computer froze while an attached controller was vibrating; the vibration continued until the computer unfroze.
  • The computer will usually unfreeze after waiting a few minutes, but sometimes it needs to be forcibly powered down. Only once did it blue-screen with the error DRIVER_POWER_STATE_FAILURE.

What I've tried:

  • Reset Windows twice, once using the built-in fresh start to get Dell's OEM image, and once from a recovery disk to get a completely base install of Windows.
  • Made sure all drivers were updated, including BIOS, using Dell Command Update.
  • Ran quick and full scans using Dell Support Assist: no issues were reported.
  • Ran tests to check SSD and HDD using CrystalDiskInfo: both succeeded with no issues.
  • Ran Intel Processor Diagnostic Tool: succeeded with no issues.
  • Ran memory check (mdsched.exe): no issues were reported.
  • Disabled Windows automatic driver updates and changed power management settings to provided minimum 100% power to CPU, per this answer: Windows 10 Freezing

What am I missing here? At this point I am assuming faulty or incompatible hardware somewhere, given that I've reinstalled Windows several times. I unfortunately do not have extra hardware available to start swapping out different components, and I don't want to be throwing money at the problem without any direction. Any help here or other suggestions on what to try would be greatly appreciated.

0

Here is what I would do with any machine with similar behaviour.

You need some silver oxide processor / heatsink paste & a can or compressed air.

You will notice that as long as the machine is cool - you don't get the problem for 30 mins to an hour after switching on. Problem starts happening just at the point you think everything is ok. The problem is most likely heat related, but this does not make it any easier to solve as the only thing you can do is swap/replace with identical/near identical items.

Start with the processor and fan - dust free? fan spinning well? Pull/unscrew the heatsink - either a clamp or 4x screws. Do the same with the video card. Carefully remove the heatsink. Processor on videocards are non-removable.

Pop out the processor. examine for bits of fluff, dust, paper etc - I had a machine which had tiny bits of paper under the processor from the guy's messy desk. Paper conducts electricity and this solved his BSOD issues.

Get a can of compressed air. ALWAYS hold upright. NEVER hold at an angle. Bend the straw into the direction you need to go. Blast some air on the socket, replace processor.

Regardless to the condition of the blueish paste on the heatsink and processor, scrape off and replace with a decent silver oxide processor paste. try not to use too much - don't google how much to use or how to spread as there's far too much info on this, and in my opinion everyone is right who replaces the paste. Just make sure it is uniform across the PC - do the same with the heatsink on the video card.

Power on. Test. If problem persists ...

Find some identical ram/compatible ram. swap. Test again. Still glitching out?

find a video card - One that can run a game would be good, but not that easy to find if you are on a budget - Ask a good local IT shop if they could loan you a card for a small fee- My local shop does this and it proves to be a useful service.

Plugin, run your tests. still borked? - You may have a buggered machine on your hands - Stay positive though. Many dead machines are not dead at all, people give up too easily.

Get hold of a power supply for it. Swap. Test. Arrgh! still freaking out.

Bios update / reset. Check the dell website for a bios update and install - try resetting the existing bios but be careful. Take pictures of each screen just in case things get reset too far.

One of the tests above should help. ALL of the tests need to fail before you either consign it to non game use (Ubuntu! Linux Mint!) or whatever else.

If it still don't work. At least you have tried all options apart from a new motherboard, which most of the time is the most expensive option, especially with some DELL units.

3
  • Before you take things apart to apply "silver oxide" (sic, see en.wikipedia.org/wiki/Thermal_paste), check the CPU and GPU temperature. If they're not overheating, don't invite damage by smearing unneeded thermal paste on things. – DrMoishe Pippik Nov 18 '20 at 2:29
  • It's cheap, and easy to do, and as long as you have a tiny amount of experience removing a heatsink, I can't imagine much in the way of damage. In his diagnosis above he has done everything apart from checking the temp, so this is what he needs to do next. It doesn't matter what I call thermal paste, nor does it matter that I called it silver oxide paste. I even have it named on a tub as such by the manufacturer. I'm pretty sure that the reader won't get it mixed up with toothpaste, shippams crab paste, wallpaper paste etc. I knew that this post would attract the 'Thermal Paste Police' ! – JohnnyVegas Nov 18 '20 at 23:46
  • I am familiar with thermal paste and have built PCs in the past, but I do appreciate the clarification. Either way, I spent an hour of fairly demanding gaming and both the CPU (all 6 cores) and the GPU were between 50°C and 60°C at the end of it, and I did experience a freeze during the hour. So I don't believe this is an overheating issue. – Mike Nov 19 '20 at 2:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.