1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

PC Crashing/Locking up after new CPU

Discussion in 'General Hardware Issues' started by nabed4t, Oct 23, 2016.

  1. nabed4t

    nabed4t Established Techie7 Member

    Hey guys,

    I’ve started having some issues with my PC after upgrading my CPU. My specs are as follows:

    Windows 10 Pro

    MSI 890FXA GD70 Motherboard

    AMD FX 6300 CPU using a Corsair Hydro Series H70 Core CPU Cooler

    Radeon R9 290x GPU

    16GB DDR3 1600 G.Skill Ripjaws F3-12800CL9D RAM

    Okay, so a few days ago up upgraded my AMD Phenom II x4 965 CPU to the FX 6300 due to needing a CPU that supported SSE4.1 instruction sets for a couple of new games. I had to update the bios on my MOBO (to version 1.15) so it was compatible with AM3+ sockets (the MSI website says the CPU is compatible with my MOBO). Since installing the new CPU I have cleared my CMOS and now experience occasional system lock ups accompanied by an audible “RRRRRRRRRRRRR” sound if the speakers are on and occasional and random system restarts where the PC will spontaneously restart. I can’t seem to see a constant factor that is causing these issues as they happen at seemingly random times like playing games (although I have played MGS V for a solid 5 hours with no issues), to the computer downloading stuff overnight to the 2 times it locked up while writing this post and listening to some music on MusicBee.

    I have stress tested the CPU for 12 hours on both the simple and torture tests on Prime95 with no errors, the CPU’s temp also never got above 40 degrees Celsius. Which brings me to a side question, I’m using Open Hardware Monitor to check CPU temps and at the moment it is telling me that the CPU is sitting at 8 degrees Celsius. But, if I used Aida64 the temp reading is 25 degrees. Has anyone had this type of temp discrepancy with Open Hardware Monitor before?

    Moving on from that, the CPU is not overclocked, it is sitting on its factory 3.5Ghz and no other component in my system is overclocked either.

    I have check the systems even log, Window logs / System and there are a few warning and error titles there, although I do not know how to interpret them.

    Has anyone got any suggestions on what my next step could be? A friend suggested updating the chipset on my MOBO as there may be a communication issue with the CPU and MOBO due to the age difference but I can’t seem to find any info on the MSI website regarding updating the chipset.

    Thanks in advance, if I’ve forgotten any important info please let me know.
     
  2. Digerati

    Digerati Super Moderator Techie7 Moderator

    Obviously, you cannot have a CPU cooler than your ambient (room) temps without some serious alternative cooling - like refrigeration! So the 8°C cannot be right unless you live in Canada and all your windows are open - or the sensor is bad (and that is possible - they are very low tech devices).

    The problem is, there is no industry standard for sensors, sensor location, or sensor names. So sensor makers, CPU makers, motherboard makers and hardware monitors all do it their own way.

    The sensors simply convert the temperature they sense to a hexadecimal value. The sensor maker publishes that information so users of that sensor know how to translate that value back into a format that makes sense to us (F° or C°). If all the players from the sensor to the monitoring program are on the same page, 23°C should always be seen as 23°C. But because there are no industry standards (the ATX Form Factor standard, for example, does not dictate temperature monitoring), it does not always work that way.

    Note this is NOT an exclusive problem with Open HW Monitors. I have seen the same issues with SpeedFan, Speecy, RealTemp, CoreTemp and others.

    You might check out HWiNFO64 and see what it says. I have found this to be the most informative (if not too informative!) and less prone to inaccuracies.

    Note the Chipset drivers are typically part of the BIOS. The only W10 motherboard drivers I found for that board are for the network interface and sound so I would not worry about them.

    The fact you were able to play and stress test for hours with no problems does not suggest a BIOS, driver or Windows problem to me. I always want to ensure I am feeding my electronics good, clean stable power - this is especially true when troubleshooting intermittent problems. So if me, I would swap in a known good power supply. You might also try running with just one stick of RAM at a time - swapping in the other(s) if no problems and see if you can isolate the problem to one stick.
     
  3. nabed4t

    nabed4t Established Techie7 Member

    That is an excellent point, as I'm in central Queensland in Australia, the ambient temperature is definitely not below 10 degrees.

    This is a great little program, I've just downloaded it and it's great. The CPU temp is around 10-15 degrees Celsius higher than the one displayed by HW monitor.

    Unfortunately I do not have another power supply to test at this stage. Although these issues were not present before I installed the new CPU, is it just a coincidence that they are occurring now?

    I spent a large chunk of yesterday playing Metal Gear Solid 5 for... testing purposes... with no crashes at all. However, I was running bitTorrent over night and the computer had locked up in the morning. I hard reset it and started bitTorrent again and was listening to some music on MusicBee when the system locked again and gave that "RRRRRRR" sound. Again, I hard reset the PC but for an FF code on the MOBO while booting and the PC just hung on a black screen. I reset again and it booted fine.

    Should I do a MemTest86 to see if the RAM is faulty?

    One last thing, in HWiNFO64, under the RAM section on of the sticks of RAM are labelled using garbled letters where the other 3 are correct. Could this be an indication of defective RAM? I've added a screenshot to explain.

    Thanks again Digerati.
     

    Attached Files:

    • DAL.jpg
      DAL.jpg
      File size:
      110.7 KB
      Views:
      3
  4. Digerati

    Digerati Super Moderator Techie7 Moderator

    I don't believe in coincidences - but I don't rule them out either. That garbled text is odd but it is showing 16GB of RAM too.

    You can run MemTest86 (note that is MemTest86, not MemTest). But note no software based tester is totally conclusive. But if MemTest86 reports any errors, even 1, the RAM is bad.
     
  5. nabed4t

    nabed4t Established Techie7 Member

    Thanks Digerati, I will start here and see how we go with the RAM.

    Do you suggest testing one stick of RAM at a time?
     
  6. Digerati

    Digerati Super Moderator Techie7 Moderator

    Well, I already did. It is not totally conclusive either because on rare occasions, some RAM can test and run just fine alone, then fail when paired with another stick.

    That said, it really is important to eliminate the PSU from the equation since everything inside your computer depends on stable power from the supply.
     
  7. nabed4t

    nabed4t Established Techie7 Member

    True you did, sorry about that.

    Other than acquiring a new PSU, is there anyway to test a power supply?
     
  8. Digerati

    Digerati Super Moderator Techie7 Moderator

    Not conclusively. To conclusively test a PSU, you need an oscilloscope or a power supply analyzer - expensive and sophisticated tested equipment that takes special skill to use and understand the results. That's why for most home users, swapping in a known good one is the only way.
     
  9. nabed4t

    nabed4t Established Techie7 Member

    I was transferring some files today and I got this error on a blue screen ""whea_uncorrectable_error"

    It said something about running into an error and needed to collect some error info then restarted.

    Does that shine a light on what might be happening?
     
  10. Digerati

    Digerati Super Moderator Techie7 Moderator

    Did you test your RAM?

    You might put your old CPU back in and see if the problem happens.
     
  11. nabed4t

    nabed4t Established Techie7 Member

    I've been swapping out sticks of RAM and then used the PC for a while and I've still been getting the hangs, restarts and that BSOD.

    I've actually sold the old CPU on ebay. I've got to post it on Monday, so I have a couple of days to test it again.

    After a little bit of googling, that "whea_uncorrectable_error" is apparently related to voltage issues with the CPU or driver/software problems?

    If it is a voltage issue to the CPU, does that strengthen you theory on the PSU going out?

    I've been looking at PSU's on the change I do need a new one, if I could get your opinion on these two that would be great, it's been ages since I bought my Thermaltake Toughpower 850AP.

    Power Supply 1

    Power Supply 2
     
  12. nabed4t

    nabed4t Established Techie7 Member

    I've been using the old Phenom CPU all day with no issues.

    Would the BIOS be causing the issue? I'm running the most recent version, should I roll back a version or would that be a a waste of time?

    The MSI website specifically states that the FX 6300 is compatible with my board provided I have updated my BIOS. I've emailed MSI, however I doubt that I will get a reply.
     
  13. nabed4t

    nabed4t Established Techie7 Member

    Pleasant surprise, MSI just replied to me (I've found in the past if you're seeking assistance on an old model item, you may not receive any help)
    They have confirmed that the CPU should work with my MOBO - they asked if any of the pins were bent on the new CPU. I've checked again and there are none that I can see.
     
  14. Digerati

    Digerati Super Moderator Techie7 Moderator

    You might reset your BIOS - especially if you are overclocking. You can easily get by with a decent 650W PSU. I like EVGA Golds.
     
  15. nabed4t

    nabed4t Established Techie7 Member

    Nothing is overclocked, I just thought that maybe the CPU is needing more voltage. I have HWInfo64 run a log up to when it crashed, it shows voltages and usage of the CPU up to the point where the system locked. I'm not sure how to interpret them but I've attached them to this email on the off chance you may see something I don't. (scratch that I tried uploading the excel spread sheet, but it is not supported on this website)

    I've also called a couple of IT services around town and they suggested that maybe the MOBO is just not compatible with the new CPU.

    I also called the suppliers I bought the CPU off, I can send it back but as I've sold the old CPU, I will be with out a PC for a while.

    So, while writing this post, I got another BSOD, this time with a different code:
    Stop Code: SYSTEM_SERVICE_EXCEPTION.

    Might have to look into getting a new MOBO...
     
    Last edited: Oct 29, 2016
  16. Digerati

    Digerati Super Moderator Techie7 Moderator

    It does not work that way. If the motherboard supports the CPU, and your PSU is adequate, the voltages will be right. If the PSU were not adequate, you would not be able to play your game for 5 hours without problems - assuming the PSU was not intermittently faulty.

    But according the QVL (qualified vendors list) for your motherboard, the FX-6300 is indeed supported by your board.

    While bad capacitors are not nearly as common as they used to be, you might inspect your motherboard for any leaking or bulging caps.
     
  17. nabed4t

    nabed4t Established Techie7 Member

    So I rolled back the bios version to 1.12 from 1.15 and, touch wood, I haven't had any issues for the last 18 hours...

    Maybe that BIOS version was buggy?
     
  18. Digerati

    Digerati Super Moderator Techie7 Moderator

    Could be. I would send MSI tech support and report your problems. They may ask you to beta test a fix but it would be good to know they are paying attention to you.
     
  19. nabed4t

    nabed4t Established Techie7 Member

    I will email them and see what they say.

    So, over 24hrs since rolling back the BIOS version and no crashes or BSOD.

    BUT, when I booted the computer on this morning it posted the FF code and wouldn't boot. I reset it and it booted fine. There is still something not quite right, but it seems that once the system is booted that it is fine... at least for now.
     
  20. nabed4t

    nabed4t Established Techie7 Member

    Emailed MSI, got a super fast reply that goes as follows.

    "Regarding your concern,it is really strange,we have tested FX 6300 processor on this MB strictly before,this MB can support it.And sorry to say that we have no plan to release new bios for this MB.Thanks!"

    At least I can now use my PC with some stability.