SOLVED help needed with 4770 CPU overheat and shutdown

Discussion in 'Troubleshooting Help' started by ibace, Apr 20, 2019.

  1. ibace

    ibace Member

    Joined:
    Aug 10, 2015
    Messages:
    142
    Location:
    Bathurst NSW...
    Recently purchased a used 4770, when it arrived swapped out the 3258 and in went the 4770...finished install and booted up...heard the 'beep' and thought...great!! It works...booted into Windows, checked it was recognised in System info; activation status unchanged...all good!

    Turned it off and installed a GTX 780 card booted up into Windows, installed new driver, rebooted and getting ready to run a benchmark or two...

    Into Windows, started 3D Mark Vantage 11 watched the screen loading up; went outside and returned a couple of minutes later to find system rebooting again...hmm, what's going on? Turned it off and removed heatsink/CPU to determine it was properly installed, and paste evenly distributed on the heatsink...all good.

    New paste, verified heatsink installed properly...so booted to Windows and ran the Heaven Benchmark...soon as it loaded and began to play, I clicked F9 to start benchmarking and within 8 seconds the system shutdown and was rebooting itself...obviously a problem of some kind...you think?

    Once up and running I decided to download and install HWiNFO as I had previously used it and knew it was suitable to watch real time temperatures...started it up then opened Aida x64 and started a Stability Test...selected all tests and very soon it was reporting issues with CPU overheating...then CPU throttling was engaged to try to limit overheating...didn't work as shutdown in 4 mins 56 seconds.

    So restarted and once up and running started the Stability Test again [which is where I got the info on overheating to 100°C and shutdown in 4:56]. This time I elected to remove CPU stress test and limited it to the ram and the GPU...off we go again...

    Came back in about 20 minutes...looked like system was shutdown, but it was only asleep...it woke up with mouse movement, though the Aida x64 & HWiNFO were not running...seemed a bit unusual?
    Restarted Aida x64 clicked on the Stability Test start and up came the error message of system shutdown due to overheating of the CPU [even though it wasn't selected to be stressed] but this time it lasted 5 mins and 17 secs [big improvement of 21 seconds longer with CPU not being stressed].

    So it seems to me there is an issue with the CPU [previous owner says it was working correctly in an office machine that was taken out of service and he parted it out] and he has offered to exchange it next week, but I am wondering if there is any possibility that delidding and reapplying compound may help with the issue...or anyone have another suggestion or thought and so, am asking opinions with regard to my options. Visually there is no evidence of any obvious issue, nor signs of overheating...
     
  2. Bold Eagle

    Bold Eagle Member

    Joined:
    Jun 28, 2008
    Messages:
    8,474
    Location:
    Brisbane
    Assess the CPU and system health in the BIOS - should be under the Hardware Health tab or something to that affect.

    It reads like someone has tried to delid it and done a very poor job?
     
    Last edited: Apr 20, 2019
  3. OP
    OP
    ibace

    ibace Member

    Joined:
    Aug 10, 2015
    Messages:
    142
    Location:
    Bathurst NSW...
    I would've thought that if an attempt had been made at delidding, there'dve been some evidence of same..and I haven't noticed any. Have booted to BIOS as per recommendation; CPU temp is shown at 56°C; and appears stable at that. Will leave system running in BIOS and check intermittently...and report back.
     
  4. Bold Eagle

    Bold Eagle Member

    Joined:
    Jun 28, 2008
    Messages:
    8,474
    Location:
    Brisbane
    What's your local temp - 56C is high at idle unless you are in a 40C room.

    What is the HS? Is it the stock Intel shit? Can you take a pic?
     
  5. OP
    OP
    ibace

    ibace Member

    Joined:
    Aug 10, 2015
    Messages:
    142
    Location:
    Bathurst NSW...
  6. Bold Eagle

    Bold Eagle Member

    Joined:
    Jun 28, 2008
    Messages:
    8,474
    Location:
    Brisbane
    You have the stock Intel heatsink (HS). Those lugs can and do fail or 'pop' and a corner lifts - making them next to useless. You said it is 14C and yet the CPU is at 56C that clearly shows the HS is not working!! A normal range is 10-15C above ambient for a loaded and working CPU. You need to try and push those lugs don't correctly or if to worn replace that HS.

    Frankly your problem is obvious - ambient is 14-15C, CPU idle should be 25-30 - that is range 10-20C above the ambient, your getting 55-56. That HS is not working! It can be simply tested by running the PC and a temp monitoring app (HWMonitr for example) and then try and wriggle the HS.

    Your problem is obvious!
     
    Last edited: Apr 21, 2019
  7. OP
    OP
    ibace

    ibace Member

    Joined:
    Aug 10, 2015
    Messages:
    142
    Location:
    Bathurst NSW...
    when I first removed the heatsink to verify that it was installed correctly; the paste/compound had been evenly spread around the whole of the heat spreader on the CPU and on the copper core of the heatsink. At the same time as I did that job, I used a different heatsink [of the exact same type] to reinstall with new compound. But as you've mentioned; the heatsink cannot be working correctly [or it is working to a decent degree but the CPU temp is extreme, and being cooled down to the 55-56°. I will remove and check the CPU again with reference to the spread of the compound...I would think that would indicate what level of pressure is impacting on the CPU. Will advise...
     
  8. Bold Eagle

    Bold Eagle Member

    Joined:
    Jun 28, 2008
    Messages:
    8,474
    Location:
    Brisbane
    Just consider you'd expect about a 10-20C variance on ambient anything above that shows the HS is not working!
     
  9. DSTM (Dougie)

    DSTM (Dougie) Member

    Joined:
    Jun 2, 2018
    Messages:
    393
    Location:
    Western Sydney
    Maybe one of the 4 lock down lugs have let go.Pull upwards on each lug, and see if one has let go.Terrible old case for ventilation.
    Those HS's are $5 at my local computer shop.
    Are those temps with the side cover on?
     
    Last edited: Apr 21, 2019
  10. adamsleath

    adamsleath Member

    Joined:
    Oct 17, 2006
    Messages:
    13,550
    Location:
    Sunnybank Q4109
    send it back; you have the offer to return it.
     
  11. philquad

    philquad Member

    Joined:
    Jan 22, 2004
    Messages:
    920
    Location:
    nelson bay
    sounds like that 4690k i sent you ha
    i hate cpu's
     
  12. DSTM (Dougie)

    DSTM (Dougie) Member

    Joined:
    Jun 2, 2018
    Messages:
    393
    Location:
    Western Sydney
  13. OP
    OP
    ibace

    ibace Member

    Joined:
    Aug 10, 2015
    Messages:
    142
    Location:
    Bathurst NSW...
    Gonna close this thread; it may well have been the forebodings that the motherboard was at fault, cause it has crapped itself [can only think it was the frequent on/off swap CPUs, switch out ram, try different ram slots, boot to windows and then bios, I was beginning to think the motherboard's temperature sensor was inaccurate as at one time as the BIOS was showing 56°C temp, I had turned it off by switching the PSU off, taken off the heatsink and removed the CPU, all within seconds and I was surprised that the CPU never even felt warm to the touch! You'dve thought that if it was 56°C it would've felt at least warm to the touch?

    There were a lot of restarts and swapping components trying to sort out the apparent overheat issue, when all of a sudden, turned it on, and then something like 50 short and fast error beeps then shuts off [very hard to get an accurate count as the beeps were so short but I do think 50 was the valid number] restarts for a couple of seconds, then off and back on again and through the 50 beeps cycle again...noisy little effer! I left it overnight with the CPU, BIOS battery and the power supply disconnected in an attempt to get the motherboard to work...but it didn't!

    So whatever the cause of the problem...it no longer matters - and I don't have another motherboard to check the CPU in...so I'll just send it back and get a refund.
     
  14. philquad

    philquad Member

    Joined:
    Jan 22, 2004
    Messages:
    920
    Location:
    nelson bay
  15. OP
    OP
    ibace

    ibace Member

    Joined:
    Aug 10, 2015
    Messages:
    142
    Location:
    Bathurst NSW...
    yeah...saw those last night, or more correctly early this morning...saved to my watchlist. Will likely get one of them...have gone the Chinese route with a Huanan ZHi mobo and a 16 thread Xeon....working great and seems to be about a 20% improvement over my 6700K...just running at stock is all I ever do...
     

Share This Page