1

I am facing an illogical problem on my wife's dell inspiron. The computer started to shutdown randomly while she was working. The problem started to evolve until the day when the computer always turned off before loading the windows. I tried to check the bios, but the computer turns off 10-20 seconds after getting into the bios.

Because the fan blew almost cold air, I thought it might not be an overheating issue, but motherboard, ram, power board or any other hardware issue instead.

In a desperate move, I tried to install ubuntu by the live CD and it surprisingly installed,loaded and works non stop since then. Hence, I ruled out the possibility of hardware failure.

However, the problem persisted in bios or if I stayed in the grub loader for more than 15 seconds. So, I thought that the only logical explanation would be "overheating" in bios because unlike the OS with halt instructions, I heard that bios tends to utilize 100% of the CPU.

I took the computer a part, cleaned the fan, heat sink and applied high quality thermal paste on the CPU die and GPU ATI chip. I know that this move worked out because the system shows idle temperature drops from 50 to 35 and 75 to 60 degrees Celsius for the CPU and GPU respectively.

Finally, I tried to load again into the bios, but the problem still persists.

I am wondering, if it is not a motherboard, ram, hdd, cpu (because once you get into ubuntu, the computer runs with no problems for very long periods) or overheating problem (because of the new thermal paste), what can the problem be?

Dell Inspiron N5010 CPU: Core i5 2.4 Ghz GPU: Mobility Radeon HD 5430/5450/5470 Memory: 2x2 GB DDR3 1333 Mhz OS: Ubuntu 16.04 / Windows 10 Battery Removed, running on AC power

8
  • if it had been internally cooking, the caps might have been cooking too. How many years was it working fine for? Are there any hardware stats you can aquire to look at , like voltages & temps and all, to observe while it is running?
    – Psycogeek
    Commented Jun 2, 2016 at 23:26
  • What utility gave you the 50 to 35 and 75 to 60 degrees C info? Great idea loading ubuntu to isolate some sources of error. The fact it still occurs if you stay in grub I would say points to a HW issue. Until you can get past the bootloader/bios causing an overheat/power-off (if it is indeed an overheating issue), I don't think you can rule out A) a real overheating/HW issue, or B) a HW issue with the temperature sensor. Is the machine/motherboard still under warranty?
    – tniles
    Commented Jun 2, 2016 at 23:27
  • @Psycogeek: The machine has been working since 2010, that is almost 6 years. I am not sure how can I retrieve voltage in Ubuntu
    – mowienay
    Commented Jun 2, 2016 at 23:32
  • @tniles09 I am using sensors in terminal. No the machine is not under warranty, it is a very old one :) issues with temperature sensors makes sense. Do you have any idea on how to troubleshoot that?
    – mowienay
    Commented Jun 2, 2016 at 23:34
  • @mowienay that could prove difficult. The place to start would be to find out where the sensor(s) are, and then find out whether you can get some replacements to swap and try. Before attempting those efforts, it would be a good idea to see if you can identify other possible issues (i.e. in case it's not an overheating issue). Not sure what that would be, though... bad upgrade on one of those chips, perhaps? Have you tried removing all non-essential cards (e.g. if the GPU is the issue, but on a card, does removing it change things?).
    – tniles
    Commented Jun 3, 2016 at 0:02

0

You must log in to answer this question.

Browse other questions tagged .