Motherboard or CPU problem

Discussion in 'Hardware' started by michael78, May 8, 2006.

  1. michael78

    michael78 Terabyte Poster

    2,085
    29
    141
    Hi guys and gals, strange problem, I just bought a case and built up a PC with spare parts I had around my flat. I built it up and it started up (I didn't have a monitor or anything wired up as it was late and just wanted to see that it powered on). The next day I coupled it up to my monitor and keyboard etc and it wouldn't power on.

    I've tried another board in the case as I thought it might of been the cases switch that was faulty but it worked. I've tried another heatsink as the one I was using came with a pentium and I thought it might of been that but it still wouldn't switch on. I've tried another power supply and it still wouldn't switch on.

    So basically it's come down to it being the motherboard or CPU. Can a faulty CPU stop the PC switching on. I don't want to buy a new motherboard and find it's the CPU thats faulty and vise versa. The power LED on the motherboard lights up as well so it's getting power though I know that doesn't mean it can't be faulty.

    I'm going to ask a stupid question that I should know but I covered the CPU with some silver paste (I've never done this before as I usually just use the supplied thermal pad with the heatsink) could I of short circuited the CPU with the silver paste?

    Cheers in advance

    Michael
     
    Certifications: A+ | Network+ | Security+ | MCP | MCDST | MCTS: Hyper-V | MCTS: AD | MCTS: Exchange 2007 | MCTS: Windows 7 | MCSA: 2003 | ITIL Foundation v3 | CCA: Xenapp 5.0 | MCITP: Enterprise Desktop Administrator on Windows 7 | MCITP: Enterprise Desktop Support Technician on Windows 7
    WIP: Online SAN Overview, VCP in December 2011
  2. r.h.lee

    r.h.lee Gigabyte Poster

    1,011
    52
    105
    slypie,

    One possibility that I ran into from real world experience, is that _something_ like an unused motherboard screw post may be shorting out the motherboard to the point the computer no longer POSTs. So to test that possibility, you need to run the motherboard outside of the case. Remove all the expansion cards from the motherboard, then unscrew the motherboard and remove it from the case. Place the motherboard onto something like the cardboard box that it originally came in. Connect the power supply connector to the motherboard. With the only components being the power source still in the case, the motherboard, the cpu, and the ram should be all the components that should be connected together. Then, using a small screwdriver or other metallic device, touch the two motherboard pins that represent the on/off switch. If the motherboard comes to life, and you hear the nice "beep" from the POST, then you have troubleshot that the problem is the case.

    If you don't get the POST beeps, then your BIOS or other simiiarly critical component on the motherboard may be dead. However, if you insert a "known good cpu" into the motherboard, then try to fire it up and it works, then you know that the first cpu may be bad. If however, the motherboard still remains dead after inserting a known good cpu into it, then you've cross-referenced a probable dead motherboard.

    Next test after the abovementioned test passes, is to double check if the cpu is ok. To do this, temporarily remove the RAM sticks. During the POST process, it should detect for RAM, and if RAM is not detected, then it should issue some sort of beep code for "missing RAM." This test also serves to verify if the CPU is ok or not because the POST process requires the CPU to be OK first before testing for RAM.

    If the abovementioned two tests pass, then you've got the nitty-gritty painful phase of troubleshooting where you reconnect each single cable, each single expansion card, to see if it could be something else causing the computer not to boot up. I've seen some unusual reasons for non POST to include: a required expansion card swap, a bad drive cable, etc.

    After verifying that the entire "computer" runs ok outside of the case, then the next phase is to determine if somehow the case is shorting something out. The simplest test is to put the motherboard back into the case as if you're going to screw it in, but don't screw the motherboard into the case. Then try to fire it up. If it fails to fire up, then you know it's due to grounding somewhere. If it fires up, then screw one motherboard screw in at a time, then fire it up. Once upon a time, I had a "No POST" problem with a motherboard and a rackmount server case. Using the abovementioned method, I determined that the motherboard was shorting underneath motherboard screw #7ish. So I used one of those small red "washer" on top of the suspect motherboard screw post in the case, then the motherboard fired up consistently after that.

    I wish you luck.
     
    Certifications: MCSE, MCP+I, MCP, CCNA, A+
    WIP: CCDA
  3. r.h.lee

    r.h.lee Gigabyte Poster

    1,011
    52
    105
    I almost forgot. Do a good ol' fashioned inspection of the cpu pins to see if any are bent. If they're bent, then there's no electrical connection between the CPU pin and the CPU socket, which also could cause no POST.
     
    Certifications: MCSE, MCP+I, MCP, CCNA, A+
    WIP: CCDA
  4. hbroomhall

    hbroomhall Petabyte Poster Gold Member

    6,624
    117
    224
    Depends on the paste - some of them do indeed have conductive properties, so great care is required. Which is why a new thermal pad is the preferred solution.

    Which brings me onto the other point - 'covered'? (Quoting). It should be a very thin smear - 'covered' sounds like rather a lot, although this may be an unfortunate choice of words! Too much paste can be as bad as none.

    First thing to do is fit a POST card and see if the processor is starting. If it isn't then either something is shot, or there is a serious incompatability in fitted hardware.

    Harry.
     
    Certifications: ECDL A+ Network+ i-Net+
    WIP: Server+
  5. michael78

    michael78 Terabyte Poster

    2,085
    29
    141
    Lee, Cheers for the reply it's appreciated. I should of mentioned I've tried building it outside the box with no luck. I've also tried starting it by touching the two points and it won't boot up. I don't have another CPU to try which is a pain. With the parts/spares I have I've managed to rule out the case/PSU/RAM. Would the motherboard at least power up with the CPU inserted? If so then this will allow me to rule that out I suppose. I never really thought of trying that but will do when I get home.
     
    Certifications: A+ | Network+ | Security+ | MCP | MCDST | MCTS: Hyper-V | MCTS: AD | MCTS: Exchange 2007 | MCTS: Windows 7 | MCSA: 2003 | ITIL Foundation v3 | CCA: Xenapp 5.0 | MCITP: Enterprise Desktop Administrator on Windows 7 | MCITP: Enterprise Desktop Support Technician on Windows 7
    WIP: Online SAN Overview, VCP in December 2011
  6. michael78

    michael78 Terabyte Poster

    2,085
    29
    141
    Harry, I think I might of over done it with the silver paste. I'm starting to think that in my stupidity I may have screwed the CPU by putting too much on. Hmm 13 years experience with PC's and I still do stupid things without thinking...:oops:
     
    Certifications: A+ | Network+ | Security+ | MCP | MCDST | MCTS: Hyper-V | MCTS: AD | MCTS: Exchange 2007 | MCTS: Windows 7 | MCSA: 2003 | ITIL Foundation v3 | CCA: Xenapp 5.0 | MCITP: Enterprise Desktop Administrator on Windows 7 | MCITP: Enterprise Desktop Support Technician on Windows 7
    WIP: Online SAN Overview, VCP in December 2011
  7. Keimos

    Keimos Byte Poster

    137
    3
    29
    Hi slypie,

    Check your PSU properly, a faulty PSU will still light up the motherboard, there is a tester available, see your local computer guys as most other diagnostics will not pick it up.

    Had a similar problem a month or two ago.

    Your CPU will work even if if the heatsink paste is dodgy, it will just overheat and computer will stop



    Keimos
     
    Certifications: Microsoft Office Specialist
  8. Bluerinse
    Honorary Member

    Bluerinse Exabyte Poster

    8,878
    181
    256
    You should only use a blob of paste about the size of a small pea.
     
    Certifications: C&G Electronics - MCSA (W2K) MCSE (W2K)
  9. michael78

    michael78 Terabyte Poster

    2,085
    29
    141
    Guys cheers for the replies, I finally got around to sorting it. I think I must of over used the silver paste. I removed it and put a small blod on and it worked.

    On a side note silver paste is probably the hardest stuff to remove in the world that stuff just gets everywhere.
     
    Certifications: A+ | Network+ | Security+ | MCP | MCDST | MCTS: Hyper-V | MCTS: AD | MCTS: Exchange 2007 | MCTS: Windows 7 | MCSA: 2003 | ITIL Foundation v3 | CCA: Xenapp 5.0 | MCITP: Enterprise Desktop Administrator on Windows 7 | MCITP: Enterprise Desktop Support Technician on Windows 7
    WIP: Online SAN Overview, VCP in December 2011
  10. hbroomhall

    hbroomhall Petabyte Poster Gold Member

    6,624
    117
    224
    Glad it worked in the end! :biggrin

    Harry.
     
    Certifications: ECDL A+ Network+ i-Net+
    WIP: Server+
  11. Mr.Cheeks

    Mr.Cheeks 1st ever Gold Member! Gold Member

    5,373
    89
    190
    As long as you dont do an American Pie job, then its all good!
     

Share This Page

Loading...
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.