Boot issue

Recently, my two and half year old computer (at home) began having issues with starting up. Originally, it would only fail to start up when the video card slipped and became unseated, but now it is afflicted with a more complex and aggravating problem. Essentially, I'll turn on the computer and all the fans will start up, but it won't POST. The video card fan will achieve the same high-revolution rate it does at the onset of of a normal start-up, but instead of settling down just prior to the POST, it will maintain this speed indefinitely. There is no signal to the monitor, despite the obvious fact that the video card is seated properly and functioning. To me, it seems as if the system is halted is pre-POST mode, as if it is missing something that signals it it should proceed forward with the boot process.

The problem would be less aggravating if the computer could never start up (because at least I could reasonably conclude it was the video card, or at least glean some insight from others who have this issue - Google only yields cases in which the computers can't start), but I've found a way to consistently make it function properly. After a bit of experimentation and frustrated troubleshooting, I noticed that if I let the system sit in its frozen state for a couple of minutes, shut it down, and then restart it, it will invariably boot properly. While it's nice to know that I can still use the system, I don't think this workaround is optimal for the long-term health of the computer, nor does actually the solve the underlying (and unidentified) problem with some aspect of the circuitry. Has anyone heard of this issue before, or is at least able to contribute something to solving it?


Gigabyte GA-P35-DS3R
Core 2 Duo E6750 with AC Freezer 7 Pro
4 GB Crucial Ballistix DDR2 800
Evga 8800GTS 512
Samsung SpinRite 400GB
Lite-On DVD Burner
Seasonic S12-550HT
20 answers Last reply
More about boot issue
  1. Well, the first thing you can conclude is that some part is working once warmed up but not cold.

    It sounds as if the power ready test completes because the video card is getting power.

    Have you visually inspected the capacitors on the board?
  2. Yeah, I kind of had that concluded, but I had no idea what part, if any, would function normally when idled, but not when cold. I have not inspected the capacitors, but I will. What should I be looking for when I do this?
  3. That depends... I'll look up your board..

    Well, they're all solid caps, so it's less likely any failed. Have a look for bulging or leaks anyway though. You also should look around the mosfets. Those are the flat black squares. Check those for any scorching.

    I don't think your board has a buitl-in speaker. If you need one, here:

    That would let you hear any BIOS beeps you may be getting when the board fails to post.

    The only other thing I can suggest would be to try with each stick of RAM separately.
  4. Okay. I checked the capacitors, and they are all intact, as far as my eyes can tell. I didn't smell anything near them, so I'm going to assume they are all working properly. I haven't visually inspected the mosfets, but most of them are near capacitors, and I didn't smell anything from them - can I presume they'd at least make a faint scent evocative of burning electronics? Once I'm finished error checking the memory, I'll look at them and corroborate.

    I do have a built-in speaker, and I do get BIOS beeps when something is really wrong, such as the graphics card being seated improperly. Moreover, it does beep affirmatively when the systems POSTs successfully. That is what makes it so perplexing - nothing the motherboard is telling me suggests hardware failure, but something is obviously wrong with some component in the system.
  5. The memory is fine, and no mosfets are burnt out. I did some more searching, and people thought it might be a PSU or motherboard issue. Is there any way to test the stability or functionality of these components that doesn't involve acquiring another physical part?
  6. A Seasonic PSU failing is not so likely as others, but of course any part can go bad. Certainly a failing PSU can produce these symptoms... the power is too far out of spec for the board to run but the PSU sends the ready signal anyway, before it's really warmed up.

    What I was getting at with the speaker is that you want to see if you can produce an error code from the PC when the POST fails. A "no memory" or "no video" code would indicate that the board is at least PARTLY working at these times, and so it would help narrow down your search.

    If it does come down to being either the PSU or board, it's fairly hard to sort out at that point. You can use a multi-meter on a molex to see if the 12V is within spec, but even if it was say 11.7V that might be just enough for a failing board part to get in the way.

    You could just contact Seasonic and tell them what is going on and what you have done. They'll probably take the unit in and service it or replace it. But, then you would have a wait, and you might still have the problem.

    Ah wait! I just dreamed up a test :p Sorry I'm not awake yet.

    1. Disconnect the PSU from all devices, including the board.

    2. Follow these instructions to make your PSU run:

    3. Let it run for a while so that it's good and warm. The MB and such should be completely cold when this is done.

    4. Turn the PSU off and pull out the paper clip. QUICKLY hook up the 24-pin, 4-pin, and 6-pin video power and turn on the PSU. then start the computer.

    5. Does it boot with no problem? Then you now have your culprit, the PSU.
  7. I followed the steps, let the PSU sit for three hours, and was met with the same issue once I restarted. Barring a multimeter reading (which I can't get at the moment because I have no multimeter) suggesting it is failing, I'll have to rule out the PSU as the cause of the problem.
  8. I'm inclined to agree. Must be the motherboard then.
  9. In my endeavors to isolated the problem further, I conducted a test to ascertain whether the need to warm up was thermal or electrical. After blasting the motherboard, concentrating around the CPU and RAM, for three to five minutes, I successfully turned the system on, which led me to conclude that something literally needed to be heated up before it could function properly, presumably on the motherboard. What possible could require such temperature change? Is there maybe a small crack somewhere near the CPU socket that inhibits normal operation while cold, but is closed once it expands due to heat? I realize I will likely have to RMA the board at some point, but I want to understand exactly what the issue is before I proceed with replacing anything.
  10. That's very hard to say. Any contact could be potentially spread when cold and closed once warm. Doesn't have to be a crack.
  11. nukchebi0 said:
    . Barring a multimeter reading (which I can't get at the moment because I have no multimeter) suggesting it is failing, I'll have to rule out the PSU as the cause of the problem.

    Without a way of measuring the PSU pins, you cannot rule out the PSU.
  12. I tested the PSU with the multimeter and got the following readings:

    3.3V: 3.42 (constant)
    5V: 5.04-5.06 (consistent fluctuation)
    12V: 12.19 (constant)

    Can I safely rule out the power supply?
  13. A multimeter can't catch the small fluctuations that might cause your system to be unstable. Ultimately, you can only really know about a PSU with an oscilloscope and load tester.

    Your voltage numbers can (and will) change with varying loads... that's not to say that a multimeter is useless. At least you know that unloaded your voltages are in range.

    Still, I think your heat test was fairly indicative of a MB issue of some sort.
  14. I tested the 5V and 12V rails under load (I used an open molex connector) and without load, and the readings were identical. I guess I can't conclusively say the PSU is not to blame, but I think the evidence strongly points to its innocence

    At this point, I'm going to assume that it is a motherboard issue. Am I right in thinking there is no way to rectify it short of getting a new motherboard?
  15. It would be worthwhile to breadboard, just to eliminate any case-type shorts. Check all the power contacts for scorching, that sort of thing. Make sure there's no conductive thermal interface material running out the side of your CPU...

    Chances are it's more an internal MB issue, but rule out the external.
  16. What is breadboarding?
  17. There is no scorching anywhere, and the thermal interface material is contained within the contact face of the external heat sink and CPU heat spreader. I didn't have time to breadboard, but based on what I can tell, it is not an external motherboard issue.
  18. That's my best guess then, the board has some small issue.
  19. Yeah. It also won't turn on anymore, so I am going to go ahead and conclude the motherboard needs to be replaced. Thank you for the help.
Ask a new question

Read More

Homebuilt Graphics Cards Computer Systems