Boot-Up can take up to 5 resets to boot to my desktop

Discussion in 'Windows 64bit' started by Denise, Apr 28, 2009.

  1. Denise

    Denise Guest

    I have not found a proven hardware problem yet. It may be that timings is off.

    I'm asking you to stay out of any threads that I post here due to the rude
    and offensive remarks that you write to me. If you continue to harass me, I
    will file a complaint against you.
    Denise, May 2, 2009
    1. Advertisements

  2. Denise

    Denise Guest

    Hi westom,

    I purchased the psu on 21Jan09 from NewEgg. NewEgg does not list the
    warranty information for the psu on its website. I placed an RMA with
    Corsair a few minutes ago on the internet. I was given a case number and the
    page said that someone will be contacting me. If the psu is still under
    warranty and an RMA is allowed, a new psu should fix the problem unless the
    problem isn't associated with the psu, but it seemed to be the concensus of
    opinion that it may be the cause of the problem. I will first remove the
    daisychain power cables and connect the hdds to their own separate psu
    connectors before I RMA the psu because it may be the reason that the psu
    isn't working properly.

    Thank you for the valuable information regarding the use of a multimeter,
    reading the numbers from the individual cable wires and when and how they
    should be read.
    Denise, May 2, 2009
    1. Advertisements

  3. If you are reading low and abnormal voltages from your power supply,
    then it's should be obvious to the most casual observer that it is NOT a
    Windows problem. Of course an intransigent will continue the course.

    If you think pointing out the obvious is 'harassment' then you have more
    than hardware problems.
    Bobby Johnson, May 2, 2009
  4. Denise

    Denise Guest

    I have not stooped to your level in these posts, nor will I. You seem intent
    on insulting and harassing me, as your posts in this thread show. I shall
    find the correct channel to submit a complaint against you and seek to have
    you banned from posting in any threads that I may start in these forums, for
    you do not edify but do harm and you have maliciously stated that you will
    not help me. There is, therefore, no reason for you to post in any thread
    that I may start in the MS discussion groups.
    Denise, May 2, 2009
  5. If you're so 'computer savvy' then why don't you know how to get your
    computer to boot up without a floppy drive installed? It's really one
    of the basic functions of setting up a computer. I think this is proof
    enough that you really don't know as much about computers as you brag about.
    Bobby Johnson, May 2, 2009
  6. Denise

    westom Guest

    Unfortunately an assumption that is false. Again, read those
    previously posted concepts. A defective PSU can boot a computer. A
    perfectly good PSU can fail in an otherwise good computer. You are
    assuming black and white when reality has many shades of grey.

    Notice I did not say power supply. I used the word 'system'.
    Another assumption: a power supply is the entire power supply
    'system'. No. But with numbers from a meter, you not only know if
    the power supply is good or bad (maybe not waste your or Corsair's
    time), but you also learn what that simple thing in front of you has
    and is supposed to be doing.

    I can provide massive amounts of assistance. However I cannot even
    post one useful sentence if you don't provide useful facts. The
    quickest and easiest facts are numbers from a meter. Your labor - 30
    seconds. Then post those numbers - which will probably required more
    than 30 seconds. Then have more information from one post than the
    past 40 combined. That is not an exaggeration.

    Currently you don't yet know anything. You don't know if the supply
    is defective. You don't know if the power supply controller is
    defective. A warranty swap will still leave you completely confused
    if the problem is with parts of the 'system' - making a good supply
    act defective.

    Do you know if the problem is hardware or software? No. Probably
    hardware. But again, I still don't see the important facts that
    define it as hardware or an OS problem. Just another answer that
    comes after 30 seconds and that meter. Appreciate how much those
    numbers report. With those numbers, the next post would be twice as
    long as this one just replying with all the information that those
    numbers might report.

    Long before waiting for an RMA number, you could have had answers
    that say whether the RMA number will be a waste of time - long before
    they call. But again, 30 seconds, a meter, and a reply because of
    posted useful facts and numbers.

    Replacing a supply in speculation is called shotgunning. Most
    repair people quickly find themselves unemployed if doing as you would
    do. Replace something only after facts are known. In this case,
    maybe only 30 seconds later.
    westom, May 3, 2009
  7. Denise

    Denise Guest

    By process of elimination, it has come down to the psu, the daisychain
    connections to the psu, RAM timings or the adapter card.

    I've run some tests in order to find the problematic component. I have
    ruled out the mobo since it passed all the qualifications of the tests
    outlined in another post, along with the cpu and the RAM sticks. I've run
    deep stress tests (3 times in a row at 75% for 45 minutes). The fdd, cpu,
    cd/dvd drive, sound, video memory, RAM and video card did not fail.

    I don't believe that the adapter card is defective because, when I performed
    the deep stress test, it included the hdds connected to the adapter card and
    there were no hdd failures, but I will test it nonetheless.

    Heat inside the case is not causing the reset problem. It is only on
    boot-up that the computer fails, even when the computer is cold at boot-up.
    (I also have to reset several times after the computer has been turned off

    Five of the drives that are connected to the mobo are now on 5 second
    staggered spin-up and the problem still occurs. When I connect the new
    cables, 6 hdds will be on staggered spin-up. While the case is open, I will
    have the opportunity to see what is really happening inside the case and
    perform several steps recommended in this thread:

    1. I will do a Windows Repair.

    2. I will remove the daisychain power cables and place each hdd on its own
    connector to the psu.

    3. I'll disconnect 6 drives to see how the computer boots up, then
    disconnect some hdds and connect others, etc, etc, shutting down when I
    disconnect and connect the power cables and then booting up. This will allow
    me to see if the adapter card is problematic and if the psu has a problem
    spinning up only 6 hdds, whether to the mobo or the adapter card.

    4. I will place a very small amount of metallic compound on the heatsink.

    5. Setting the timings of the RAM may be needed but I need to read a lot
    more about it.

    After these steps are taken, I will know if a multimeter is needed to test
    the system.
    Denise, May 3, 2009
  8. Denise

    westom Guest

    Only deep stress test is also called burn-in testing. IOW execute
    the comprehensive hardware diagnostics (that the computer manufacturer
    provides), then heat the system up to something about 100 degree F,
    and repeat. Then cycle temperature down to a minimum (40 degrees F?)
    and execute diagnostics again.

    No software will 'stress' digital hardware. Those digital gates go
    on and off at standard speeds set by a clocks. Only thing that
    changes how digital hardware works is temperature and voltage - to
    change both internal gate timing and digital thresholds. You don't
    make digital hardware work harder. You simply cause change how it
    works by changing those two environments.

    Not sure what that stress test is. But the only valid stress test
    is one that executes all internal functions including those that you
    never use. That is called diagnostics.

    I don't know what 'daisy chain' is for the power supply. The power
    supply connects to various parts from a central point. Wires to each
    disk drive are so large as to not be relevant. For example, how many
    watts does a disk drive draw? Worst case 10? Average? More like 2.
    Those wires are good for at least 30 watts. But again, if cables
    voltages are a problem, 30 seconds with the meter would have
    identified the problem. As I said, that meter and only 30 seconds
    report numerous things you don't even yet know.

    Don't know why RAM is suspect because I do not see anything that
    says what does and does not execute. Don't waste time changing Ram
    timings - for too many reasons I will not even begin to explain until
    it becomes relevant.

    Myths constantly promote heatsink problems. Your previous BIOS
    numbers say that is not a problem. If the CPU is Intel, the system
    would only slow down; not crash.

    Hard drives do not crash any computer. If any harddrive has a
    problem, the system simply keeps retrying (keeps talking to the disk
    drive computer), reports any errors in system (event) logs, or even
    declares the drive as non-existent. IOW drives are not related to
    your problem. And that would be obvious again from the manufacturer
    comprehensive diagnostics or from disk drive manufacturer
    comprehensive diagnostics.

    But again, your power supply system is unknown. Three conditions
    exist - definitively bad, definitively good, or unknown. Until that
    supply 'system' is definitively good, then any other strange problem
    may only be a defective supply. Again, 30 seconds with the meter is
    essential before anything else can be moved from unknown to
    definitively something.

    Warning - metallic compound around the CPU? That will create more
    problems. Even thermal compound must never get out to the edge
    where CPU meets heatsink. That thermal compound can even cause
    electrical problems between CPU pins. AND if the heatsink is properly
    selected and machined, then it is more than sufficient even without
    thermal compound.

    But again, why are you trying to fix things that were not yet found
    defect? That is shotgunning. And shotgunning has a history of also
    exponentially complicating problems.

    Until the power supply system is 'definitively good', then most
    anything else can act defective. So you try to fix something that is
    not bad? That's how problems get more complex.

    Meanwhile, other facts I did not see and that you must have before
    making any changes. What do the system (event) logs report? What
    does Device Manager report? And what do comprehensive hardware
    diagnostics (only provided by better computer manufacturers) report?
    More examples of information that must be known before making even one
    hardware or OS change.
    westom, May 3, 2009
  9. If anyone should be banned from posting here it should be you since you
    are unable to comprehend the difference between 'hardware' and 'software.'

    Even though this is a group for SOFTWARE and specifically Windows 64-bit
    operating system, you insist on posting your multitude of HARDWARE
    problems in a SOFTWARE group!

    What gives you the right to post your hardware problems in a software
    forum and tell me that I don't have the right to tell you you're in the
    WRONG group?

    A bit hypocritical isn't it?
    Bobby Johnson, May 3, 2009
  10. Denise

    Denise Guest

    The deep stress test that I used was the Burn-In Test. I ran it 3
    consecutive times at 75% for 45 minutes, not using the method that you
    described. I had all tests turned on and nothing failed.

    A daisychain power cable contains multiple power connectors on one cable.
    My daisychain connectors contain 4 power connectors. The end of one cable of
    a daisy chain cable is connected to a connector on the psu and each connector
    on the daisychain cable is connected to a hdd. Since 4 SATA power connectors
    are on the daisychain, 4 SATA hdds are connected to one daisychain, thus 4
    SATA hdds are connected to one power connector on the psu. If the wires are
    good for 30 watts, the daisy chain may be pulling 10 watts per hdd at startup
    for a total of 40 watts each until spinup is complete, which may exceed the
    number of watts that each connector can provide.

    The 7 drives that I purchased are Western Digital Caviar Black WD1001FALS
    1TB 7200 RPM 32MB Cache SATA 3.0Gb/s 3.5". The voltage that they require to
    spinup isn't stated on the website.

    The remaining 5 drives are a combination of different drives that I removed
    from their external enclosures and installed in the case, some WD and some

    I'll remove the daisychain power cables and see if I can allot each hdd it's
    own connection to the psu. I'm not sure if this is possible. I haven't been
    inside of the case for a number of months but I remember that there wasn't a
    sufficient number of connectors for each hdd but someone in this thread said
    that there was. It may be that the connectors are molex power connectors so
    that I needed an adapter that the daisychain connector supplied. If there
    is a sufficient number of SATA connectors on the psu and I can connect each
    hdd to its own connector and it fails to fix the problem, I'll then buy a
    multimeter and test as you explained in a previous post.

    System event logs show mostly Information with several Warning for Avast.
    I've uninstalled and reinstalled Avast but I still receive the Warnings.
    Just today, I noticed some Errors for Application Popup. I noticed that
    there is a different type of popup that appears on the net. They hang like
    shadows on the page and will either disappear when the cursor is moved over
    something else or the popup is closed. There are no Warnings or Errors about
    the resets.

    The mobo and adapter card bioses and drivers are up-to-date, as well as all
    of my drivers.

    Thanks for your explanations . . . they're very much appreciated.
    Denise, May 3, 2009
  11. How do we get a moderator to close this thread since it is not related
    to this group?
    Bobby Johnson, May 3, 2009
  12. Denise

    westom Guest

    That sound like the stress test that is used to create hot spots in
    the CPU. Testing heat between larger bus driver trasistors inside the
    CPU does nothing for your system. The only 'stress' you can put on a
    computer is to temperature cycle it or change voltages. The only
    useful test at the system level is one that tests every function in
    the machine including those that your OS does not use. That is called
    a comprehensive system diagnostic. That diagnostic is unique to each
    manufacturers machine AND is provided for free with every machine on
    the disk drive, on a CD-Rom, and on the manufaturer's web site ... for
    free and for problems exactly like yours.

    Again, if the daisychain is a proble, you need the meter to see it.
    Nothing else will find it or identify the problem. But again, I do
    not see an exact description of what comes up on the screen before the
    machine crashes. Therefore we don't even know it the OS is involved
    or even if any disk drive is even talked to before the system crashes.

    Your various error messages (ie from Avast) tell us nothing because
    you did not quote any error messages and disply its critical numbers.
    Again, your replies are only as useful as information provided. That
    means the most important information are tidbits that you don't
    comprehend or that you think are irrelevant. And numbers from the
    multimeter (which those previous BIOS numbers say you desperately need
    now). Since you provide no critical error messages and numbefs,
    again, this post and all others remain useless. A reply here will
    only be as useful as the information you provide.

    The items that can actually crash a truly multitasking system are
    few and specific. For some reason, others have you even 'checking'
    things that are not in that list. For example, the BIOS are
    completely irreelvant. Disk drives will not crash a computer. But
    others will only recommend fixing the few things they understand.
    Almost nothing checked or tested is revelant until the multimeter has
    measured those voltages. Scary is fixing that heatsink with thermal
    paste when your own numbers said nothing is wrong.

    A parallel example. All doors were sticking. So the homeowner
    planed down all doors. Then the doors stuck again. He tried to fix
    everything before first finding the problem. He never noticed the
    foundation was crumbling. When the foundation was restored, all
    doors were defective due to plaining. Your computer's foundataion is
    its power supply 'system'. A foundataion is first verified by visual
    inspection. The visual inspection of a power supply 'system' are
    numbers from a multimeter.

    Remember those Speedfan numbers that make no sense? They said one
    thing. Nothing else should be looked at, tested, or even tried until
    a multimeter is purchased. Those numbers could not make more obvious
    why that meter is the next and only thing you do. Why did you ignore
    the most obvious fact? Those numbers suggested that anything else you
    do will be wasted time and labor.

    Get the meter because even if you replace that power supply, those
    measurements are still necessary. But again, a defective supply can
    still boot a computer. We see this often. Someone replaces the
    power supply. Sees computer work. Assumes supply is OK when the
    supply was defective. Only a meter can determine a supply
    definitively good or definitively bad. You need the meter if you do
    or do not replace that supply. So again, another post that tells you
    nothing new or useful because you did not provide any error messages
    and did not use the meter. Instead you keep suspecting even things
    that could never create your problem - ie BIOS, power cables, etc.
    westom, May 3, 2009
  13. Denise

    Denise Guest

    I have to start somewhere. Your recommendation is to buy a multimeter and
    test the wires coming from the psu. I prefer to remove the daisychain cables
    and connect the hdds to their own separate psu connector before I purchase a

    Different people have offered different solutions or suggestions to see if
    the problem is caused by other factors. I won't ignore those suggestions
    because I feel it would be foolish to do so. I started this thread to see
    what people think is causing my computer to require resets in order to
    completely boot to my desktop. I am taking your suggestion as seriously as
    the suggestions made by other people.

    I decided to start from the source . . . Windows. I just finished a
    Windows Repair and downloaded some updates from MS Update Center. My
    computer runs much better and faster than it did prior to the Repair. So it
    seems that this may partially be a Windows XP x64 problem afterall. As an
    added benefit of the Windows Repair, when it rebooted, it gave me a Critical
    Error message regarding my video card and gave me steps to fix the problem.
    One of the recommendations was to go into Display and lower graphics hardware
    acceleration to None and to uncheck the box for Enable Write Combining. The
    other recommendation was to check my video card's driver. I did so and it is
    the most up-to-date drive available. Since the Windows Repair, I have
    received no Warnings for Avast.

    I very much appreciate your advice but I'm planning to check into each and
    every suggestion given to me in this thread. It may just be that Gigabyte
    mobos are flaky and cause resets, as one person mentioned, but I'll reserve
    that opinion until I'm sure that I've done everything I can to find a source
    of the problem.
    Denise, May 3, 2009
  14. Denise

    Denise Guest

    P.S. The BurnIn Test will test many computer components and offers options
    to turn a particular test off. I chose to leave all the tests on except for
    Printer and Network because I don't have my printer connected and I don't
    have a network. All of the other components (which covered 14
    areas/components, in my case) were tested at the same time, not one after the
    other and, as I said, I ran them 3 consecutive times which equalled 45
    minutes and I ran them at 75 instead of the recommended 50, with the cpu at
    highest temperature and I ran the RAM Torture Test. There were no failures.
    It could very well be that the power supply cannot boot up 12 hdds at the
    same time, even though 5 are now staggered. But without running these types
    of tests and trying other simple things, I could miss out on a chance to find
    the problem.
    Denise, May 4, 2009
  15. Denise

    westom Guest

    Do you invent a solution from near zero knowledge? Or first learn
    from one who has been doing this for multiple generations?

    Start with a meter because that is where one starts - by using
    knowledge. If a power supply 'system' is unknown, then anything or
    everything can act defective - no matter how good or bad it is.

    Solution means moving each component from 'unknown' to 'definitively
    something'. After so much labor, what have you accomplished?
    Nothing. Not even one component on a list of 'definitively good' -
    even after a test that you somehow called a burn-in test.

    Why would you spend another 86,000 seconds doing nothing when 30
    second means immediate results? Why do you continue to ignore the

    Eventually you will fix the computer by repeatedly replacing parts
    until something works - spend how much money and learn nothing. You
    choice to keep doing what would get any tech quickly fired for being
    incompetent and hopeless. 30 second means immediate results.
    Instead you continue doing what gets technicians fired for

    Daisychained cables? More specuating on things that could not
    possiblity be a problem. Just another fact that 30 seconds with the
    meter would have made obvious. Why are you wasting everyones time
    with what could not be?

    Staggering drives: does nothing. Just another example of you
    wasting time using wild speculation - and no multimeter. Meter would
    have made quickly obvious how foolish that 'staggered drive' idea was.
    westom, May 4, 2009
  16. Denise

    Denise Guest

    It was a Windows problem. Ever since I did the Windows Repair yesterday, my
    computer boots right to my desktop. It takes about about 1.5 minutes to boot
    up all the way but I expect that with 12 drives. I've had my graphics
    accellerator on Full and my printer connected and turned on since this
    morning and I booted and rebooted my computer about 15 times since then to
    see how it would go and it booted all the way to my desktop every time. It
    also runs faster.

    So I was in the right forum all along. That can be put in somebody's pipe
    and smoke it, lol.
    Denise, May 4, 2009
  17. Denise

    Michael Mol Guest is not moderated, AFAIK. Most
    newsgroups aren't. Those that are typically end in .moderated.
    Michael Mol, May 6, 2009
  18. Denise

    Michael Mol Guest

    Keep in mind what I said about sudden poweroffs and reboots leaving
    your computer in an indeterminate state. In the process of one of
    these poweroffs, you may have wound up with a corrupted driver or
    other piece of software, which the "repair" fixed. It's still quite
    possible that there are other pieces of broken or corrupted software
    or data on your system that Windows Repair isn't able to recognize or

    Since this thread started, I even wound up able to provide anecdotal
    evidence; My Vista workstation locked up hard last week, forcing me to
    manually power it off. Ever since then, I've found that I've lost
    settings in a number of applications and subsystems. What I'm going
    to do, and what I would recommend that you would do now that your
    system seems to be behaving, is export my application settings from
    the registry, and then do a wipe-and-reinstall to guarantee that all
    of my software was in good shape.

    I do know of a number of my data files that were corrupted when my
    system locked up, and there's nothing I can or could do about them;
    They're gone. (No backups in my case; They were files with an intended
    lifetime of about twenty minutes.) But it's quite possible that I've
    wound up with corrupted tools and operating system components, and a
    reinstall will clear that up.
    Michael Mol, May 6, 2009
  19. Denise

    westom Guest

    Sudden power off does not destroy files (drivers). That problem was
    why we upgraded from FATxx filesystems to NTFS. Just another reason
    why better informed computer users avoided Windows 95 and upgraded
    directly to Windows NT.

    Sudden power off means the machine restarts from a default state
    where every hardware item restarts the same way with every power on.

    Anecdotal evidence without first learning underlying priniciples is
    junk science. A perfect example of junk science reasoning. In your
    case, files were probably destroyed or corrupted when your system was
    running. Later when you restarted the machine, those corrupted and
    missing files caused problems. So you blamed power off rather than
    first learning what was damage, why, and when.

    Massive numbers of OS files can be deleted from the drive and
    Windows will still keep operating. But on the next power on, Windows
    will never boot. So junk scientists blame power cycling.

    Sudden power off does not damage files stored on a drive for a long
    list of reasons. Worse are the many reasons why others just know
    sudden power off can harm a drive. All directly traceable to junk
    science reasoning because the 'expert' did not even know how hardware
    westom, May 7, 2009
  20. Denise

    Michael Mol Guest

    I'm familiar with the concept of journaled filesystems. (This is what
    you're referring to, I take it.) I also never said that it destroyed
    any drivers.
    To an extent; There's the CMOS battery which maintains certain data
    and live processes in the event of full power-off. In the case of
    soft power, there's still Wake-On-Lan, et al, which necessarily means
    a different type of wake. There's also resuming from different ACPI
    suspend states, though it's arguable at what point the operating
    system takes over from the BIOS.
    No, I wasn't blaming the poweroff for the corruption of my data. A
    poweroff merely means that anything not in persistent storage at the
    time of the poweroff is lost. What was remarkable was that the
    poweroff had to occur when the data in persistent storage was in an
    inconsistent state. I lost such things as file associations, registry
    settings and a number of source files I'd been working on were
    corrupted, and had to be retrieved from source control.
    I never said that Windows didn't boot. I never said that I had wound
    up with corrupted drivers or other pieces of information necessary to
    boot. I specifically said that I lost settings and data. I might add,
    though, that a number of the things that became corrupted as a result
    of the event weren't even files I was aware had been open or
    modified. Though I suppose it's plausible that my user registry file
    could count as "open" throughout the lifetime of my login session, and
    therefore subject to corruption if not completely written to disk.
    One of the source files I'd been working with in Notepad, upon
    reexamination, had mysteriously changed from a text file (I forget the
    encoding) to a file with garbage data. I didn't think to check at the
    time if the file encoding had been switched, but I did notice that the
    size on disk had been altered.

    However, if you'd like an example of a sudden poweroff that *could*
    result in a failure of Windows to boot, consider what could happen if
    tweaked device driver settings got corrupted, or if the HKLM registry
    hive were incompletely serialized to disk. Either could cause a
    failure that would necessitate booting into Safe Mode to clear things
    up. If one is in the process of installing device drivers or some
    such, an inopportune poweroff could render that process incomplete as
    Well, if you want to quibble on that point, there's the small matter
    of the disk cache. I'm not talking about the in memory caching of
    disk data that all modern operating systems do, though that can be a
    problem if not done correctly. I'm talking about the 2-8MB of data
    that most hard disks use as a buffer when reading and writing, so as
    to improve response times. In the event of a sudden poweroff, that
    cached data can be lost if the drive isn't capable of sustaining its
    own power long enough to flush it to disk. (Which I imagine is one
    reason you don't typically see such caches exceed 8MB.)

    I appreciate your expert attempt at preventing people from getting the
    wrong idea about the dangers of sudden poweroffs, though. You're
    right; it's not the poweroff that causes instability. It's the
    processes and system state mutation that the poweroff interrupted.
    NTFS and journaling filesystems can only go so far.
    Michael Mol, May 8, 2009
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.