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

    There were about 5 of those reboots out of about 300 boots so I'm not
    concerned about them right now. I also thought that any "fix" would fix them
    too but I mentioned them in order to give all the info for a diagnosis, or at
    least a thought.

    I'm running a memory test right now and will be running a different memory
    test later. I'm also going to be running BurnIn tests for other pc
    components. I'll at least be able to disqualify some hardware components as
    a possible culprit.

    I ran HD Tune tests for 6 random hdds. I averaged out the results:
    Transfer Rate Minimum 55.4
    Transfer Rate Maximum 108.7
    Average 88.5
    Access Time 12.2
    Burst Rate 137.3
    CPU Usage 2.3%
     
    Denise, Apr 28, 2009
    #21
    1. Advertisements

  2. Denise

    Denise Guest

    Wow. . . Get a grip man! Everyone who comes to these groups and forums ask
    questions in order to resolve their pc problems.

    If people, such as you, don't like my questions, they don't have to answer
    them but it's not going to stop me from asking. I've learned a lot by
    asking. If you don't want to answer my questions, don't, but insulting a
    person who's asking questions and trying to learn is bad form.
     
    Denise, Apr 28, 2009
    #22
    1. Advertisements

  3. Denise

    Michael Mol Guest

    Denise, you may or may not be aware, but the forum you're posting to
    also appears on the wider Internet as a newsgroup. In fact, you can
    see this particular thread here:
    http://groups.google.com/group/micr...read/thread/de9eb18171500844/5cf689d47c7e8519

    To answer your simple question, Yes, it is possible. But that depends
    on whether or not there is a bug in your motherboard's southbridge
    hardware, a bug in the driver for your southbridge, or a bug in some
    other driver that manages those components. As an addendum to that
    answer, No, it's not very likely. I, along with everyone else in
    here, would likely wager that that isn't the root cause of your
    problem. Which is why that particular question hasn't been answered,
    and why people have been trying to help you find a more likely cause.

    As far as your problems, some questions and observations:
    * Do you have temperature monitoring inside your system?

    Most motherboards these days have thermal sensors which the operating
    system makes available to programs the user can run. Likewise, all
    modern hard drives have internal thermal sensors and other diagnostic
    features (It's called "SMART", if you want to look into it). If
    you're planning on running twelve drives, the temperature and health
    of the electronics in your system are going to be a very big issue.
    Once you have everything installed, you should set up some sort of
    monitoring tools to keep an eye on system and disk health. As the
    temperature inside your computer increases, you're more likely to see
    transient (temporary) hardware failures which will often manifest as
    system lockups and reboots.

    * Simply having an 850W PSU is Not Enough.

    Having an 850W power supply is not necessarily going to be sufficient
    to spin up all of those drives at the same time. For one thing, a
    computer's PSU provides power at a few different voltage levels, and
    with multiple sets of circuits. Different manufacturers balance the
    power capacity of a PSU between these different voltage levels
    differently; The cheapest often design their PSUs to provide an
    incredible amount of capacity on the least-used voltage level, but
    still market their units as 850W. Additionally, there's the question
    of how many "rails" the PSU provides. The more rails provided, the
    better the quality of the power the PSU will provide.

    Why is all of this important? Consider what happens when you start a
    powerful microwave or an electric drier in a building with poor
    electrical wiring. While the breaker might not flip, the lights will
    dim. This is because the appliance is trying to draw more power than
    the circuit can provide, and the voltage level on the line drops.
    This is unavoidable, to an extent. The way electricity works, *any*
    time you draw current, the voltage level will drop. Most of the time,
    we don't notice this because the wiring we deal with is of a high
    enough quality to avoid it.

    A similar thing happens inside a computer when a hard drive spins up.
    As the motor inside the hard drive spins up the platters, it draws a
    large amount of current. (By "large amount", I mean much more than it
    does after it's already running.) There will be a voltage drop as a
    result of this, though if you only have one or two hard drives, you
    almost certainly won't notice. But you've got twelve...

    To tie this back into the discussion of how many rails a PSU provides,
    think of each rail as a separate circuit in your home. Each circuit
    is capable of carrying a certain amount of the PSU's overall power
    capacity, but if you put too many devices on the same circuit, and
    each of those devices powers on at the same time, you're going to
    overload the circuit. In your home, this is like having a big old TV,
    stereo, refrigerator and wall-unit air conditioner automatically turn
    on after a power loss--you'll pop the breaker. In a computer's PSU,
    you won't usually pop a breaker, but the voltage on the rails will
    drop out of spec, and that could potentially cause other problems.

    The generally accepted solution for this problem, when you need to use
    a large number of hard drives, is to stagger the powering-up of the
    drives. There may be a feature in your motherboard's CMOS to
    configure this. Professional shops often use hardware RAID cards,
    which will usually have such a feature. Failing that, getting
    secondary drive controllers with such a feature might be a solution.
    And if that doesn't work, you could get external eSATA drive
    enclosures that will at least allow you to give the drives a power
    supply separate from your core system's.

    * Random lockups and freezes are usually indicative of a hardware
    failure.

    Lockups and freezes such as the ones you've described are more often
    than not the result of temporary hardware failures. More often than
    not, those temporary failures are the result of high internal
    temperatures, which is why I've focused on discussing temperature and
    power issues up to this point. However, it's *also* possible that
    there's a hardware incompatibility between your motherboard and your
    RAM modules. There's more to RAM specs than simply being "PC6400";
    There's also timings to consider. All modern motherboards will
    attempt to identify the correct timing settings for your RAM modules,
    but it doesn't always work; Sometimes a motherboard's BIOS has a bug
    in it that prevents it from Just Working, other times the RAM modules
    are out of spec.

    Some things to try:
    1. Install system temperature monitoring tools. If you regularly see
    temperatures higher than 50C on your motherboard sensors, or 40C in
    your hard drives, you should take steps to get those numbers down.
    This might involve using a better thermal paste between your heat sink
    and CPU, more fans blowing air into and through your system's case,
    and hard drive cooling bays. In fact, for a faux-server setup like
    you describe, those should be a given measure, anyway.

    2. Install and run stress-testing tools. Prime95 is a useful tool to
    run under Windows. You'll need to run several copies at once, though,
    if you want to stress each of your CPU cores. Prime95 is quirky in
    that if you want to run it multiple times, you need to install it,
    then copy its directory once for each additional simultaneous run. If
    you want to stress four cores, you need to install it, and copy the
    directory to three other places, and then run each copy. (I had to
    test an eight-core Xeon system this way, once.) This is for testing
    the CPU. Prime95 also has tunable options for testing RAM, which I
    would also suggest you do. However, when you use those options, you
    should only run one copy of Prime95; Running multiple copies for RAM
    testing would likely lead to spending more time swapping to disk than
    actually testing RAM, which defeats the purpose of stressing the
    hardware memory infrastructure.

    3. Try tuning your RAM module's timings. Read up on RAM timings, and
    use what you learn to tweak your timings. After each tweak, reboot.
    If your system boots, stress-test your RAM. If the stress-test
    doesn't report any errors after a day or two (Yes, it can take that
    long for problems to appear), try using the system for your planned
    workload.

    One final thing to remember. Every time your system hangs, freezes or
    spontaneously reboots, your system has been left in an indeterminate
    state. This means that you've very likely lost data, and you risk
    your will be less stable than before, even if there is no other
    hardware or software cause for instability. As a result, I would
    strongly recommend a clean wipe and reinstall, once you've got all the
    other issues taken care of. (As I've explained told students I
    tutored, it's always easier the second time. And moreso for each
    subsequent time, until it becomes second nature.)

    Hope that helps.
     
    Michael Mol, Apr 29, 2009
    #23
  4. Denise

    andy Guest

    It doesn't matter. When you have more than one disk drive, the drive
    that the Bios boots is determined by its position in the Hard Disk
    Boot Priority setting under Advanced BIOS Features in Bios setup.
    Have you checked the Event Viewer?
    The information you need is the correspondence between this list
    and the SATA2_# connectors on the motherboard.

    For the GA-EP45-UD3R motherboard, it's
    IDE0M - SATA2_0 - Disk 0
    IDE0S - SATA2_2 - Disk 1
    IDE1M - SATA2_1 - Disk 2
    IDE1S - SATA3_3 - Disk 3
    and if four drives are connected, the Disk # in Disk Management would
    be as shown.

    So if you want the C drive to always be Disk 0 in Disk Management,
    connect it to SATA2_0.
     
    andy, Apr 29, 2009
    #24
  5. Denise

    Denise Guest

    Hi Andy . . . thanks for replying and your suggestion.

    "So if you want the C drive to always be Disk 0 in Disk Management, connect
    it to SATA2_0."

    I have Drive C connected to port SATA2_0 on the mobo, as recommended in the
    mobo manual.
     
    Denise, Apr 29, 2009
    #25
  6. Denise

    Denise Guest

    P.S. I didn't see this in your post earlier . . .

    "Have you checked the Event Viewer?"

    I've checked it on a number of occasions. It's doesn't show anything about
    the resets. Most of the entries are Information with one or two Warnings a
    day, usually about Avast. I uninstalled and reinstalled Avast but I still
    get some Warnings.
     
    Denise, Apr 29, 2009
    #26
  7. Denise

    Denise Guest

    Last night, I ran MemTest and BurnInTest at the same time. The settings for
    the
    BurnInTest were set higher and/or to the highest for each category, i.e,
    RAM, video
    memory, all hdds, sound, optical data drive, fdd and cpu to 75% instead of the
    recommended 50%. I ran the BurnInTest 3 times in a row with no failures. I
    would
    have run them at 100 but we had hot weather on the east coast and the temp
    in my pc room was 82º, and the pc had been working for hours so that heat was
    already a
    factor. While the tests were running, I opened Microsoft Word and Paint,
    and opened and closed folders that contain hundreds of video files, and I was
    on the internet. I ran the BurnInTest RAM Tortue Test. The results were that
    it had 249,550 coverage, cycled 2, and performed 1,896 trillion operations.
    I also tested my RAM using RightMark Multi-Threaded Memory Test and MemTest.
    MemTest ran for approximately 10 hours with no problems.

    Due to the results of the tests, I believe that I can say that my RAM, cpu,
    hdds, fdd, cd/dvd drive, audio and video memory have no problems and are
    causing no problems.

    Without having a multimeter or a power supply tester to perform a psu test,
    I can't say with 100% certainty that the psu is not faulty but, with 12 hdds,
    the optical data drive, the fdd and the cpu being tested along with sound and
    video memory (all at the same time) at a minimum of 75% for 45 continuous
    minutes, I believe that my psu is ok and that it has sufficient power for my
    system.

    To test the video card, I downloaded and installed Video Card Stability
    Test/Stress
    Testing & Benchmark Test. It ran non-stop for over 10 hours. The video
    card did not fail.

    To test the mobo, I followed these steps provided at several sites on the
    internet:

    1. Download free diagnostic tools from your motherboard manufacturer's
    website or other reliable websites. Follow the directions to test the
    motherboard. If a BIOS upgrade is available, download and run it. (I COULD
    FIND NO DIAGNOSTIC TOOLS AT GIGABYTE'S WEBSITE. INFORMATION ON THE NET
    INDICATES THAT IF THE LED LIGHTS ARE ON, THAT IS THE MOBO TEST. ALL THE LED
    LIGHTS TURN ON WHEN I BOOT UP. ALSO, THE LATEST BIOS UPDATE HAS BEEN
    INSTALLED)

    2. Change the CMOS battery if the computer will not keep the right time or
    save configuration settings. (COMPUTER KEEPS THE RIGHT TIME AND SAVES
    CONFIGURATIONS SETTINGS SO BATTERY REPLACEMENT NOT NECESSARY)

    3. Turn on the computer. Press the appropriate key to get into the Setup or
    BIOS configuration. Write down all BIOS settings. Set the BIOS to default
    settings.

    4. Remove unnecessary cards and peripherals, like sound cards and modems.
    Disconnect the hard drive.

    5. Take out the memory and processor. Reseat them gently and firmly in their
    slots.
    Check for loose connections on chips and jumper caps. Turn on the computer
    and note whether it powers on and displays a hard drive error. If not,
    continue to troubleshoot. (NO LOOSE CONNECTIONS ON CHIPS OR JUMPER CAPS. NO
    MISSING JUMPERS. NO HARD DRIVE ERROR DISPLAYED.)

    6. Swap the power supply and video card with parts from a working computer.
    Test the computer again. If it powers on, replace the components one at a
    time until you find the one causing a problem. (NO DIFFERENCE WAS NOTED)

    7. Replace components one at a time until you find the card causing a
    problem. Turn on the computer. (THE AFOREMENTIOED WAS NOT AN ISSUE) Enter the
    BIOS configuration screen. Configure the settings to the previous setup.

    Since the above steps showed no problems, I can only assume that my mobo is
    in good working order.

    I couldn't test the controller card on another computer. I have Driver
    Genius which has a hardware troubleshooting tool. It shows that the device
    is working properly.

    In addition, the drives that are connected to it are recognized in My
    Computer and Disk Management and these drives were also tested during the
    BurnInTest at 75% with no failure.


    I believe that the results show that my mobo, cpu, psu, fdd, cd/dvd drive,
    sound, video memory, RAM, controller card and video card are working properly
    and well. If there is a hardware problem, it is not with the hardware itself
    but with a setting. I changed no settings in BIOS or CMOS except for the
    boot order sequence. I set the CD/DVD drive as first in the boot order
    sequence, followed by the hard drive and third is the fdd. I also turned off
    the Gigabyte splash screen and enabled keyboard.

    I thought that I would post my findings, as I usually attempt to do, so that
    they may help someone else who is having the same/similar boot problem.
    Unforturnately, I didn't find a fix for the need to reset several times
    before my pc boots to my desktop, but it's fortunate that there are no
    problems with my hardware.
     
    Denise, Apr 29, 2009
    #27
  8. Yes, this is ok.

    Drvies are really identified by the physical path (the ARC Path) as
    defined by rules set out in your hardware BIOS. Generally, it's either:

    * on-board IDE
    * on-board SATA
    * add-in controllers

    or

    * on-board SATA
    * on-board IDE
    * add-in controllers

    Sometimes, you can swap on-board IDE and SATA around with a setting in the
    BIOS.

    If you have multiple add-in controllers, they will be ranked according to
    expansion bus rules, and then searched in ranking order. Basically,
    whichever is reported first by the BIOS has the highest priority.

    Drives are also generally numbered (from 0) in the order in which they're
    found using the ARC path.

    If your drive with the Windows installation on it is attached to an add-in
    controller, and other drives have subsequently been connected directly to
    the motherboard, or to higher-priority ports on the same controller, or to
    ports on a higher-priority controller, I would expect the drive number to
    change.

    Whether the Windows partition is identified as "disk 0", "disk 4", or
    "disk 63490" doesn't really matter - as long as the bootstrap loader (ie
    NTLDR) is found (first Active partition found going through the drives in
    ARC Path order), and the ARC Path it uses (in BOOT.INI or BCD) to find the
    OS is correct.

    Your PSU may not be up to this challenge. 12 is an awful lot of drives.
    SATA does not require an FDD to be present (and FDDs don't use the IDE
    master/slave terminology). You might need an FDD to supply SATA drivers
    during OS installation (dependent on OS).

    However, by default, most BIOSes are configured to expect to see an FDD,
    and will generally halt the POST if they don't find one - but those
    behaviours can be changed with settings in the BIOS.
     
    Steve Foster [SBS MVP], Apr 29, 2009
    #28
  9. Denise

    Denise Guest

    Denise wrote:

    "Drive C was reallocated as Disk 4 in Disk Management.
    Is this okay?"

    Steve Foster [SBS MVP]" wrote:

    "Yes, this is ok."

    Thank you for letting me know. :)


    To sum it up then, Drive C is connected to the correct port on the mobo
    according to the mobo manual and is not connected to a port on the adapter
    card, and since it's ok that Drive C is allocated as Disk 4, I can then look
    elsewhere for the boot problem. Since the computer finds my OS, I would
    assume that the bootstrap loader and ARC path is correct. If this isn't
    true, then I would have to either perform the steps that MS outlined to
    correct this or disconnect all drives except Drive C and format it again.

    I have only 1 controller card, unless the video card is considered a
    controller card. All of my hdds are SATA.


    Steve Foster [SBS MVP]" wrote:

    "Your PSU may not be up to this challenge. 12 is an awful lot of drives."

    This is a possibility although I was assured that it would be sufficient.
    When choosing the parts, I was told "The only time you're coming (I THINK HE
    MEANT TO WRITE 'COMPUTER') will require a lot of power is during the initial
    spin up. You're computer will not support staggered spin up, so you're drives
    are likely to pull 30W a piece when you boot up, then they'll pull about 6-10
    watts depending on the drive after that. 650W is more then sufficient for
    this build." I decided to purchase the 850W psu though, due to the number of
    drives that I was going to install and he later agreed with me.

    I was also warned that there could be problems if the psu was too large. ".
    .. . A psu works in it's most efficient state when it's outputing 50-70% of
    it's maximum sustained output. Anything below that normally means the PSU
    works less efficiently. I'm stating this because going over board with the
    PSU can be bad." This person knew that I would have 12 internal hdds. It
    may still be that my psu isn't large enough, though, as you said.


    Steve Foster [SBS MVP]" wrote:

    "Sometimes, you can swap on-board IDE and SATA around with a setting in the
    BIOS." I've seen an option in BIOS or CMOS to set the type of configuration
    I wanted to set up. The choices were IDE, RAID and I believe the third one
    was ACHI (or something like that). Since I wanted to configure my pc in the
    IDE format, I was instructed to chose IDE even though my hdds are SATA. This
    may not be the option you had in mind though.

    Thanks very much for your advice, Steve. I very much appreciate it! I grew
    up in the 50's where our toys were rollerskates and jumpropes, and pong (if
    anyone remembers that game) didn't come out until I was in my 20's, and a 1T
    computer filled an entire room. I didn't have a computer until about 8 years
    ago. I enjoy learning about them, how the components work, about drivers and
    software. Three years ago I didn't know a cpu from a psu but I learned
    about them from people like you who are willing to share their knowledge. :)
     
    Denise, Apr 29, 2009
    #29
  10. Denise

    Paul Shapiro Guest

    The total PSU power sounds completely adequate, even for startup. But many
    of these higher output psu's have multiple 12-volt output lines. While the
    total supported is 850 watts, you should check how that spec is distributed
    among the connectors. If you have all 12 drives on the same 12-volt line,
    you could have plenty of total power capacity in the psu but still get too
    much voltage drop on the 12 volt line being used by all the drives. The more
    power you draw, the more the voltage decreases from it's idle value.

    Since your problems are only during boot, this seems reasonably likely. As
    you stated, the peak power draw at startup is the critical demand. What
    brand and model psu and sata controller? Check the sata controller
    manufacturer for firmware updates, and check your motherboard manufacturer
    for bios updates.

    You could also do some testing with half the drives removed. Try using 6 at
    a time, and then swapping with the other 6, to see if any
    drives/cables/controller ports are problematic. The good part of having the
    worst problems at boot is testing can be reasonably fast.

     
    Paul Shapiro, Apr 30, 2009
    #30
  11. Denise

    Ron Guest

    Toss up between your gigabyte motherboard (had a lot of problems with mine
    and many other reported cases of this on the net - do a search) and your PSU
    with all your hard drives!

    I am a professional photographer and only use 5x 1TB internal hard drives,
    all others for backup are external drives. Many gigabyte motherboards are
    flaky.

    Cheers

    Ron
     
    Ron, Apr 30, 2009
    #31
  12. Denise

    Denise Guest

    Hi Paul,

    Thanks for the info. :) I'll run the checks you suggested as it seems
    that it's possible that it may be the cause of the boot problems.

    I don't have all 12 drives on one connector. I purchased 3 daisychain power
    connectors so that 4 drives are on one psu connector.

    I just recently purchased new data cables because the ones that came with
    the adapter card were over 40"+ and I needed only 24" so I folded them over
    several times and used a rubberband to keep them folded. I'll be opening the
    case soon . . . it's very heavy with all that's in it and I need someone to
    lift it, lol.

    These are the specs of the psu:
    Brand CORSAIR
    Model CMPSU-850TX

    Spec
    Type ATX12V 2.2 / EPS12V 2.91
    Maximum Power 850W
    Fans 140mm Thermally Controlled
    PFC Active
    +12V Rails Single
    PCI-E Connectors 4 x 8Pin
    CrossFire Ready
    SLI Ready
    Modular No
    Efficiency > 80%
    Over Voltage Protection Yes
    Overload Protection Yes
    Input Voltage 100 - 240 V
    Input Frequency Range 50/60 Hz
    Input Current 12A
    Output [email protected], +5V30A, [email protected], [email protected], [email protected]
    MTBF >100,000 Hours
    Dimensions 5.9"(W) x 3.4"(H) X 5.9"(L)

    Features:
    Guaranteed to deliver rated specifications at 50°C

    80%+ energy efficiency at 20%, 50% and 100% load condition for less heat
    generation and lower energy bill.

    Active Power Factor Correction (PFC) with PF value of 0.99.

    Dedicated single +12V rail offers maximum compatibility with latest
    components.

    Supports the latest ATX12V v2.2 and EPS12V 2.91 standards and is backwards
    compatible with ATX12V 2.01 systems.

    High quality Japanese capacitors provide uncompromised performance and
    reliability.

    Powerful +5Vsb rail with 3A rating.

    Guaranteed compatible with dual-GPU configuration.

    Auto switching circuitry provide universal AC input 90~264V.

    Over Current/Voltage/Power Protection, Under Voltage Protection, and Short
    Circuit Protection provide maximum safety for your critical system components



    The controller card:
    Brand HighPoint
    Model RocketRAID 2320

    Specifications
    Type SATA II
    Internal Connectors 8 x SATA II
    Interface PCI Express x4
    Transfer Rate Up to 300MB/s for each SATA II drive port
    RAID RAID 0/1/5/10/50 JBOD
    Operating Systems Supported Microsoft Windows 2008/2003/XP/ 2000/x64
    Edition/Vista 32 and 64 bit
    Linux
    Mac OS X 10.4 (Tiger) and 10.5 (Leopard)
    FreeBSD
    GPL Licensed Linux Open Source Driver into Kernel 2.6.25
    Dimensions 6.3" x 2.6"

    Features
    Features Online Capacity Expansion and Online RAID Level Migration (OCE/ORLM)
    Native Command Queuing (NCQ)
    SAF-TE enclosure management
    Hard drive activity and Failed LED support
    Staggered drive spin-up support
    Hot swap and hot spare
    Write-through and write-back cache support
    Quick and Background initialization for quick RAID configuration
    Online array roaming
    BIOS booting support (INT13)
    Single RAID Cross Adapter Support
    64bit LBA for over 2TB support
    Automatic RAID rebuild of failed drive
    S.M.A.R.T drive monitoring for status and reliability
    Browser-based RAID management software
    Command Line Interface (CLI)
    SMTP for email notification

    RAID Management Utilities:
    BIOS
    Browser-based RAID Management
    CLI

    Packaging
    Package Contents RocketRAID 2320
    User Guide
    Driver Disk
    4 x SATA Cable
    Other Accessories



    I've updated the driver for the controller card and other drivers and BIOS.

    --
    Denise



     
    Denise, Apr 30, 2009
    #32
  13. Denise

    Denise Guest

    Hi Michael,

    Thank you so much for the information and for your recommendations!

    I put the specs of my psu and controller card in a reply to Paul Shapiro's
    post (2nd from last post) so I won't post them again here as they're quite
    lengthy. I understand what you said about the draw of power and, from what
    others have said in this discussion, it seems that my psu might not be
    sufficient. There is reserved space in the case to install an additional
    power supply and I will give it serious consideration. Does the psu have to
    be the same size and brand?

    My mobo has settings for monitoring heat and it has S.M.A.R.T. If a
    component is too hot, I have it set to beep to alert me but it has never
    beeped. The temperature of my hdds are approximately 31ºC when I first boot
    up. The temperature for any given hdd has never risen above 41ºC after my pc
    has been on for a day or two. I use HD Tune to check the temperatures but
    I'm not sure how accurate it is. The case has 4 fans and the video card has
    its own fan.

    I use CPUID Hardware Monitor which shows the temperatures of the cpu, psu
    and hdds.

    CPU:
    Value Min Max
    Core 0 40º 39º 40º
    Core 1 36º 36º 37º
    Core 2 33º 33º 35º
    Core 3 33º 33º 35º

    PSU (the values change ±.1 as the test runs):
    Voltages Value Min Max
    CPU VCORE 1.14V 1.14V 1.28V
    VIN1 1.92V 1.92V 1.92V
    +3.3V 0.08V 0.08V 0.08V
    +5V 5.05V 5.03V 5.05V
    +12V 2.24V 2.18V 7.17V
    -12V -0.45V -0.58V -0.45V
    -5V -0.45V -1.54V -0.45V
    +5V VCCH 3.66V 3.66V 3.66V
    VBAT 3.17V 3.17V 3.17V

    Again, I don't know how accurate the test is and I don't understand the psu
    voltages, whether they're good or bad.

    When I purchased the cpu, I didn't know if it would come with compound so I
    purchased metalic compound. The cpu came with compound so I didn't use the
    metalic compound but I'll remove the cpu and place a very small amount of it
    on the heatsink.

    I have a GIGABYTE GA-EP45-DQ6 mobo and mushkin 4GB (2 x 2GB) 240-Pin DDR2
    SDRAM DDR2 800 (PC2 6400) Dual Channel Desktop Memory. I was told that the
    RAM is compatible with the mobo. Regarding timing, I would be very hesitant
    to try something like that and would prefer to take it to a professional.
    Those types of setting changes I've never tried. I'll read up on it though
    and decide if I want to try it myself.

    I haven't done a clean boot on this pc yet so I'm going to try that. I was
    planning on doing a Windows Repair before a format to see if it helps, along
    with uninstalling and reinstalling some of the drivers. If it doesn't help,
    I'll format the drive. I've done it a number of times on other computers.
    The worst part of a format is the amount of time it takes to reinstall
    everything.


    Michael Moi wrote:

    "The generally accepted solution for this problem, when you need to use a
    large number of hard drives, is to stagger the powering-up of the drives.
    There may be a feature in your motherboard's CMOS to configure this.
    Professional shops often use hardware RAID cards, which will usually have
    such a feature. Failing that, getting secondary drive controllers with such
    a feature might be a solution."

    My mobo has the ability to stagger spin-up for a RAID configuration but
    doesn't offer it in an IDE format. I have a controller for the controller
    card but how would I get "secondary drive controllers"? I Googled "secondary
    drive controller" but came up with very little.

    I have 4 external IDE hdds. They're always turned off at boot-up. They
    contain back-up files and I turn them on for less than 5 minutes at a time.
    once every week or two, in order to copy files onto them. My sole purpose in
    buying this case and moving all my externals into it was to eliminate
    externals. I had 23 externals at the time, various sizes and brand drives,
    and with all the cables, hubs, and controller cards that I used for them made
    it one big mess. I found that my internal drives never failed but my
    externals failed at a rate of 4 per year, with 6 failing in the 6 months
    prior to building this new system (the last straw that made me decide to
    build this system with 12 internals).

    Again, thanks for your help, Michael. I very much appreciate it.
     
    Denise, Apr 30, 2009
    #33
  14. Denise

    Paul Shapiro Guest

    Corsair sells very good psu's. No problem there. You mentioned updating the
    motherboard bios and drivers, but didn't specifically say you checked the
    raid controller firmware. That's essentially the bios on the controller
    card, and you would check Highpoint's website for updates. That can be quite
    important, so don't skip that step.

    The fact that you are using 3 psu connectors for your 4 drives doesn't
    necessarily mean that they aren't all on the same 12 volt power bus. From
    Corsair's website: "Dedicated single +12V rail offers maximum compatibility
    with latest components". This means that all the connectors are on the same
    12 volt bus, which is actually a good thing. Their spec says max load on the
    12 volt line is 70 amps, giving 840 watts available. Unless there's
    something wrong with your particular power supply, or the wiring is
    problematic, power should be ok. Corsair's specs say there are 8 sata power
    connectors and 8 4-pin peripheral connectors. Seems like you should have
    enough connectors to give each drive it's own connector. Instead of using
    daisychain power connectors, you might use all 8 available sata power
    connectors and then use 4 of the 4-pin to sata converters.

    In another post you listed these measurements on your system. Either
    something is wrong with the measurements, or with the psu. The listed
    voltages are WAY WAY below spec. The 3.3V line should be between 3.2 and 3.4
    volts, more or less. Your list shows 0.08 volts, at which point the computer
    wouldn't run. Similarly with the 12 volt line varying from 2.24 volt to 7.17
    volt. If someone built the system for you, they should be able to measure
    the 5 volt and 12 volt lines easily, with a voltmeter.
    PSU (the values change ±.1 as the test runs):
    Voltages Value Min Max
    CPU VCORE 1.14V 1.14V 1.28V
    VIN1 1.92V 1.92V 1.92V
    +3.3V 0.08V 0.08V 0.08V
    +5V 5.05V 5.03V 5.05V
    +12V 2.24V 2.18V 7.17V
    -12V -0.45V -0.58V -0.45V
    -5V -0.45V -1.54V -0.45V
    +5V VCCH 3.66V 3.66V 3.66V
    VBAT 3.17V 3.17V 3.17V

    If power is ok, then the raid controller would seem the next likely suspect.
    After verifying you have the latest firmware, you could see if it has some
    boot-time diagnostic/testing capabilities. The 2320 manual says it does
    support staggered spinup, so go into the boot-time configuration for your
    controller and see what you can find. It says staggered startup is off by
    default, but enabling it gives a 2-second delay for each drive. With 12
    drives I would definitely enable the staggered startup if it's available.
    The also mention "Some hard disks may require the use of a software utility,
    or jumper setting to enable this feature", so you may need to check the
    drive specifications. They have a drive compatibility list on their website,
    and it wouldn't hurt to confirm that your drives are on that list.

    Under the Advanced Options the 2310 manual mentions:
    EBDA Reallocation – this function refers to “Extended BIOS Data Areaâ€.
    Disabling this feature may remedy boot problems associated with
    motherboards that halt after the RR231x BIOS screen is displayed.

    You might call Highpoint tech support and ask their advice. They should be
    aware of any issues with your motherboard and/or drives.

     
    Paul Shapiro, Apr 30, 2009
    #34
  15. No, video cards are not considered a controller card. I just covered the
    possibility of multiple add-in HDD controllers to be on the safe side (and
    because with 12 drives, you might have used multiple controllers).
    I would have thought an 850w PSU to be enough, but it does also depend on
    what else you have drawing power - some of the latest video cards and CPUs
    demand a lot of power.
    No, it wasn't. I was talking about the priority allocated to the various
    drive types (that affects ARC Path and numbering), and you're talking
    about the options on how the motherboard SATA controller presents drives
    connected to it to the OS (which, coincidentally, can also affect ARC Path
    and numbering, but for different reasons). I'm not sure why you've chosen
    to configure it to present the drives as IDE though - that is normally
    only needed for older OSs that don't understand SATA (or where SATA
    drivers for the OS are lacking/poor).

    Changing the drive types back to SATA now would probably cause the ARC
    Path to change, and then Windows likely wouldn't boot (updating
    boot.ini/BCD to reflect any changed ARC Path would fix that).
    You're welcome.
     
    Steve Foster [SBS MVP], Apr 30, 2009
    #35
  16. Denise

    Denise Guest

    Hi Paul,

    :
    "Corsair's specs say there are 8 sata power connectors and 8 4-pin
    peripheral connectors. Seems like you should have enough connectors to give
    each drive it's own connector. Instead of using daisychain power connectors,
    you might use all 8 available sata power connectors and then use 4 of the
    4-pin to sata converters."

    I don't remember the psu having 8 4-pin peripheral connectors, or 8
    available for the hdds, the reason I bought the daisy-chain power connectors,
    but I'll look again.

    I downloaded and installed SpeedFan. Its findings for the psu are:

    Vcore1: 1.14V
    Vcore2: 1.92V
    +3.3V: 0.08V
    +5: 5.05V
    +12V: 2.24V
    -12V: -16.31V
    -5V: -7.72V
    +5V: 3.66V
    Vbat: 3.17V

    I don't expect these reading to be any more accurate than the ones taken by
    CPUID. I think I'll need a multimeter to get accurate readings. Some of the
    voltages readings differ between the two tests but it's obvious that the 3.3V
    and 12V readings from both tests are too low (from what you stated).


    I went into the mobo BIOS/CMOS a little while ago. The option to have
    staggered spin-up of drives connected to the mobo would never maintain the
    number of seconds that I entered. After allocating 5 seconds, saving and
    exiting BIOS, I would go back into BIOS and the field would again be blank.
    For some reason, after I allocated 5 seconds a short while ago, saved and
    exited, when I went back into BIOS, 5 seconds was still there. (Same as when
    you take your car to an auto mechanic because it's making a noise but the
    noise isn't there for the mechanic to hear!)


    From what I read in either the mobo or controller card manual, I was under
    the impression that staggered spin-up was only available in RAID, but I'll be
    going into the BIOS for the adapter card to enable staggered startup.


    I checked the BIOS and drivers for my mobo and they're all current, the same
    for the RocketRaid card.


    Paul Shapiro wrote:
    "Under the Advanced Options the 2310 manual mentions:
    EBDA Reallocation – this function refers to “Extended BIOS Data Areaâ€.
    Disabling this feature may remedy boot problems associated with
    motherboards that halt after the RR231x BIOS screen is displayed."

    I'll check this out too.

    Thanks so much :))
    --
    Denise



    --
    Denise



     
    Denise, Apr 30, 2009
    #36
  17. Denise

    Denise Guest

    Hi Steve,

    :

    "I'm not sure why you've chosen to configure it to present the drives as IDE
    though - that is normally only needed for older OSs that don't understand
    SATA (or where SATA drivers for the OS are lacking/poor)."

    I used the IDE format because my existing drives were about 80% to 90% full
    (also the need for additional hdds). If I set up a RAID Array, the info on
    those drives would have been erased. I'm also more comfortable with the IDE
    format, although I would have eventually become comfortable with RAID Arrays.
    It just seemed the best way to go with it.
     
    Denise, Apr 30, 2009
    #37
  18. Denise

    Denise Guest

    Hi Steve and Paul,

    I went into the controller card's BIOS. There was no Advanced Options tab
    and nothing that said EBDA Reallocation or Extended BIOS Data Area. There
    was a Settings tab for staggered start-up but this is the one that I spoke
    about in another post that wouldn't stay enabled. I don't remember where I
    read it but I remember reading that it would only stagger start-up of hdds in
    a RAID array, unfortunately. But at least the drives connected the mobo are
    staggered (5 seconds).

    Until I get it checked out, I'm concerned about the psu. The software tests
    that I ran could be completely wrong but even from a cold boot, I have to
    reset several times before it completely boots to my desktop. It could be
    the timing of the RAM or jumpers on the hdds also . . . other things that
    were mentioned that I need to look into. It could also be that Gigabyte
    boards are flaky and do require resets, as other people have said. So I'll
    be doing a lot of testing, checking and reading, replacing cables, try to
    rewire without the daisychain power connectors, and probably a format.

    One thing that I've been thinking about is the electric current where I
    live. The house has only a 100 amp service and more power is upstairs
    (raised ranch) and my pc is downstairs. I can't have 2 air conditioners on
    downstairs when I vacuum because the circuit breaker flips. This could be a
    factor also.

    The problem is definitely not the cause of old drivers/BIOS/firmware. I
    have all the latest and I checked again and made sure that they were right
    ones.

    It's too bad that the problem wasn't as simple as I first thought . . .
    that the OS wasn't on Disk 0.

    I learned a lot in this discussion, which is really good.

    --
    Denise



     
    Denise, May 1, 2009
    #38
  19. Denise

    westom Guest

    No program reads voltages. The same hardware reads voltages no
    matter what software is used to only display those numbers. That
    hardware must be calibrated with a 3.5 digit multimeter before the
    numbers are useful

    As others noted, your voltages are so bad that even the software
    should not have worked - implying multiple failures. One possible
    reason for so many bad numbers - one bad voltage. Just another reason
    to get a meter (ie Wal-Mart for less than $18). Measure the VDC
    numbers from any one of purple, orange, red, and yellow wires where
    those wires enter the nylon connector to motherboard.

    Those numbers should be taken both when the system boots AND when
    the system is multitasking to all peripherals. Ie. displaying complex
    video graphics, while downloading from the Internet, while playing
    audio loudly, while playing a CD-Rom, while searching the hard drive,
    etc.

    Both the power up and the fully loaded system numbers may also
    report other facts beyond what is obvious. Those who can report more
    will remain silent if those resulting numbers are not provided.

    Nothing in a BIOS has relevance until numbers from the power supply
    system (more than just a power supply) are analyzed. If voltage
    numbers are unknown, then everything else may make no sense. Some
    defective voltages can mean all Speedfan numbers make no sense.

    Also useful are voltages on the green and gray wire both before and
    as the power switch is pressed. In normal operation, the gray wire
    voltage will change slower after the power switch is pressed. What
    all this means (what signals do and what voltages say for you) can be
    explained after numbers are posted.
     
    westom, May 2, 2009
    #39
  20. So, she still has a Hardware problem as I stated before.

    And, as I stated before this is not a group for hardware trouble shooting.

    This is a Windows 64-bit Operating System SOFTWARE group!
     
    Bobby Johnson, May 2, 2009
    #40
    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.