Connect with us

faking computer fan RPM signals

Discussion in 'Electronic Design' started by Joseph, Apr 18, 2005.

Scroll to continue with content
  1. Joseph

    Joseph Guest

    I have a few server PCs and find their fans too noisy. I want to
    remove them and put in some of my own fans that are very quiet.
    However, low and behold the motherboard shuts down when I remove its
    internal fan. This is because the motherboard actually reads an RPM
    signal from the fan and shuts down when it doesn 't see it. I am
    looking for a RPM circuit that can fake this kind of signal. I figure
    its a ON-OFF-ON-OFF type of signal that gets fed into the motherboard.
    Does anyone know of any website or links that has this type of simple
    circuitry.
     
  2. OBones

    OBones Guest

    Well, I'd generate some with a NE555, that's the most simple astable
    oscillator I know of.
     
  3. Jim Thompson

    Jim Thompson Guest

    Before I replaced it completely, my previous machine lost a case fan.
    It was a three-wire RPM output type.

    I found a setting in the BIOS that allowed me to shut off the RPM
    sensing so I could use a two-wire fan.

    ...Jim Thompson
     
  4. colin

    colin Guest

    as another poster said this feature is usualy enabled/diasbled in the bios
    setup, ive replaced several fans in my system with very quiet ones but they
    have fan speed sense signals.

    the suposedly quiet antec psu i have has 2 fans on it, one on outside and
    one on the inside, however there is a seperate set of holes that has no fan
    so i put a third fan on this, this meant when the cpu was using 100% it no
    longer needed to speed up its fans and so remains quiet, handy it has a fan
    conector.

    i also managed to modify one fan ages ago so it had a speed sensor just by
    adding 2 diodes conected to the coil drive.

    you could always add an extra quiet fan with sense signal and put it
    somewhere usefull and use this, or i suspect you would be able to use the
    sense signal of one fan and conected it to more than one input.

    - i gues this is what they call 'fan out' heh

    quiet fans usualy move less air ... i would hope this isnt a server thats
    relied upon 100% of the time.



    Colin =^.^=
     
  5. Joseph

    Joseph Guest

    The computer I am running is a Compaq Proliant 3000 server class
    computer, inside are 2 extremely noisy computer fans + the one in the
    power supply. So 3 noisy fans. Its like you have a vacuum cleaner
    going at maximum speed. So far I have not found the key switch to
    allow me into the bios to check this enable/disable mechanism. I am
    assuming it can't be disable.

    PS: Wow Compaq engineers are so smart, I guess that's why they merge
    with HP. That's right, design machines that once retired from service
    must be landfilled instead of being put to good use by some good
    people.
     
  6. Why not go back to the source. On www.compaq.com you can search for - and
    find - (almost) all the doc you need. The manual part
    http://h20000.www2.hp.com/bc/docs/s...ual/TPM_179764-001_ch4/TPM_179764-001_ch4.pdf
    shows the switches you need to use for changing configuration and wiping the
    CMOS-data. Other chapters of the manual, diagnostic tools and packs are also
    available, for free. If all that fails you can simply ask a question and get
    an answer by e-mail. That's to say I always got one so far.

    petrus bitbyter
     
  7. Guy Macon

    Guy Macon Guest

    On all Compaq machines, you get into the BIOS by hitting F10 when you
    see the blinking square at the top right of the screen during powerup.

    Also, your BIOS is on a hard disk partition, not on the motherboard.
    Get yourself a Compaq SmartStart CD on eBay so you can restore it.
    It doesn't matter what version of smartStart; the newer ones handle
    newer servers but the files for tthe old servers stay the same.

    BTW, you are correct about there being no way to disable the fans.
    The machine will overheat very quickly without them. A fan that
    makes less noise will work, but you really should get a three-wire
    fan with a tach. You do NOT want a compaq server to keep running
    after a fan failure; it will overheat and may never work again.
     
  8. Sell it to someone to use in a computer room and use the money to
    buy a quiet PC, like the DELL I'm typing this on.
     
  9. Guy Macon

    Guy Macon Guest

    Your Dell doesn't have quad Xeon processors, eight hot-plug SCSI
    drives in a RAID array, dual redundant hot-plug power supplies,
    or seamless switchover to a backup system upon hardwatre failure.
     
  10. Ken Taylor

    Ken Taylor Guest

    Excuse me, on the hard disk partition?? How can it be a BIOS if the HD has
    to load it's own driver?

    Ken
     
  11. Iwo Mergler

    Iwo Mergler Guest

    The BIOS is still in ROM/FLASH. The setup program and some sort of
    diagnostic tool lives on a disk partition. Hitting F10 at boot time
    just redirects to the bootsector of the Compaq partition. It loads
    something which looks suspiciously like Windows 3.1.

    Kind regards,

    Iwo
     
  12. Of course, but who needs one of those at a workstation?
    Put it in the other room and use a smaller machine to access it.
     
  13. Guy Macon

    Guy Macon Guest

    The problem is that you don't know the difference between a BIOS and
    a bootloader.

    Yes, it is true that BIOS's made by Pheonix and Award contain
    bootloaders, but that doesn't make them the same thing.

    Compaq systems (the real ones made by the real Compaq - I don't
    know about the ones sins HP ate Compaq) have a BIOS thatis
    seperate from and loaded by the bootloader.
     
  14. Ken Taylor

    Ken Taylor Guest

    Which is what I'd expect, and such an obvious mistake devalues what was
    otherwise sound advice.

    Ken
     
  15. Guy Macon

    Guy Macon Guest

    These kids today... :)

    One could argue that we are simply using different terms for the
    same thing, but in my opinion I am using the correct terminology
    and you are not. The term "BIOS" was defined by the OS that
    invented the first BIOS - CP/M. CP/M had the BIOS on the floppy
    with a bootloader in ROM. Compaq followed this pattern, IBM did
    not (they put the BIOS in ROM), and most manufactures followed
    IBMs lead.

    The idea of a bootloader existed long before either CP/M or
    it's BIOS were invented. It was originally toggled in to
    core memory using front panel switches, and it gave the ability
    to fire up the paper tape reader and to start loading the data
    on the paper tape into core. (RAM came along and replaced
    core later).
     
  16. Guy Macon

    Guy Macon Guest

    You should be careful about calling correct information
    "an obvious mistake."

    I am using the correct terminology and you are making a mistake.
    Yours is a common error among those who have never used a computer
    other than PC clone.

    The program that runs first is called a bootloader, not a BIOS.
    See [ http://www.wikipedia.org/wiki/Booting ]. The confusion
    came when IBM combined a bootloader and part of a BIOS and
    put them in the ROM of the original IBM PC and called it a
    "BIOS." (The entire BIOS didn't fit, so part of it was put
    in the file io.sys).

    Bootloaders existed long before either CP/M and it's BIOS were
    invented. The term comes from the term "bootstrapping" which
    goes back to the pre-computer phrase "pulling yourself up by your
    bootstraps." See [ http://www.wikipedia.org/wiki/Bootstrapping ].

    A bootloader can be in Flash, ROM, RAM, core, or even hardware.

    The first computer booted from a hardware plugboard.
    See [ http://www.columbia.edu/acis/history/eniac.html ].
    It had no BIOS because BIOSs (and operating systems!) had not
    been invented yet.

    The early DEC PDP series booted from a bootloader in core or RAM
    that was loaded one location at a time from the front panel.
    See [ http://www.psych.usyd.edu.au/pdp-11/bootstraps.html ].
    By this time operating systems existed, but BIOSs were yet to be
    invented.

    The DEC first stage bootloader was originally toggled in to core
    or RAM memory using front panel switches, and it gave the ability
    to fire up the paper tape reader and to start loading the second
    stage bootloader the paper tape, mag tape, or disk. Going directly
    to tape/disk involved a cumbersome process of setting up a boot
    block, so most users used paper tape for the second stage bootloader.
    (There was also a hardware bootloader that was a bunch of discrete
    diodes on a PCB - essentially a ROM - but it was rare.

    The term "BIOS" is defined by the OS that invented the first
    BIOS - CP/M. CP/M had the BIOS on the floppy with a bootloader
    in ROM. Compaq followed this pattern, IBM did not (they put the
    BIOS in ROM with the bootloader), and most manufacturers followed
    IBM's lead.

    Apple correctly calls their bootloader in ROM "Firmware" instead
    of the incorrect "BIOS." (The correct term for a PC would be
    "bootloader and BIOS", but that horse is out of the barn already).
    See [ http://www.kernelthread.com/mac/osx/arch_boot.html ]

    The next time you think someone is wrong, I suggest a gentle
    "I thought that X was true, was I mistaken?". If it turns out
    that a mistake was made you can always escalate to flaming
    whoever made the mistake, but if you start out with an insulting
    tone it becomes quite embarassing for you when it turns out that
    it was you who made the error.
     
  17. Iwo Mergler

    Iwo Mergler Guest

    I may be wrong, but AFAIK the BIOS is the "Basic I/O System"
    and traditionally offers some rudimentary drivers (floppy,
    hdd, keyboard, etc.) which are accessible via software interrupts.
    The first PCs even had a BASIC interpreter in int15.

    The bootloader makes use of int13 disk services to get at the
    bootsector. In DOS, the filesystem is based on int13 functions.

    Nobody uses the (real-mode-only) disk services any more, but
    other functions have protected mode support. The Linux ACPI
    driver, for instance, is just an interface to a BIOS driver.

    I understand a bootloader to be a piece of code which sets up
    the memory timing, initialises a few devices and loads the OS.
    In particular, a bootloader isn't used after the OS starts.
    The PC BIOS has all kinds of useful functionality which is
    still used by most OSes at runtime.

    The Compaq setup utility is not a BIOS. No code from the
    Compaq partition ever runs unless you hit F10 at boot time.
    I would clasify it as an application instead, which allows
    you to set a few flags in the NVRAM for the Bootloader/BIOS
    to use on the next boot.

    Kind rgeards,

    Iwo
     
  18. Ken Taylor

    Ken Taylor Guest

    Sorry, I meant to be a bit less up myself when I started, with an "I could
    be wrong", but I did a couple of re-writes before firing it out and it lost
    it's saving graces.

    So, humble pie in mouth, (by the way, nice history!) are "all" DEC servers
    like that? We run a few and I didn't think that was the case. Mind you, on
    reflection, if only one drive goes then the redundancy (at least should)
    take care of that. I don't think we've had to even go to the install CD for
    recovery.

    I have vague recollections of salivating while 'helping' out the computer
    tech back in college and finally being allowed to fire up the PDP-8. Ah, the
    toggle switches. To be able to do that was to be a true geek. Which didn't
    matter as there were only guys on campus anyway!

    Would you consider that the definition of 'BIOS' has changed over the years
    to refer to the firmware that the CPU uses to boot up from, rather than the
    boot sector code itself?

    Again, apologies for the clumsy post.

    Cheers.

    Ken
     
  19. Guy Macon

    Guy Macon Guest

    No problem; I get that way myself at times.
    I doubt it, but my experience stops with the PDP-8. As soon as smaller
    machines were available for embedded systems, I made the switch.
    I can still toggle in the first-stage bootloader from memory. We used
    to race each other.
    Common usage has certainly changed, but there are still many folks
    like me who use CP/M on a daily basis, and tend to use the original
    definition. Not always, though; I myself used the common usage when
    I called the Compaq Diagnostics and Setup Partition a "BIOS"; the
    part that does the setting up is part of "BIOS" as redefined by IBM,
    Pheonix, and Award, but in CP/M it would be considered a utility.
    Then there is Linux, where the usual way of changing a configuration
    is to edit a textfile rather than running a program - but x86 Linux
    runs on machines that have a IBM/Pheonix/Etc. "BIOS". Interestingly,
    good old DOS uses the BIOS to run but Linux/BSD/OS X and Win NT/2K/XP
    only use it as a bootloader.
     
  20. Spajky

    Spajky Guest

    lower Fan speeds by inserting some diodes serially to + (red supply
    line) or replace them with 3-wire silent ones; you can find certain
    schematics under electronics on my site ... (there´s also a circuit
    for 2-wire fans to show RPMs ...
     
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.
Electronics Point Logo
Continue to site
Quote of the day

-