how much is too much

Discussion in 'Asus' started by RzR, Dec 19, 2005.

  1. RzR

    RzR Guest

    hi...i have this non-standard mounting procedure for thermaltake big
    water SE water have to tighten the nuts to get the
    waterblock tight do i have to go? is there a rule of
    thumb...i wand the waterblock to have a good contact, but dont want to
    damage the cpu...tnx
    RzR, Dec 19, 2005
    1. Advertisements

  2. RzR

    Paul Guest

    I'll assume this is an Intel LGA775 processor. You can go to and in the processor section (technical documents),
    you can get a datasheet for your processor. For example, the
    600 series processors (the ones with 2MB cache - doc 306382), the
    datasheet says the "normal force" may range from 18 lbf to 70 lbf.
    There is a higher number for dynamic force, which I'm guessing has
    something to do with a shock specification (if you bump or drop
    the computer). "Normal force" is the force pressing down on the
    heat spreader, and the force is assumed to be uniform across the
    surface of the spreader.

    The amount of force you apply, actually depends on the thermal
    interface material used. If you use thermal grease, the amount of
    force pressing down on the IHS can be lower, than if using other
    materials. Grease does not need quite as much force as other

    Part of the downward force is intended to prevent the heatsink
    from separating from the CPU, if the computer is bumped or dropped.
    So, you want enough force to prevent the water block from moving
    out of place, plus just a little more to give it some shock
    resistance. The water block probably has less mass than your
    average high performance air-cooled heatsink, so again, that
    means you don't need quite as much shock resistance as you would
    need with a more massive heatsink.

    If you are using the "H-plate" to hold the water block in place,
    and there are four screws, then 70lbf divided by 4, gives a
    _maximum_ of 17.5lbf for each screw. The H-plate will probably bend
    a bit, and the legs of the plate will function as a spring
    (f = k * delta_x, but we don't know the value of k for the plate).
    All I can suggest, is imagine what 17.5 pounds of force would be
    like on each screw, and make sure you use less than that amount
    of force :)

    Also, make sure you attach your hoses to the block, before fitting
    the block to the processor. You don't want to be cranking on
    that assembly, except to do up the nuts on the H place.

    Paul, Dec 19, 2005
    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.