Total Fan Noise - how to calculate ??

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

For example, if I have the following:

5 case fans - 28 dBA
1 CPU fan - 15 dBA
1 PSU fan - 23 dBA
1 Video Card fan - 12 dBA
------------------------------------------
Total = 190 dBA

Now, I am certain that you cannot calculate the amount of noise you
will hear by adding all the fan's dBA rating together. I am also sure
it would be an extremely complex formuyla to calculate, because there
are so many variables (placement of fans, distance from user's ears,
vibration, etc).

What I am wondering though, is if there is a very general ballpark way
to estimate how much noise a system will create. TIA
12 answers Last reply
More about total noise calculate
  1. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    Al Dykes wrote:
    > In article <l68f61tf70ec0l84iu99efir0kc82tp9bc@4ax.com>, Ken Roberts
    <> wrote:
    > >For example, if I have the following:
    > >
    > >5 case fans - 28 dBA
    > >1 CPU fan - 15 dBA
    > >1 PSU fan - 23 dBA
    > >1 Video Card fan - 12 dBA

    I wouldn't even consider it. db's doesn't even indicate how the sound
    will be percieved either. High pitch noises may not sound as 'noisy'
    as low pitch ones which are just as noisy. There are also loads of
    other things to take into consideration, such as how much sound will be
    absorbed before it leaves the case etc.

    Your dBA for some of those fans seems very low. CPU fan 15dBA, and
    video card 12dBA.
  2. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    In article <l68f61tf70ec0l84iu99efir0kc82tp9bc@4ax.com>, Ken Roberts <> wrote:
    >For example, if I have the following:
    >
    >5 case fans - 28 dBA
    >1 CPU fan - 15 dBA
    >1 PSU fan - 23 dBA
    >1 Video Card fan - 12 dBA
    >------------------------------------------
    >Total = 190 dBA
    >
    >Now, I am certain that you cannot calculate the amount of noise you
    >will hear by adding all the fan's dBA rating together. I am also sure
    >it would be an extremely complex formuyla to calculate, because there
    >are so many variables (placement of fans, distance from user's ears,
    >vibration, etc).
    >
    >What I am wondering though, is if there is a very general ballpark way
    >to estimate how much noise a system will create. TIA


    Google for "how to add dba" came up with this...

    http://www.overclockers.com/articles607/

    To simplify, if you have two 28dBA fans, combined they are 31dBA. Do
    that again (2x2 fans) and you're up to 34dBA, so your 5 case fans are
    about 35dBA. The rest of the fans, combined bring you up to maybe
    37dBA.

    It'll generate high numbers (ie lower perceived sound) because, for
    example, given identical fans, one moving air into the case and one
    exhaustiong air out, the exhaust fan will be a bit louder.

    Nobody needs 5 case fans for anything like a standard desktop PC. It
    the're not set up right they are fighting each other. One on and one
    out should do it.
    --
    a d y k e s @ p a n i x . c o m

    Don't blame me. I voted for Gore.
  3. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    Ken Roberts wrote:
    > For example, if I have the following:
    >
    > 5 case fans - 28 dBA
    > 1 CPU fan - 15 dBA
    > 1 PSU fan - 23 dBA
    > 1 Video Card fan - 12 dBA
    > ------------------------------------------
    > Total = 190 dBA
    >
    > Now, I am certain that you cannot calculate the amount of noise you
    > will hear by adding all the fan's dBA rating together. I am also sure
    > it would be an extremely complex formuyla to calculate, because there
    > are so many variables (placement of fans, distance from user's ears,
    > vibration, etc).

    Yes, and more. In particular, humans find higher pitched sounds more
    objectionable than lower tones so even if the 'noise level' is the same the
    'objection level' is different. (which is not the same thing as dbA
    weighting. dbA weighting takes into account that the human ear is more
    sensitive to certain sound frequencies than others but not that they're
    emotionally 'objectionable')

    A 92mm fan putting out a measured 26 db of 'noise' will 'sound better' than
    a 60mm fan with the exact same measurement because the 92mm job does it at
    a lower RPM, which translates (a significant amount, anyway) to a lower
    frequency.


    > What I am wondering though, is if there is a very general ballpark way
    > to estimate how much noise a system will create. TIA

    Sorta, but not really.

    At any rate, the basic decibel (actual base unit being the Bel and deci the
    base ten prefix for .1 ) equation is logarithmic and is measuring one level
    against a reference point (ratio), which is essentially arbitrary (but
    standardized for particular db scales).

    http://www.analogrules.com/dbwatts.html

    For sound power levels the equation is

    dB = 10 * log 10 (power in W/m 2 ) + 120, to get the dB(A) sound level

    (the equation presumes the standard reference level and is why you do not
    see a ratio in the power section but a + 120 to the log)

    To add db you need to convert back to power level and add the power levels
    together, then reconvert to db.

    There is a shortcut and that's when the db levels are the same. I.E. two is
    a doubling, 4 is a doubling of the doubling, etc. And a doubling of the
    power adds 3db. So two 21 db fans would add to 24 db, four 21 db fans would
    add to 27 db, 8 to 30, ... etc.

    However, what may seem counter intuitive is that when the db levels are
    greatly disparate, as like one is 28 and the other is 12, then the increase
    in db over the larger one, I.E. 28 in this example, is rather minuscule
    because the louder one is *so* much louder that the quieter one adds little
    to the scale which, if you remember, is logarithmic.

    I.E. it looks sorta like

    s .. to equal a small db change up here.
    o ..
    u ..
    n ..
    d ..
    ..
    p ..
    o ..
    w ..
    e ..
    r ..
    db -------------------------------->

    takes a lot down here


    So, the point is, in your case you can significantly discount the 12 and 15
    db jobs since the 28 and 23 db fans will dominate.

    Now, it gets even more confusing because the human ear does not interpret a
    doubling of sound power as a doubling of 'the sound'. I.E. even though a 26
    db fan is putting out twice as much noise power as a 23 db fan humans do
    not perceive that as 'twice as loud'. Rather, blind studies show that
    humans tend to indicate something is 'twice as loud' when the db level
    changes by 10. (this is a source of great confusion with people arguing
    about what a 'doubling of sound' is and it depends on what one is looking
    at. I.E. if you're looking at noise levels to determine ear damage and
    hearing loss, like standing next to a 747 jet engine, then the actual power
    level is the item of interest and 3 db is a doubling of it. If your talking
    about how one perceives noise, or the local opera singer, then 10 db is a
    doubling of the *perceived* sound)

    To make matters even worse, db levels are based on coherent sound and
    'noise' is non coherent, As such it tends to add together less (non
    coherent cancellation). So what would be a 3 db increase for coherent sound
    adds to less, say between 1.5 and 2.5 db, depending on the nature of the noise.

    Which is why I said "Sorta, but not really."

    And all that is before we even get to case dampening, how far away one is,
    direction, conducted vibrations and secondary radiation from things (such
    as case walls, etc.) and the fact that the fan rating is it sitting all by
    its little lonesome in an acoustic chamber and not whistling air through a
    stamped case grill or fan blade tone beating it across heatsink fins.
  4. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    petermcmillan_uk@yahoo.com wrote:

    > Al Dykes wrote:
    >
    >>In article <l68f61tf70ec0l84iu99efir0kc82tp9bc@4ax.com>, Ken Roberts
    >
    > <> wrote:
    >
    >>>For example, if I have the following:
    >>>
    >>>5 case fans - 28 dBA
    >>>1 CPU fan - 15 dBA
    >>>1 PSU fan - 23 dBA
    >>>1 Video Card fan - 12 dBA
    >
    >
    > I wouldn't even consider it. db's doesn't even indicate how the sound
    > will be percieved either. High pitch noises may not sound as 'noisy'
    > as low pitch ones which are just as noisy.

    You have the basic idea right but the scale backwards. High pitched sounds
    seem 'louder', and more objectionable.

    Which, btw, is why people perceive that cheap stereos can put out a 'lot of
    sound'. They can't really handle the power needed for 'loud' and, so, clip
    the hell out of the audio, which generates all kinds of harmonics up the
    frequency scale that the human ear interprets as 'loud', and 'painfully so'
    (so it must be *really* 'loud'). It's the high frequency distortions that
    give the perceived effect.

    The reverse is true for bass. It takes a *lot* of power to get 'volume',
    which is why people can comfortably (that's a joke folks) listen to bass
    that's shaking the house off it's foundation.

    Sensitivity also changes with the volume, which is what the 'loudness'
    control on stereos is for: to boost them when the volume level is turned down.

    The frequencies we hear the easiest are between 3 and 4 KHz.

    When you get above 9KHz, or so, human hearing drops off again, like on the
    bass end, but the 'objectionable' nature of it remains.

    > There are also loads of
    > other things to take into consideration, such as how much sound will be
    > absorbed before it leaves the case etc.
    >
    > Your dBA for some of those fans seems very low. CPU fan 15dBA, and
    > video card 12dBA.
    >
  5. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    > Total = 190 dBA

    Ah, noooooo :-)

    For 2 fans of A1 dB(A) + A2 dB(A)...
    o Combined = 10 * LOG (10 ^ (A1 / 10) + 10 ^ (A2 / 10))
    So for 2 fans of 21dB(A) + 21dB(A)...
    o Combined = 10 * LOG (10 ^ (21 / 10) + 10 ^ (21 / 10)) = 24dB(A)

    For n fans just keep adding onto the end...
    o Combined = 10 * LOG (10 ^ (A1 / 10) + 10 ^ (A2 / 10) + 10 ^ (A2 / 10))

    Thus for 5 fans of 21dB(A) the combined total is 28dB(A).
    o Multiple exhaust fans are additive in cfm
    ---- eg, 5x exhaust fans of 25cfm each = 125cfm
    ---- subject to case resistance & c/sectional area of the case intake
    o Adding intake fans is not additive in cfm, but assistive in pressure
    ---- so allowing exhaust fans to get closer to their Free Air airflow

    If you end up at 90dB(A), worry as that's the limit for machine rooms :-)
    --
    Dorothy Bradbury
    www.dorothybradbury.co.uk for quiet NMB & Panaflo fans
  6. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    Matt wrote:

    > David Maynard wrote:
    >
    >> dB = 10 * log 10 (power in W/m 2 ) + 120, to get the dB(A) sound level
    >
    >
    > W/(m^2) is a unit of intensity, not power.

    Quite right.

    The audio industry is lax about that terminology because the m^2 term
    cancels in the ratio.
  7. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    > Sensitivity also changes with the volume, which is what the 'loudness'
    > control on stereos is for: to boost them when the volume level is turned
    > down.

    To boost what? Does the loudness control do more than increase/decrease
    the volume?
  8. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    >> Sensitivity also changes with the volume, which is what the 'loudness' control on stereos is for: to boost them when the volume level is turned down.
    >
    >
    > To boost what? Does the loudness control do more than increase/decrease the volume?

    It exaggerates the base and treble, fooling your ear into thinking
    that you have the volume higher than it actual is.
  9. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    Matt wrote:

    > sorry, having a little trouble with my news server ...

    Are you with HOTMAIL? They had a major burp at 1421 ET.
  10. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    Matt wrote:

    > David Maynard wrote:
    >
    >> Sensitivity also changes with the volume, which is what the 'loudness'
    >> control on stereos is for: to boost them when the volume level is
    >> turned down.
    >
    >
    > To boost what? Does the loudness control do more than increase/decrease
    > the volume?
    >

    Yes, the 'loudness' control boosts the low and high frequencies, at low
    volume, because the ear's sensitivity to those frequencies drops off,
    relative to the middle frequencies, as the volume is turned down.
  11. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    Matt <matt@themattfella.zzzz.com> wrote:

    (for the third time)

    > To boost what? Does the loudness control do more than
    > increase/decrease the volume?

    Besides double and triple replies to the same post, your posts
    have been showing up in duplicate and triplicate for weeks. Once
    is plenty.

    Message-ID: <86C1e.1950$_g7.44@news02.roc.ny>
    Message-ID: <WCB1e.2088$hg.1479@news01.roc.ny>

    Message-ID: <DWS9e.4229$hg.1557@news01.roc.ny>
    Message-ID: <NPS9e.4226$hg.4139@news01.roc.ny>
    Message-ID: <hRS9e.4227$hg.595@news01.roc.ny>

    Message-ID: <dnIbe.281$Xa5.159@news02.roc.ny>
    Message-ID: <bIkce.89$8g.82@news01.roc.ny>

    Message-ID: <Pm2ce.20$Gu7.4@news02.roc.ny>
    Message-ID: <3v2ce.21$bJ7.7@news02.roc.ny>
    Message-ID: <eQ2ce.19$8g.11@news01.roc.ny>

    Message-ID: <xSFZd.608$Va5.74@news02.roc.ny>
    Message-ID: <OTa_d.774$Ls6.106@news02.roc.ny>


    >
    >
    >
    > Path: newsdbm06.news.prodigy.com!newsdst02.news.prodigy.com!newsmst01a.news.prodigy.com!prodigy.com!newscon06.news.prodigy.com!prodigy.net!news-feed01.roc.ny.frontiernet.net!nntp.frontiernet.net!news01.roc.ny.POSTED!53ab2750!not-for-mail
    > From: Matt <matt themattfella.zzzz.com>
    > User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.6) Gecko/20050328 Fedora/1.7.6-1.2.5
    > X-Accept-Language: en-us, en
    > MIME-Version: 1.0
    > Newsgroups: alt.comp.hardware.pc-homebuilt
    > Subject: Re: Total Fan Noise - how to calculate ??
    > References: <l68f61tf70ec0l84iu99efir0kc82tp9bc@4ax.com> <d488a8$3cq$1@panix5.panix.com> <1114094811.025888.149470@g14g2000cwa.googlegroups.com> <116fgrjbcsua213@corp.supernews.com>
    > In-Reply-To: <116fgrjbcsua213@corp.supernews.com>
    > Content-Type: text/plain; charset=us-ascii; format=flowed
    > Content-Transfer-Encoding: 7bit
    > Lines: 9
    > Message-ID: <hRS9e.4227$hg.595 news01.roc.ny>
    > X-Complaints-To: abuse-news@frontiernet.net
    > X-Trace: 52616e646f6d495629c9c8003ab14adb58e5ca2d64b74b90a5aedf1361f9daa68e1495b4e9b7bf6292986a3da989a706b9f781c8fa086495621fe06c16cc160ae0598a9263f6c2e872bedd6722bca60428372752bf238664692df204d3da587c
    > X-Abuse-Info: Please be sure to forward ALL headers so that we may process your complaint properly.
    > NNTP-Posting-Date: Thu, 21 Apr 2005 19:06:53 UTC
    > Date: Fri, 22 Apr 2005 18:21:57 GMT
    > Xref: newsmst01a.news.prodigy.com alt.comp.hardware.pc-homebuilt:433972
    >
  12. Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

    John Doe wrote:
    > Matt <matt@themattfella.zzzz.com> wrote:
    >
    > (for the third time)
    >
    >> To boost what? Does the loudness control do more than
    >> increase/decrease the volume?
    >
    > Besides double and triple replies to the same post, your posts
    > have been showing up in duplicate and triplicate for weeks. Once
    > is plenty.
    >

    God you are a bitchy little thing
Ask a new question

Read More

Homebuilt Fan Systems