Home All Groups Group Topic Archive Search About

How to Keep a Noisy LinkSys Gigabit Switch Quiet?

Author
23 Feb 2005 6:51 PM
jaykchan
I need to find a way to keep a LinkSys Gigabit switch (EG008W) quiet.
>From day one, it makes a relatively loud noise when I power it up; I
can hear the noise even I have closed the door of the equipment closet.
This sounds like noise from a noisy fan. If this is the case, I may be
able to replace the noisy fan with a quiet one. The problem is that I
don't see how it can have a fan inside its low-profile chasis.

No, I cannot get a refund because I left it in the box over 15 days
before I powered it up for the first time, and I bought it from CompUSA
who has that short refund period.

I have already sent it back twice and get two different replacement
units (of the same model). And all three of them are noisy in exactly
the same way. Seem like this is by design, and is not a random
manufacturing defect. I cannot afford to send it back because the
shipping cost is becoming more expensive than the switch itself.

My questions are:
-  Where is the noise coming from?
-  How can I keep the noise down?
-  How can I open it and look inside without
   breaking it apart? The case is really hard
   to open, and there is no screw.
-  Does anyone try this before?

Thanks.

Jay Chan

Author
24 Feb 2005 11:43 AM
kony
On 23 Feb 2005 10:51:42 -0800, jaykc***@hotmail.com wrote:

>I need to find a way to keep a LinkSys Gigabit switch (EG008W) quiet.
>>From day one, it makes a relatively loud noise when I power it up; I
>can hear the noise even I have closed the door of the equipment closet.
>This sounds like noise from a noisy fan. If this is the case, I may be
>able to replace the noisy fan with a quiet one. The problem is that I
>don't see how it can have a fan inside its low-profile chasis.

How tall does a fan need be?
I'd "guess" they just plopped a heatsink w/fan on the
processing chip.  Sure, you could do something-or-other
about that, and since your warranty is up you have nothing
to keep you from opening it, so why haven't you yet?

>
>No, I cannot get a refund because I left it in the box over 15 days
>before I powered it up for the first time, and I bought it from CompUSA
>who has that short refund period.

See?  I forcast the future and do cute parlor tricks too!
;-)

Show quoteHide quote
>
>I have already sent it back twice and get two different replacement
>units (of the same model). And all three of them are noisy in exactly
>the same way. Seem like this is by design, and is not a random
>manufacturing defect. I cannot afford to send it back because the
>shipping cost is becoming more expensive than the switch itself.
>
>My questions are:
>-  Where is the noise coming from?
>-  How can I keep the noise down?
>-  How can I open it and look inside without
>   breaking it apart? The case is really hard
>   to open, and there is no screw.
>-  Does anyone try this before?

Did you look for screws behind the label(s) and under any
feet?  If you're SURE there are no screws, what's left is to
take a knife in an inconspicuous area (after gently heating
the plastic so it's a little more flexible, less likely to
crack) and slide knife inbetween the crack somewhere,
determining how the (top and bottom?) seat together,
probably with a tongue and groove or half-groove lip so the
outer portion of the lip, shell, should be flexed outward a
little while you're peering along the seam with a strong
light, looking for little plastic friction-tabs.  You'd just
sorta work your way around the seam unhooking each set of
tabs, and having some strips of cardboard (or whatever,
thin) material handy to shove in to keep the case seam open
as you proceed the rest of the way around it.  At least, IMO
that's the best way to do it if there aren't any screws.

So you have it open finally and there it is, the fan.
Having no idea how well ventilated the chassis is I can't
begin to guess how much slower the fan can be, whether a
really low RPM fan will be sufficient or perhaps just
throttling back the current fan as much as tolerable.  Tiny
fans are also harder to come by, they're out there but if
you want to pick and choose exactly what replacment
make/model/speed/noise you get, it may end up premium
priced.  Another alternative would be to check the voltage
with a multimeter and calculate a resistor value to place
inline on the power lead...  Somewhere in the ballpark of
10-200 Ohms 2W would be my first guess as to a suitable
value(s) to try if you dont' have alternate methods of
determining the voltage/noise relationship of a particular
fan.
Author
24 Feb 2005 5:51 PM
jaykchan
OK, I will attempt to open it.  If I find that I may be about to break
it into pieces, I will try another approach, such as building a box to
house it.

Thanks.

Jay Chan
Author
24 Feb 2005 9:31 PM
kony
On 24 Feb 2005 09:51:53 -0800, jaykc***@hotmail.com wrote:

>OK, I will attempt to open it.  If I find that I may be about to break
>it into pieces, I will try another approach, such as building a box to
>house it.
>
>Thanks.
>
>Jay Chan

Usually one finds they're about to break it right after it
breaks, though if it's the plastic friction tabs I mentioned
then the common "break" would be a tab or two coming off.
Whether a tab coming off matters could depend a lot on how
much it needs to do unusual things like withstand abuse or
stand up sideway on a pedistal-thingy, as some of them are
designed to sit vertically as an option.
Author
25 Feb 2005 1:07 PM
jaykchan
That kind of break will be OK with me.  Worse comes to worse, I can use
a velco strip to hold it together.  I am more afraid of breaking
anything inside that will make it stop working.  I will see how it go
in this weekend or the next weekend.

Jay Chan
Author
25 Feb 2005 8:46 PM
kony
On 25 Feb 2005 05:07:23 -0800, jaykc***@hotmail.com wrote:

>That kind of break will be OK with me.  Worse comes to worse, I can use
>a velco strip to hold it together.  I am more afraid of breaking
>anything inside that will make it stop working.  I will see how it go
>in this weekend or the next weekend.
>
>Jay Chan

They're "usually" built with all parts surface-mounted on a
single PCB, if any (tools) stay near the edge of the casing
at most you should have a bent or cracked casing... that is,
IF there are no screws that they cleverly concealed.
Author
28 Feb 2005 4:23 PM
Ric
Just thought I would add my two cents on this. I have run into the
same problem. Well except for the fact that with the Linksys I am
working with fan has gone bad. It rattles and you can hear it hit the
housing sometimes. At any rate what I have found, though I am unable
to open the housing so far is that there are four prongs that hold
the case to the board inside. To top it off under the rubber feet in
the back are two clips as well. Now this do not pop out they slide to
an opening. One of the problems is the way it needs to slide the pegs
that hold it to the board will not allow, and to top it off the front
housing blocks it from sliding. I even called Linksys about it and
they would not share how to get the dumb thing open. I think I will
check into Cisco and see what they have like it. Our company deals
with Cisco so I have no idea why this one department went with
Linksys. Well really I do know why. They did not bother checking with
IT. Well anyway I hope this gives someone an idea and that they share
it with the rest of us.
Author
1 Mar 2005 1:49 PM
jaykchan
Thanks.  I will need to print it out and read it a couple times with
the LinkSys Switch in front of me.

Seem like you are saying that it really has a fan inside.  This sounds
promising.  May be the noise will go away if I replace it with a quiet
one.

Jay Chan
Author
2 Mar 2005 1:42 AM
SRay69
Just got the Linksys EG008W myself.  I have 3 other LinkSys brand
switches and routers and none of them have fans in them!  If I
probably wouldn't have bought it if I knew how loud this thing is.  I
have a machine with ten 120mm fans in it and this LinkSys is OVER
TWICE as loud!

There are no screws on it that I can see, unless they're under the
rubber feet.

There *are* these little holes that look like some kind of access for
a proprietary security tool.  There's also a security sticker near
these holes warning that tampering or opening the unit will void the
warranty.

I really don't care about voiding warranty, I just wanna replace that
POS fan!

Anyone know how to open it nicely?
Author
2 Mar 2005 12:16 PM
SRay69
Ok, figured it out and replaced the fan tonight.  Luckily I had just
the perfect fan lying around.  Man, LinkSys engineering has
definitely gone down hill with the fan and this switch.  I see anger
all over the place about their cooling solution.

My solution makes the fit a good bit tighter, but it has so far worked
great the past couple hours and seems to give decent airflow.

I don't have time to make a tutorial and I shouldn't even be taking
time to make pictures, but I know I'd be grateful if someone helped
me out.

Check it out:
http://members.dslextreme.com/users/jperry69/LinkSysGigabitFanFix.jpg

By the way, that tiny fan was OVER TWICE as loud as my new 14 fan
rig:
http://members.dslextreme.com/users/jperry69/PimpRig.jpg


If you have any questions you can reach me at jim_perry at
hotmail.com.  Although I don't know where to get right size fan, the
size I used is in the pic!

Disclaimer:
THIS WILL VOID YOUR WARRANTY, SO I TAKE NO RESPONSIBILITY FOR ANY
DAMAGE IT MAY CAUSE!  :)
Author
2 Mar 2005 1:33 PM
jaykchan
YOU ARE GREAT!!

I am really glad that you have found a solution to this noisy switch
problem.

> Ok, figured it out and replaced the fan tonight.  Luckily I had just
> the perfect fan lying around.  Man, LinkSys engineering has
> definitely gone down hill with the fan and this switch.  I see anger
> all over the place about their cooling solution.

You are exactly right. I have returned the noisy switch twice to
LinkSys, and each time they sent a noisy one back. The last time their
tech support promised to personally examine the replacement unit before
sending it to me. Yes, he indeed opened the box and I assume that he
had examined it. But the replacement unit is just as noisy as the
returned one. Something is really wrong: either the product
specification is wrong (this means all the same model of their GigaBit
Switch are noisy), or their tech support really didn't care and just
went through the motion. Currently, I am under the impression that all
from the same model are noisy because all units that I have received
are noisy in the exact same way.

> My solution makes the fit a good bit tighter, but it has so far
worked
> great the past couple hours and seems to give decent airflow.
>
> I don't have time to make a tutorial and I shouldn't even be taking
> time to make pictures, but I know I'd be grateful if someone helped
> me out.
>
> Check it out:
> http://members.dslextreme.com/users/jperry69/LinkSysGigabitFanFix.jpg

I am really glad that you have taken the time to take pictures and show
the dimension of the new fan (original fan 40x10mm, new fan 60x15mm low
speed fan). Now, I am sure that the fan is the source of the noise
(instead of some other hard-to-replace electrical components); I can go
about replacing it.

I assume you need to oversize the fan in order to maintain the same air
flow while running it in low speed to reduce noise. This makes sense.

One question though:
   How do you open the case? What was the trick to open it? Do you
simply slide the black plastic cover in the front forward? Do you need
to press down any specific area to get the cover off?

> By the way, that tiny fan was OVER TWICE as loud as my new 14 fan
> rig:
> http://members.dslextreme.com/users/jperry69/PimpRig.jpg

Your PC is really nice -- and BIG. It must be your game machine that
you need to overclock it to get the last bit of performance out of it
-- that will explain the reason why you need 14 fans.

> Disclaimer:
> THIS WILL VOID YOUR WARRANTY, SO I TAKE NO RESPONSIBILITY FOR ANY
> DAMAGE IT MAY CAUSE!  :)

I know this. Anyway, I had already broken the warranty seal when I
attempted (and failed) to open the case. This warning may be useful for
other people who may attempt to do the same thing.

Jay Chan
Author
4 Mar 2005 10:33 PM
SRay69
> YOU ARE GREAT!!

Aww, shucks.  :roll:

> Currently, I am under the impression that all from the same model
are noisy because all units that I have received are noisy in the
exact same way.

Yeah, I'm sure most of them are noisy.  Take a look at Amazon user
reviews on this model and most complain loudly about it (no pun
intended).

> I assume you need to oversize the fan in order to maintain the same
air flow while running it in low speed to reduce noise. This makes
sense.

A larger fan is the right way to go, but even the fan I have in it is
blocked some by the front cover, the blue one without holes.  There's
good airflow and I can feel it when I bring it to my face.

> One question though:
> How do you open the case? What was the trick to open it? Do you
simply slide the black plastic cover in the front forward? Do you
need to press down any specific area to get the cover off?

I guess I wasn't very clear in the pic.  The blue front piece it what
comes straight off.  You'll want to pull it off towards you if you
have the unit's front facing you.  One side seems to come off better
than the other, so you'll need to find out which side works for you.
There's no real trick, you just basically pull like heck until it
comes loose.  It snaps back together with no problem, but like you
say, it will void the warranty of course.  LinkSys makes sure of this
with it's little security sticker on the bottom of the unit.

If you need a picture to see how it comes off, let me know.

> Your PC is really nice -- and BIG. It must be your game machine that
you need to overclock it to get the last bit of performance out of it
-- that will explain the reason why you need 14 fans.

Actually I haven't OC'd it at all yet, mainly because I haven't had
time.  I put it together to replace an old LOUD 5 year old PC.  It's
mostly a development machine right now, again no time to play games
on it, YET!  :)  I wanted a quiet case with good airflow and this was
the best one out there.  It can be bought from www.MountainMods.com if
anyone's interested.  You can't beat powder-coated aluminum.
Definitely worth the price.  Those are all huge 120mm fans too.
Author
6 Mar 2005 3:27 AM
jaykchan
> I guess I wasn't very clear in the pic.  The blue front piece it what
> comes straight off.  You'll want to pull it off towards you if you
> have the unit's front facing you.  One side seems to come off better
> than the other, so you'll need to find out which side works for you.
> There's no real trick, you just basically pull like heck until it
> comes loose.  It snaps back together with no problem, but like you
> say, it will void the warranty of course.  LinkSys makes sure of this
> with it's little security sticker on the bottom of the unit.

I finally managed to partially disassemble the case by doing these:
-  The case is made from three plastic pieces:
   o  The front black cap
   o  The top silver cover
   o  The silver base.
-  The left and right sides of the front black cap are
   snapped together with the left and right sides of
   the top silver cover.
-  By squeezing the left and right sides of the top
   silver cover at area where the front black cap meets
   the top silver cover, we can easily pull the front
   black cap out.
-  Then, we can easily remove the top silver cover from
   the silver base.

This was the easy part. The difficult part is how to remove the circuit
board from the silver plastic base. I still cannot figure out the
"right" way to remove the circuit board in order to remove the screws
that hold the mounting plate of the fan (The screws are inserted from
the bottom of the circuit board). I may need to drill a large hole
through the plastic silver base to reach each of the screw head.

I have found sources for a quiet 60mm fan (60mm x 60mm x 15mm) as shown
in the following links to replace the noisy 40mm fan:

http://www.quietpcusa.com/acb/showdetl.cfm?&DID=8&Product_ID=222&CATID=7
   http://www.endpcnoise.com/cgi-bin/e/papst_60mm.html

I have these questions:

1. It has 2 wires; but it has two 4-pin plugs.
   Can I simply cut out the 4-pin plugs and solder
   the 2 wires to the 2-pin plug of the noisy
   40mm fan?

2. The 40mm fan has two wires: red and black.
   The 60mm fan has two wires: red and blue.
   I assume I should connect the red wire to the
   red wire, and the black wire to the blue wire,
   right?

3. The power consumption of the 40mm fan is
   supposed to be 0.6W. The power consumption of
   the 60mm fan is rated as 0.4W. This is OK, right?

Thanks in advance for any info.

Jay Chan
Author
6 Mar 2005 5:18 AM
kony
On 5 Mar 2005 19:27:28 -0800, jaykc***@hotmail.com wrote:


>I have found sources for a quiet 60mm fan (60mm x 60mm x 15mm) as shown
>in the following links to replace the noisy 40mm fan:
>
>http://www.quietpcusa.com/acb/showdetl.cfm?&DID=8&Product_ID=222&CATID=7
>   http://www.endpcnoise.com/cgi-bin/e/papst_60mm.html

$21?  Yikes!
With a bit of searching you might find a much cheaper fan
that's only a little noisier.  $21 is very expensive for
such a fan, I used to buy other decent name-brands for $1
each.  Unfortunately I don't know where to find that size
cheap now.

>
>I have these questions:
>
>1. It has 2 wires; but it has two 4-pin plugs.
>   Can I simply cut out the 4-pin plugs and solder
>   the 2 wires to the 2-pin plug of the noisy
>   40mm fan?

yes

>
>2. The 40mm fan has two wires: red and black.
>   The 60mm fan has two wires: red and blue.
>   I assume I should connect the red wire to the
>   red wire, and the black wire to the blue wire,
>   right?

Odds are that's right, but you could also touch, hold the
wires together and confirm the fan spins before soldering,
or take voltage readings (to determine polarity).  I assume
you've confirmed that the two fans use same voltage, rarely
a router/switch/etc might use a 5V fan instead.

>
>3. The power consumption of the 40mm fan is
>   supposed to be 0.6W. The power consumption of
>   the 60mm fan is rated as 0.4W. This is OK, right?

Yes that's fine.

When fans are shoehorned into tight spots, their mounting
can introduce a lot of additional noise.  I don't know the
clearance on the new fan to the other parts but it may be
something you'd need to modify for quietest results.  One
simple aid can be putting washers under the fan on the
mounting studs, either to absorb vibration (if washers are
flexible) or at least raise fan blades further away from
components.
Author
8 Mar 2005 4:14 PM
jaykchan
Yes, the price of the quiet 60mm fan is quite a bit higher than what I
expected.  Unfortunately, I cannot find it in local retail stores, and
I cannot find anything in internet other than the expensive model.  You
see.  The fan must meet this criteria:
-  40mm or 60mm (height and width)
-  15mm thick or below in order to fit into the tight space
-  Relatively Quiet (hard to judge from description in the web page)
-  0.6W or lower power consumption.

Thanks for the confirmation that I can simply cut the wires and solder
them together to get around with the little problem of incompatible
plugs.

I figure the space is around 17mm.  Therefore, a 15mm fan should have a
little clearance left.  Worse comes to worse, I can file the edge of
the fan to fit.

Jay Chan
Author
8 Mar 2005 6:45 PM
kony
On 8 Mar 2005 08:14:15 -0800, jaykc***@hotmail.com wrote:

>Yes, the price of the quiet 60mm fan is quite a bit higher than what I
>expected.  Unfortunately, I cannot find it in local retail stores, and
>I cannot find anything in internet other than the expensive model.  You
>see.  The fan must meet this criteria:
>-  40mm or 60mm (height and width)

What about 50mm?


>-  15mm thick or below in order to fit into the tight space
>-  Relatively Quiet (hard to judge from description in the web page)
>-  0.6W or lower power consumption.
>
>Thanks for the confirmation that I can simply cut the wires and solder
>them together to get around with the little problem of incompatible
>plugs.


Well if you have no aversion to soldering, that opens up a
LOT more possibilities as it's not hard at all to find "any"
fan with those dimensions then simply solder on a 20 cent ~
47-120 Ohm 2W resistor onto the + power lead.

>
>I figure the space is around 17mm.  Therefore, a 15mm fan should have a
>little clearance left.  Worse comes to worse, I can file the edge of
>the fan to fit.
>
>Jay Chan

If you really wanted to get fancy (and don't have anything
sitting on top of the switch), you could cut out a hole on
top, mount the fan to the casing, and put a filter on top.
Having the fan draw in outside air would certainly result in
lower temps per same (or even lower) RPM, noise.

There are a ton of fans out there though, most significant
might be if you wanted to purchase anything else, somewhere,
as shipping a single fan can/is often more costly than the
fan itself.
Author
9 Mar 2005 4:33 PM
jaykchan
The problem is not the size, it is the thickness of the fan.  There are
many 60mm fans that are in 20mm or 25mm thickness.  But very few are
15mm thick, and those are the expensive kind.

Moreover, the power consumption is also a problem because most requires
more than what the existing fan use, and I want to avoid any trouble by
making sure that the replacement fan is equal to or less than what the
existing fan uses.  This further reduces the available pool of 60mm
fans.

I cannot use a 50mm fan because there is no mounting holes for it.

I ordered a $17 Papst quiet 60mm fan last night.  It matches all the
requirements.  I will know how it goes when I receive it.

BTW, the shipping costs as much as half of the cost of the fan itself.
Oh well...

Jay Chan
Author
6 Mar 2005 6:15 AM
SRay69
> I finally managed to partially disassemble the case by doing these:
> - The case is made from three plastic pieces:
> o The front black cap

Are you sure it's black?  Mine is a dark blue.  Try turning on a
bright light!  :)

> I still cannot figure out the "right" way to remove the circuit
board in order to remove the screws that hold the mounting plate of
the fan (The screws are inserted from the bottom of the circuit
board). I may need to drill a large hole through the plastic silver
base to reach each of the screw head.

Please, don't go drilling holes!  See my picture and you'll know
exactly how to remove the circuit board from the silver base.  This
threw me too at first.  Yes, there are two screws you can't get to.
BUT, there is a third screw near the back of the unit that holds the
fan mounting plate AND the circuit board to the plastic base.  See
the middle picture here:

http://members.dslextreme.com/users/jperry69/LinkSysGigabitFanFix.jpg

It's actually down right now, but hopefully should be up later.  Gotta
love ISP hosting.

Anyway, see the arrow pointing to that screw?  Do you not have the
same screw?  If you have a black front cap, then you might have a
different version than I.


> I have these questions:
>
> 1. It has 2 wires; but it has two 4-pin plugs.  Can I simply cut out
the 4-pin plugs and solder the 2 wires to the 2-pin plug of the noisy
40mm fan?

Yep, that's what I did!

> 2. The 40mm fan has two wires: red and black.  The 60mm fan has two
wires: red and blue.  I assume I should connect the red wire to the
red wire, and the black wire to the blue wire, right?

Yes, that should be correct, but you should test it first on a
computer power supply if possible just so there's no chance of
hurting the Gigabit switch if it happens to not have any polarity
protection.  You shouldn't be able to hurt the fan, but you might
want to buy 2 just in case.  It's always good to have an extra fan
lying around!

> 3. The power consumption of the 40mm fan is supposed to be 0.6W. The
power consumption of the 60mm fan is rated as 0.4W. This is OK, right

I'm starting to wonder if we have the same switch.  I have the
EG008W:
http://www.linksys.com/products/product.asp?grid=35&scid=42&prid=529

My original 40mm LinkSys fan shows 0.06 AMPS and the new 60mm fan
shows 0.1A.  The unit has been running with the new fan for a couple
days now with no problems.

Note, the 60mm fan fits tightly in the case all around.  Not one fan
hole would line up for any screw, so I strategically used 2 small tie
wraps to secure the fan to the metal plate.  You might be able to get
away with not securing it since the silver top holds it pretty
tightly, enough to bulge it a bit, but you don't notice it after it's
all together.

Don't forget to position the fan so the air is pushed up, NOT down.
Author
8 Mar 2005 4:22 PM
jaykchan
The front cap looks like black with a hint of blue.  It may be a very
very dark blue.

Thanks for the picture.  I take a look of the picture one more time.
This time I understand what that screw is for.  I thought the screw on
the picture was for the metal fan mount.  I unscrew it and I can easily
remove the circuit board from the silver plastic base.  I am very glad
to be able to get it off.  Now, I can proceed to order a quiet
replacement fan.

Thanks for the confirmation that I can simply cut the wires and solder
them together with the plugs that come from the noisy fan.

The label on the noisy 40mm fan states that its power consumption is
0.6W.  I don't use any instrument to confirm this number.

Based on the pictures, I have a feeling that the metal fan mount in
yours is slightly different from mine.  I figure that I should not have
any problem screwing a 60mm fan onto the metal fan mount.

Jay Chan
Author
4 Apr 2005 3:10 PM
jaykchan
Just want to let you know that I finally get the LinkSys Gigabit switch
to quiet enough by doing exactly as what you have done:
-  Replace the noisy 40mm fan with a quiet 60mm fan.
-  Undervolt the fan with a power resistor.

I believe the noise from the original setting was coming from two
sources:
-  Primarily the noise was from the noisy 40mm fan.
   It was a noisy beast when I ran it without any
   enclosure.
-  Secondarily the noise was from the air rushing
   through the small vent holes on the plastic case
   right on top of the fan. When I put my finger over
   other part of the case, the noise didn't reduce.
   When I put my finger over the vent holes, the noise
   was significantly reduced.
The total of these two sources of noises combined was greater than any
one of the noise above.

The very quiet Panasonic Panaflo 60mm fan that I use now was virtually
silent when I ran it without any enclosure. This means using this fan
eliminates the primary source of noise mentioned above. Unfortunately,
this oversized fan produces so much more air flow than the 40mm fan
that it generates a lot more air rushing through the vent holes that it
actually generates more secondary noise than what the 40mm fan
generated. This explains the reason why I still heard noise after I
have just replace the noisy 40mm fan with the quiet 60mm fan. I can
tell just by putting my hand over the 60mm fan and the 40mm fan that
the 60mm fan can produce much more air flow than the 40mm fan. This
means I need to find a way to reduce the secondary noise.

Luckily, you mentioned the use of power resistor to reduce the speed of
the fan to further reduce the noise. I follow your advice and add a
"100 ohm 2W 5% carbon composite" power resistor inline with the fan.
Now the gigabit switch is much more quiet -- to the point that I cannot
hear the noise if I stay 5-ft away. Because of the fact that I put the
gigabit switch inside a closet, this means I cannot hear the noise if I
close the door of the closet. Problem solved!

I find that even after I have cut down the speed of the 60mm fan, I
still can feel (using my hand) that it is generating more air flow than
the 40mm fan can in full speed. This means I could have reduced the
speed even further to cut down the noise to the minimum. I will let
other people to try this. For now, I am more than happy of the result,
and will not change a thing.

Thanks for the great practical advice that you have provided. Without
those advices, I probably still leave the gigabit switch in the box.

Jay Chan
Author
4 Apr 2005 11:11 PM
kony
On 4 Apr 2005 08:10:41 -0700, jaykc***@hotmail.com wrote:

>Just want to let you know that I finally get the LinkSys Gigabit switch
>to quiet enough by doing exactly as what you have done:
>-  Replace the noisy 40mm fan with a quiet 60mm fan.
>-  Undervolt the fan with a power resistor.

<snip>

>"100 ohm 2W 5% carbon composite" power resistor inline with the fan.

<snip>

>I find that even after I have cut down the speed of the 60mm fan, I
>still can feel (using my hand) that it is generating more air flow than
>the 40mm fan can in full speed. This means I could have reduced the
>speed even further to cut down the noise to the minimum. I will let
>other people to try this. For now, I am more than happy of the result,
>and will not change a thing.
>
>Thanks for the great practical advice that you have provided. Without
>those advices, I probably still leave the gigabit switch in the box.
>
>Jay Chan

Good to hear it worked out well for you.  While it's likely
you could reduce the fan speed a little further, beyond a
certain point there might be the issue of minimal spin-up
voltage.  That is, the fan needs a certain voltage to begin
spinning from a complete stop at power on, and that voltage
could easily result in the fan still having more flow than a
smaller fan would... those 40mm fans really don't move much
air at all, I suspect most 60mm fans at their absolute
minimum voltage would do as well.

Personally I try to avoid above 120 Ohms when unsure,
instead settling for a series of diodes which won't drop the
current like a resistor would.  It's much more tedious to
determine numbers of diodes to use though, so I seldom
recommend that method.  For my own use I have a 12 position
switch which additively puts more diodes in series to
determine optimal values for a fan, but without such an easy
method it wouldn't be practical except for someone wanting
the ultimate control over fan speed without a larger
controller assembly. 

Most often it isn't necessary to get a fan THAT near it's
lowest-possible voltage, especially with smaller fans.  I
have a few 6 1/2" monsters that need it a lot more than "PC"
fans but those monster fans have their own sets of issues,
actually have a linear regulator inside to set same RPM from
a variety of input voltages, a circuit that must be
circumvented to even begin to use external RPM control
(except when manufacturered with such a feature).
Author
5 Apr 2005 6:18 PM
jaykchan
> Personally I try to avoid above 120 Ohms when unsure,
> instead settling for a series of diodes which won't drop the
> current like a resistor would.  It's much more tedious to
> determine numbers of diodes to use though, so I seldom
> recommend that method.  For my own use I have a 12 position
> switch which additively puts more diodes in series to
> determine optimal values for a fan, but without such an easy
> method it wouldn't be practical except for someone wanting
> the ultimate control over fan speed without a larger
> controller assembly.

Thanks for suggesting the use of a 12-position switch. Although I don't
need to use anything like this at this moment, I may need it when I
need to build a HTPC that needs to be very quiet. What's the keywords
to search for a 12-position switch anyway?

> Most often it isn't necessary to get a fan THAT near it's
> lowest-possible voltage, especially with smaller fans.

You are right. In my case, I have lowered the speed of the 60mm fan
enough that I cannot hear its noise from 4-to-5-ft away (in a very
quiet environment). This is good enough even though I could have
reduced the noise even further.

Jay Chan
Author
5 Apr 2005 7:19 PM
kony
On 5 Apr 2005 11:18:25 -0700, jaykc***@hotmail.com wrote:

Show quoteHide quote
>> Personally I try to avoid above 120 Ohms when unsure,
>> instead settling for a series of diodes which won't drop the
>> current like a resistor would.  It's much more tedious to
>> determine numbers of diodes to use though, so I seldom
>> recommend that method.  For my own use I have a 12 position
>> switch which additively puts more diodes in series to
>> determine optimal values for a fan, but without such an easy
>> method it wouldn't be practical except for someone wanting
>> the ultimate control over fan speed without a larger
>> controller assembly.
>
>Thanks for suggesting the use of a 12-position switch. Although I don't
>need to use anything like this at this moment, I may need it when I
>need to build a HTPC that needs to be very quiet. What's the keywords
>to search for a 12-position switch anyway?

Well "12 position" might work, but generally they're listed
by number of poles and throws, so a 12 position switch would
be a Single Pole, 12 throw, "SP12T" or 1P12T.  It may be
easier to just browse through an online store's categories
though to get to them.  The aforementioned electronics
oriented suppliers like Mouser, Digikey, Newark, Allied,
etc, etc, will likely have a few alternatives.

The one I used was here,
http://www.allelectronics.com/cgi-bin/category.cgi?category=700425&item=RS-117&type=store
but i wasn't being picky at the time, since i had no
intentions of mounting the switch anywhere or using it for
more than sizing up fan diodes.  Just about any switch can
handle the few hundred mA from a fan so I suggest buying
whatever's available at a good price when you next need
"other" things from same store... I'd never buy individual
components like this except in an emergency, shipping and
minimum order fees will get quite expensive that way so I
find it easier to just buy at good spot-prices, more and
other than what I need at any particular time and then have
spare parts available. 

We may have different perspectives though, and different
parts needs.  If it were me building *only* another system
with slower fans I'd probably not go to the trouble of
assembling a 12 position switched diode controller, rather
starting out by using some panaflo "L" speed fans and seeing
if they'll run OK from 5V via a simple 4-pin molex fan
adapter (swapping it's 5V & 12V pins)  like one of these:

http://www.svcompucycle.com/conad.html
Most commonly this one:
http://www.svcompucycle.com/3pinto4pinad.html

SVC.COM is about as cheap as it gets for small fan adapters
as they'll throw them into a USPS envelope... though
sometimes their shopping cart shipping calculator is
malfunctional so scrutinize their quoted shipping cost to
see if it looks appropriate for something so small and
light.



>
>> Most often it isn't necessary to get a fan THAT near it's
>> lowest-possible voltage, especially with smaller fans.
>
>You are right. In my case, I have lowered the speed of the 60mm fan
>enough that I cannot hear its noise from 4-to-5-ft away (in a very
>quiet environment). This is good enough even though I could have
>reduced the noise even further.

It is harder to get a fan quiet with obstructions so
nearby... and yet the obstructions further reduce the
airflow.   I'd consider just putting the router
under/behind/etc  a desk if you need it quieter, plus that
gets it out of the way, out of sight too... seems like
tethered computer gear keeps multiplying all around us.
Author
6 Apr 2005 5:26 PM
jaykchan
> Well "12 position" might work, but generally they're listed
> by number of poles and throws, so a 12 position switch would
> be a Single Pole, 12 throw, "SP12T" or 1P12T.  It may be
> easier to just browse through an online store's categories
> though to get to them.  The aforementioned electronics
> oriented suppliers like Mouser, Digikey, Newark, Allied,
> etc, etc, will likely have a few alternatives.
>
> The one I used was here,
>
http://www.allelectronics.com/cgi-bin/category.cgi?category=700425&item=RS-117&type=store

Thanks for the link and the keywords.

Jay Chan
Author
7 Mar 2005 10:17 PM
SRay69
> When fans are shoehorned into tight spots, their mounting can
introduce a lot of additional noise. I don't know the clearance on
the new fan to the other parts but it may be something you'd need to
modify for quietest results. One simple aid can be putting washers
under the fan on the mounting studs, either to absorb vibration (if
washers are flexible) or at least raise fan blades further away from
components.

Hey there.  I doubt washers would be a good idea since there's no
space for anything even a penny high when using a 60mm fan.  I
wouldn't recommend using anything smaller since the airflow would be
compromised.  After all I do have some experience since I already
replaced the LinkSys fan.

Perhaps you have a picture of your installation?  I posted one earlier
that I did.

Here it is again:
http://members.dslextreme.com/users/jperry69/LinkSysGigabitFanFix.jpg
Author
8 Mar 2005 9:21 AM
kony
On 7 Mar 2005 17:17:09 -0500,
jim_pe***@hotmail-dot-com.no-spam.invalid (SRay69) wrote:

Show quoteHide quote
>> When fans are shoehorned into tight spots, their mounting can
>introduce a lot of additional noise. I don't know the clearance on
>the new fan to the other parts but it may be something you'd need to
>modify for quietest results. One simple aid can be putting washers
>under the fan on the mounting studs, either to absorb vibration (if
>washers are flexible) or at least raise fan blades further away from
>components.
>
>Hey there.  I doubt washers would be a good idea since there's no
>space for anything even a penny high when using a 60mm fan.  I
>wouldn't recommend using anything smaller since the airflow would be
>compromised.  After all I do have some experience since I already
>replaced the LinkSys fan.
>
>Perhaps you have a picture of your installation?  I posted one earlier
>that I did.
>
>Here it is again:
>http://members.dslextreme.com/users/jperry69/LinkSysGigabitFanFix.jpg

If there is no space, you're not getting good airflow.
There was supposed to be space above the fan, between it and
the casing... your pictures cannot make it clear whether
there is any space. 

Regardless, the sweeping idea that "wouldn't recommend
anything smaller since airflow would be compromised" is
quite premature.  On the contrary, just about any fan you
put in there is going to make marginal difference (compared
to a different fan)  until you devise a method to remove
that heat from the casing itself rather than just
recirculate it inside.  Otherwise it's not a matter of
trying to get same flow rate, only a matter of keeping
device cool enough.  If they'd used a casing with more
ventilation holes it's quite possible an entirely passive
heatsink couldn've been used instead, as most modern
consumer-level  switches/routers/etc don't have one.
Author
10 Mar 2005 1:37 PM
SRay69
> If there is no space, you're not getting good airflow.  There was
supposed to be space above the fan, between it and the casing... your
pictures cannot make it clear whether there is any space.

As long as there is a place for air to come from under the fan and a
place for it to go above the fan, then you're getting airflow.  I
feel about the same dissipation blowing out as when the smaller and
much higher RPM fan was in it.  It's working great.  If Chan wants to
do it your way, great.  I just hope he has the kind of time you do.

> Regardless, the sweeping idea that "wouldn't recommend anything
smaller since airflow would be compromised" is quite premature. On
the contrary, just about any fan you put in there is going to make
marginal difference (compared to a different fan) until you devise a
method to remove that heat from the casing itself rather than just
recirculate it inside. Otherwise it's not a matter of trying to get
same flow rate, only a matter of keeping device cool enough. If
they'd used a casing with more ventilation holes it's quite possible
an entirely passive heatsink couldn've been used instead, as most
modern consumer-level switches/routers/etc don't have one.

I think YOU'RE premature for posting here since you haven't even
replaced one of these.  The point of it was to reduce the noise to a
whisper while giving at least the same heat dissipation as the
original fan and I've done that.  Who knows if using a slower version
of same original-size fan would be enough to keep the box from
overheating.  I wanted to get this right the first time and move on.
Replacing the fan with the same flow rate fan, but larger, has worked
perfectly.  It's been running for a week now with no problems.

I don't have time to sit and try different fans to cool this thing in
the best possible way.  I just wanted to get rid of the irrating
noise without it overheating, in the least possible time, just like
most other buyers of this unit.

And by the way, I didn't solder the my wires together at all.  There
is no strain at all on the wires in this thing, so simply twisting
them together well and using some black electrical or rubber tape
does the trick.  I don't think they'll be coming lose until this
thing is recycled or in a landfill.

I'm a perfectionist too, but only to a point.  I've learned where to
draw the line and not to f-around so much.  I've definitely spent way
more time posting here than it took to ever fix this LinkSys switch
issue.  It's not a big enough deal to continue the conversation.
Author
10 Mar 2005 4:00 PM
kony
On 10 Mar 2005 08:37:08 -0500,
jim_pe***@hotmail-dot-com.no-spam.invalid (SRay69) wrote:

>> If there is no space, you're not getting good airflow.  There was
>supposed to be space above the fan, between it and the casing... your
>pictures cannot make it clear whether there is any space.
>
>As long as there is a place for air to come from under the fan and a
>place for it to go above the fan, then you're getting airflow.  I
>feel about the same dissipation blowing out as when the smaller and
>much higher RPM fan was in it.  It's working great.  If Chan wants to
>do it your way, great.  I just hope he has the kind of time you do.

It takes a lot of time to do it?  Hardly, refitting new fans
is a fairly routine thing though usually on old equipment
rather than new.


Show quoteHide quote
>
>> Regardless, the sweeping idea that "wouldn't recommend anything
>smaller since airflow would be compromised" is quite premature. On
>the contrary, just about any fan you put in there is going to make
>marginal difference (compared to a different fan) until you devise a
>method to remove that heat from the casing itself rather than just
>recirculate it inside. Otherwise it's not a matter of trying to get
>same flow rate, only a matter of keeping device cool enough. If
>they'd used a casing with more ventilation holes it's quite possible
>an entirely passive heatsink couldn've been used instead, as most
>modern consumer-level switches/routers/etc don't have one.
>
>I think YOU'RE premature for posting here since you haven't even
>replaced one of these. 

Sure, I've never seen a fan before let alone a device that
uses them, LOL.  Pull head out of arse and look around, the
only thing unusual about this particular fan swap was
figuring out how to get the plastic case open.


>The point of it was to reduce the noise to a
>whisper while giving at least the same heat dissipation as the
>original fan and I've done that.

Oh?  Did you take temp measurements?  Do you know the
thermal margins of the chip so you'd even know if trying to
match the original spec is important, rather than trying ot
match the actual needs of the device?  This isn't a knock
against you as much as the original engineers of the device
(or the penny-pinchers who changed the design) so it ended
up sub-optimal to begin with.


> Who knows if using a slower version
>of same original-size fan would be enough to keep the box from
>overheating. 

I do but that's beside the point?
They don't have a nuclear fusion reactor in there, as I
wrote previously it should be cool enough with only a good
heatsink as plenty of other brands have demonstrated.


>I wanted to get this right the first time and move on.
>Replacing the fan with the same flow rate fan, but larger, has worked
>perfectly.  It's been running for a week now with no problems.

You did fine for the first time.  Having been dealing with
different size fans for so long that I can't even remember
where they all are (brand new), I have a different
perspective.  What you did do right which many do wrong is
choosing a thicker fan.  10mm fans are too thin for
longevity due to a decent dual bearing (or long sleeve)
practically requiring 12-15mm thickness.


>
>I don't have time to sit and try different fans to cool this thing in
>the best possible way.  I just wanted to get rid of the irrating
>noise without it overheating, in the least possible time, just like
>most other buyers of this unit.

I can appreciate that, and yet I was also mentioning my
experiences with this... there's more than one way to get
the job done and I've done it several ways on different
equipment.  Happen to have an old Lantronix unit just thrown
onto the to-do pile that needs one replaced too.

If nothing else we can see that there are multiple choices
and it hurts nothing to discuss them all.


>
>And by the way, I didn't solder the my wires together at all.  There
>is no strain at all on the wires in this thing, so simply twisting
>them together well and using some black electrical or rubber tape
>does the trick.  I don't think they'll be coming lose until this
>thing is recycled or in a landfill.


Maybe not though electrical tape  has a way of degrading and
leeching black goo over time, especially in a warmer
environment.  I'd always advise soldering and heatshrinking
or crimping the wires, else putting a mating connector on
each which is fine if you happen to have the connectors but
possibly excessive if not.

>
>I'm a perfectionist too, but only to a point.  I've learned where to
>draw the line and not to f-around so much.  I've definitely spent way
>more time posting here than it took to ever fix this LinkSys switch
>issue.  It's not a big enough deal to continue the conversation.

Ok. I may've been too critical with my comments, but one of
the points might've been that it's not so necessary to try
and match the original fan's flow rate.  In these types of
applications a tiny bit of airflow goes a long way.
Author
2 Apr 2005 1:44 AM
timcuculic
FYI

newegg.com has a Panasonic Panaflo 60 x 15mm Cooling Fan, Model
"FBA06T12L" (Noise Level: 24dBA) for 11.98 including shipping.

Power consumption is 0.13 Amps @ 12 V (1.56W) though. Might be a bit
high.
Author
2 Apr 2005 3:01 AM
kony
On Sat, 02 Apr 2005 01:44:33 +0000,
timcucu***@yahoo-dot-com.no-spam.invalid (timcuculic) wrote:

>FYI
>
>newegg.com has a Panasonic Panaflo 60 x 15mm Cooling Fan, Model
>"FBA06T12L" (Noise Level: 24dBA) for 11.98 including shipping.
>
>Power consumption is 0.13 Amps @ 12 V (1.56W) though. Might be a bit
>high.

That's still quiter than most, it would be a good choice
with a resistor or other method of RPM reduction added.
Author
4 Apr 2005 2:48 PM
jaykchan
> newegg.com has a Panasonic Panaflo 60 x 15mm Cooling Fan, Model
> "FBA06T12L" (Noise Level: 24dBA) for 11.98 including shipping.
>
> Power consumption is 0.13 Amps @ 12 V (1.56W) though. Might be a bit
> high.

That dimension is the same as the expensive $19.x Panasonic Panaflo
that I have bought. But is power consumption is higher than what the
original 40mm fan is speced for. Therefore, I am very hesistant to
order that one.

But your info may become useful for someone who wants the same
dimension fan, and don't mind the higher power consumption.

Jay Chan