Help with Cooling Fan (Question)

[Shawn]

Two fans in my computer (Dell Dimension 2300) are making noise. One is in
the power supply and the other
is the processor fan. I know I can replace the processor fan (with a green
schroud) if I can find one. But can I
replace just the fan that resides in the power supply. Dell says no, but it
looks like a might be able to if I can
find the correct size and power. I figured 10 bucks for a fan, not 50 or 60
bucks for a power supply.

Thanks in advance,

Shawn

 

[John McGaw]

Two fans in my computer (Dell Dimension 2300) are making noise. One is in
the power supply and the other
is the processor fan. I know I can replace the processor fan (with a green
schroud) if I can find one. But can I
replace just the fan that resides in the power supply. Dell says no, but it
looks like a might be able to if I can
find the correct size and power. I figured 10 bucks for a fan, not 50 or 60
bucks for a power supply.

Thanks in advance,

Shawn

Yes, power supply fans can be replaced but there is an important caveat:
a power supply can kill you (or at least zap you badly enough that you
will not forget it unless the jolt whacks your head against the wall
hard enough to induce amnesia).

With care it could be done by almost anyone who can remove the screws
and splice some wires. Suitable care includes disconnecting the power,
removing the supply completely, allowing it to sit for long enough to
discharge the filter capacitors, touching no other conductors except the
fan leads, insulating the leads properly after splicing them, and
putting everything in the supply back together exactly as it was
including shielding, insulation, and wire routing.

If the PS uses a thermally controlled fan (I don't know if Dell spends
the extra $0.05 for that feature or not) some care must be taken to make
sure that the replacement fan has a similar power draw to the original
to avoid damaging the fan control circuitry.

 

[Bob]

If the PS uses a thermally controlled fan (I don't know if Dell spends
the extra $0.05 for that feature or not) some care must be taken to make
sure that the replacement fan has a similar power draw to the original
to avoid damaging the fan control circuitry.

Chances are that the entire PSU needs replacement if the fan is acting
up. In fact, it might be time for a new computer.


--

Violence Against Women and Department of Justice Reauthorization Act:

"Whoever...utilizes any device or software that can be used to
originate telecommunications or other types of communications that are
transmitted, in whole or in part, by the Internet... without
disclosing his identity and with intent to annoy, abuse, threaten, or
harass any person...who receives the communications...shall be fined
under title 18 or imprisoned not more than two years, or both."

 

[Chuck F.]

Chances are that the entire PSU needs replacement if the fan is
acting up. In fact, it might be time for a new computer.

This is a ridiculous attitude. First, replacing a Dell PSU may
cause unnecessary anguish, because Dell uses non-standard shapes
and pinouts. Second, replacing a computer makes no sense if it is
doing the job. Keeping up with the Jones boys just adds to the
expense and waste generation. So, barring unusual circumstances,
the OP should simply replace the fan while taking the appropriate
precautions.

--
Some informative links:
http://www.geocities.com/nnqweb/

http://www.caliburn.nl/topposting.html
http://www.netmeister.org/news/learn2quote.html

 

[kony]

Two fans in my computer (Dell Dimension 2300) are making noise. One is in
the power supply and the other
is the processor fan. I know I can replace the processor fan (with a green
schroud) if I can find one. But can I
replace just the fan that resides in the power supply. Dell says no, but it
looks like a might be able to if I can
find the correct size and power. I figured 10 bucks for a fan, not 50 or 60
bucks for a power supply.

Yes, both can be replaced. It seems curious that both are
now making noise, I wonder if fine dust is accumulating
elsewhere as well.

On a similar Dimension I've serviced, I took out the rear
exahaust (what you're calling the processor fan), peeled
back the label, took out the plug, and put in some heavy oil
(almost grease) and that did very well to quiet the fan.
You might try it- at worst it is only a temporary fix for a
few months but generally if you use the right lube and the
fan hasn't degraded horribly yet, you may get a lot longer
life.

Similar with the power supply fan if it has a sleeve bearing
rather than ball bearing (check label, it may list "ball" if
it has one or two ball bearings, or after peeling back label
it should be obvious if a sleeve bearing because you'll see
a bronze colored sleeve around the shaft rather than a
stainless steel colored bearing. Trying to lube a
ball-bearing is often only an emergency fix as it may seize
suddenly or become very noisey, but the sleeve bearing type
will also repond well to lubing.

Either can be replaced instead of course, just note the fan
dimensions, voltage (almost certainly 12V but it's good to
be sure of it), and current rating- since it probably won't
have an RPM rating on it. You might choose a "slight" bit
lower current rating but not much, both of the fans in your
system should be thermally controlled and if you chose too
low a current it could mean either insufficient airflow or
complete failure to spin at all.

If you don't have a spare system you can use during
downtime, you might need to do a quick exploratory operation
on the power supply to see what the fan rating is and how
it's connected for power- whether a plastic plug-in
connector or soldered onto the circuit board. The
difference suggest the obvious solution, to seek a fan with
similar connector if it has connector or have soldering iron
and heat-shrink tubing handy (or electrical tape plus a
nylon wire tie to secure the tape to be certain it never
unwraps when the PSU heats up and adhesive does too).

While you definitely need to unplug the power supply and
wait a couple minutes, after that period of time there is no
inherant danger inside, the caps have then drained and you
could rub your hands all over everything with no danger-
just don't plug it in till you have double-checked
everything, made sure the fan leads are secured away from
the fan blades and fixed in place. In other words, try to
minimize slack in the fan wiring and if possible secure it
so it doesn't fly around- a common method is to use a nylon
wire-tie to attach to a fixed part of the assembly- fan
corner hole or an indentation on the casing, or tied around
another wire- how the original wire was secured is
presumably a good method to use.

Providing the power supply fan had continued spinning at
reasonable RPM, it is likely you can just replace it. If
the fan had been severely effected and the RPM was extremely
low, such that there was insufficient airflow to keep PSU
cool, then the PSU may now be heat-stressed and would be
best replaced.

A Panaflo fan would be a good choice for the rear exhaust,
perhaps an "M" speed, maybe "H" (compare each current rating
to the original fan). I would choose a dual-ball bearing
fan in a quality brand for the PSU exhaust. Perhaps NMB.
You can find their model #s on their website to facilitate
easier web searches.

 

[Bob]

This is a ridiculous attitude. First, replacing a Dell PSU may
cause unnecessary anguish, because Dell uses non-standard shapes
and pinouts.

Are you saying that Dell doesn't supply replacement PSUs for their
computers?

Second, replacing a computer makes no sense if it is
doing the job.

There is a point where putting more money into something that is old
is not economical.

Keeping up with the Jones boys just adds to the
expense and waste generation.

Operating systems and applications become more bloated each cycle. You
may not want to keep up with Jones but you do have to keep up with
Micrisoft.

So, barring unusual circumstances,
the OP should simply replace the fan while taking the appropriate
precautions.

You apparently are unaware that electrolytic capacitors age.

--

Violence Against Women and Department of Justice Reauthorization Act:

"Whoever...utilizes any device or software that can be used to
originate telecommunications or other types of communications that are
transmitted, in whole or in part, by the Internet... without
disclosing his identity and with intent to annoy, abuse, threaten, or
harass any person...who receives the communications...shall be fined
under title 18 or imprisoned not more than two years, or both."

 

[Bob]

put in some heavy oil

Ever notice that PSU fans make noise in the colder months.


--

Violence Against Women and Department of Justice Reauthorization Act:

"Whoever...utilizes any device or software that can be used to
originate telecommunications or other types of communications that are
transmitted, in whole or in part, by the Internet... without
disclosing his identity and with intent to annoy, abuse, threaten, or
harass any person...who receives the communications...shall be fined
under title 18 or imprisoned not more than two years, or both."

 

[kony]

Ever notice that PSU fans make noise in the colder months.

I try to keep mine inside or at least give them ear muffs.

 

[Bob]

I try to keep mine inside or at least give them ear muffs.

I keep mine inside too, but I set the furnace at 65F so it gets a lot
cooler in winter than in summer.

I once had a tower near a slider window and when the window was
cracked open, the PSU would make a racket you could hear all over the
house. As it warmed up it quieted down.


--

Violence Against Women and Department of Justice Reauthorization Act:

"Whoever...utilizes any device or software that can be used to
originate telecommunications or other types of communications that are
transmitted, in whole or in part, by the Internet... without
disclosing his identity and with intent to annoy, abuse, threaten, or
harass any person...who receives the communications...shall be fined
under title 18 or imprisoned not more than two years, or both."

 

[kony]

I keep mine inside too, but I set the furnace at 65F so it gets a lot
cooler in winter than in summer.

I suppose it matter what the resultant temp is where these
systems are- might be even lower than 65F. I have systems
that don' exhibit fan noise at 65F but I have heard fans get
noisier when colder.

I once had a tower near a slider window and when the window was
cracked open, the PSU would make a racket you could hear all over the
house. As it warmed up it quieted down.

I suspect you need a new fan if the temp wasnt' much lower
than 65F... though if the fan were right next to the opening
it might be colder? You live in TX though, yes? Recalling
pictures sent from relatives near Austin, it's a kodak
moment when it's even cold enough for snow to settle- that's
not very cold.

 

[professor]

Yes, power supply fans can be replaced but there is an importan
caveat: a power supply can kill yo

You have to love it when you're told, "Sure you can do it, but i
might kill you." I guess this is another reason to build you
own computer. Usually you'll have around a 2-3 year warranty on th
PSU, so if it goes and it's not under warrantee, you might as wel
get a new one. Three years after you build your computer, you'll b
able to buy a power supply that will power it rather easily fo
probably around $50 or less. Three years ago a 350w PSU was all yo
needed for just about anything. Today, good 350w PSUs are rathe
cheap

 

[Bob]

I have systems
that don' exhibit fan noise at 65F but I have heard fans get
noisier when colder.

It was an old cheap PSU.

I suspect you need a new fan if the temp wasnt' much lower
than 65F.
Yep.

You live in TX though, yes?
Houston.

Recalling
pictures sent from relatives near Austin, it's a kodak
moment when it's even cold enough for snow to settle

It snows in the Panhandle.

that's not very cold.

It gets colder than 65F.


--

Violence Against Women and Department of Justice Reauthorization Act:

"Whoever...utilizes any device or software that can be used to
originate telecommunications or other types of communications that are
transmitted, in whole or in part, by the Internet... without
disclosing his identity and with intent to annoy, abuse, threaten, or
harass any person...who receives the communications...shall be fined
under title 18 or imprisoned not more than two years, or both."

 

[kony]

caveat: a power supply can kill you

You have to love it when you're told, "Sure you can do it, but it
might kill you." I guess this is another reason to build your
own computer.

Except that it won't. ATX power supplies do not retain a
lethal charge after unplugged. You'd have to have the
casing already off, power plugged in or be unplugging it
while challenging yoruself to touch something as quickly as
possible during a few seconds there's any HV remaining.

 

[professor]

Except that it won't. ATX power supplies do not retain

lethal charge after unplugged. You'd have to have th
casing already off, power plugged in or be unplugging i
while challenging yoruself to touch something as quickly a
possible during a few seconds there's any HV remaining

Never underestimate the lethal power of human stupidity. If peopl
can find a way to install a memory module backwards, they can find
way to kill themselves with a PSU

 

[kony]

Never underestimate the lethal power of human stupidity. If people
can find a way to install a memory module backwards, they can find a
way to kill themselves with a PSU.

Has nothing to do with stupidity, only recognizing what
plugged vs unplugged for a few seconds means.

 

[UCLAN]

Except that it won't. ATX power supplies do not retain a
lethal charge after unplugged. You'd have to have the
casing already off, power plugged in or be unplugging it
while challenging yoruself to touch something as quickly as
possible during a few seconds there's any HV remaining.

ATX supplies are no different from AT supplies when it comes to
the discharge rate of the 300v buss. It varies with design. I
always kept a 47K bleed resistor handy when playing with _any_
SMPS, AT or ATX.

 

[kony]

ATX supplies are no different from AT supplies when it comes to
the discharge rate of the 300v buss. It varies with design. I
always kept a 47K bleed resistor handy when playing with _any_
SMPS, AT or ATX.

Yes they are different! Surely I dont need to tell you why?
Hint: 5VSB

Again I ask for an example, show one power supply that is
ATX that does not self-discharge the caps. AFAIK, there is
not one. When you used that 47K bleed resistor, did you
actually meaure the voltage first? Part of the typical
manual-voltage selection type PSU is also a stabilization
gained by the resistors across the voltage doubling
subcircuit.

It doesn't vary by design unless you're talking about
something else as one design and an ATX power supply as the
other design.

 

[UCLAN]

Yes they are different! Surely I dont need to tell you why?
Hint: 5VSB

Again I ask for an example, show one power supply that is
ATX that does not self-discharge the caps. AFAIK, there is
not one. When you used that 47K bleed resistor, did you
actually meaure the voltage first? Part of the typical
manual-voltage selection type PSU is also a stabilization
gained by the resistors across the voltage doubling
subcircuit.

It doesn't vary by design unless you're talking about
something else as one design and an ATX power supply as the
other design.

Not all ATX PSU designs have +5vsb circuits that run off the
300v buss. Many early designs used a small bias type transformer
to provide the input for the +5vsb. As +5vsb current requirements
increased, more and more +5vsb designs went to a switching output
fed by the 300v buss. That's why I said it depends on the design.
And even in those PSUs where the +5vsb _is_ fed by the 300v, a
failure in that circuit *will* slow discharge rate.

As far as resistors across the 300v buss go, I would love to hear
explain why they are different/lower value in ATX type supplies.

 

[kony]

Not all ATX PSU designs have +5vsb circuits that run off the
300v buss. Many early designs used a small bias type transformer
to provide the input for the +5vsb.

So where are you claiming this transformer input is
filtered? The issue is not whether there was a separate
transformer as that is quite common today.

As +5vsb current requirements
increased, more and more +5vsb designs went to a switching output
fed by the 300v buss.

No, that was by far the most common arrangement since day 1
of ATX.

That's why I said it depends on the design.
And even in those PSUs where the +5vsb _is_ fed by the 300v, a
failure in that circuit *will* slow discharge rate.

"Slow" in context, nobody suggested a race to see if you can
unplug the AC power cord with one hand while simultaneously
lunging at the HV parts with the other to see if you can
manage to touch them before caps have drained. Rather,
unplugging the supply before it's opened is sufficient.

As far as resistors across the 300v buss go, I would love to hear
explain why they are different/lower value in ATX type supplies.

Who claimed they're different/lower values? Not I.

 

[professor]

Has nothing to do with stupidity, only recognizing wha

plugged vs unplugged for a few seconds means

In my old A+ class, I watched someone use a hammer to insert a socke
A CPU. I do not mean using the back of the hammer like a screwdrive
to push the heatsink’s connectors; I mean actually hammering the thin
into place, or at least trying to. You want to know what "ste
1" was for him in the assembly phase? Plug the power supply in
Trust me, there are people out there dumb enough to leave a PS
plugged in while opening it up. Then they may proceed to hit it wit
a hammer..