Replace 3 wire fan with 4 wire fan ?

I got the wrong fan. It is one for cooling the computer chassis instead of the cpu cooling fan.

The new one has a 4 wire connector while the old is 3 wire.

Can I use the new one in place of the 3 wire fan ?

New fan does draws twice the amperage of the old, so I quess it would be louder.

Andy

Andy wrote:
I got the wrong fan. It is one for cooling the computer chassis instead of the cpu cooling fan.

The new one has a 4 wire connector while the old is 3 wire.

Can I use the new one in place of the 3 wire fan ?

New fan does draws twice the amperage of the old, so I quess it would be louder.

Andy

The alignment tab on a four pin fan, is the same
size as a three pin fan. So that a three pin can be
connected to a four pin motherboard header.

Pin Signal Function
1 GND
2 +12V
3 RPM Fan sends out two pulses per rotation, to the computer
4 PWM Fan accepts 5V level, pulse width modulation for speed control.
Computer sends 25KHz signal, where the pulse width varies with
desired speed setting.

Pin Signal Function
1 GND
2 +12V
3 RPM Fan sends out two pulses per rotation, to the computer

If the PWM pin is left "floating", it is pulled high by the fan.
A permanently high signal on PWM, is like a 25KHz signal where
the "logic 1" width is 100%. The floating pin ends up being a
request for "maximum speed". The 25KHz was chosen, to be out
of the audible human hearing range.

*******

Fans fit roughly into four CFM categories. Low, medium, high, ultra.
I have an ultra here, at around 110CFM and maybe one ampere of current.
A low or medium might be in the 35CFM range, maybe 100-200ma current,
and those are relatively quiet. You can't really sit next to the 110CFM
fan, which is 120mm outside dimensions, and 37.5mm deep. That fan
has to be turned down, to be useful.

A CPU fan doesn't have to be a monster, if the heatsink fins have a
relatively large surface area, and the fin spacing is such that
air moves easily (low arrestance). Then, a lower CFM fan can be used,
or a medium fan can be speed controlled to good effect.

You tell by the CPU temperature, while a 100% loading program
(Prime95) is running, as to whether you got the right fan. By
matching the CFM rating and size dimensions, that should get
you into the right ballpark. If you don't have a CFM to work with,
then current draw will give a rough idea.

The only computer designs that use the ultra idea, are things
like single-fan Dells, where one fan cools the CPU and computer
case. Some of those have a high rating, which under normal
circumstances is never used. The fan in that case, stays turned
down, unless the machine goes nuts. And then the vacuum cleaner
noise of the machine, alerts you a failure is occurring on your
machine.

Paul

Paul wrote:
Andy wrote:
I got the wrong fan. It is one for cooling the computer chassis instead
of
the cpu cooling fan. The new one has a 4 wire connector while the old is
3 wire.

Can I use the new one in place of the 3 wire fan ?

New fan does draws twice the amperage of the old, so I quess it would be
louder. Andy

The alignment tab on a four pin fan, is the same
size as a three pin fan. So that a three pin can be
connected to a four pin motherboard header.

Pin Signal Function
1 GND
2 +12V
3 RPM Fan sends out two pulses per rotation, to the computer
4 PWM Fan accepts 5V level, pulse width modulation for speed
control.
Computer sends 25KHz signal, where the pulse width varies
with
desired speed setting.

Pin Signal Function
1 GND
2 +12V
3 RPM Fan sends out two pulses per rotation, to the computer

If the PWM pin is left "floating", it is pulled high by the fan.
A permanently high signal on PWM, is like a 25KHz signal where
the "logic 1" width is 100%. The floating pin ends up being a
request for "maximum speed". The 25KHz was chosen, to be out
of the audible human hearing range.

*******

Fans fit roughly into four CFM categories. Low, medium, high, ultra.
I have an ultra here, at around 110CFM and maybe one ampere of current.
A low or medium might be in the 35CFM range, maybe 100-200ma current,
and those are relatively quiet. You can't really sit next to the 110CFM
fan, which is 120mm outside dimensions, and 37.5mm deep. That fan
has to be turned down, to be useful.

A CPU fan doesn't have to be a monster, if the heatsink fins have a
relatively large surface area, and the fin spacing is such that
air moves easily (low arrestance). Then, a lower CFM fan can be used,
or a medium fan can be speed controlled to good effect.

You tell by the CPU temperature, while a 100% loading program
(Prime95) is running, as to whether you got the right fan. By
matching the CFM rating and size dimensions, that should get
you into the right ballpark. If you don't have a CFM to work with,
then current draw will give a rough idea.

The only computer designs that use the ultra idea, are things
like single-fan Dells, where one fan cools the CPU and computer
case. Some of those have a high rating, which under normal
circumstances is never used. The fan in that case, stays turned
down, unless the machine goes nuts. And then the vacuum cleaner
noise of the machine, alerts you a failure is occurring on your
machine.

Paul

Interesting! I'm assuming they use PWM to control the motor speed since
it's more cost effective than simply supplying a variable DC voltage (like
out of a D-A converter) to the fan motor(?)

Bill in Co wrote:
Paul wrote:
Andy wrote:
I got the wrong fan. It is one for cooling the computer chassis instead
of
the cpu cooling fan. The new one has a 4 wire connector while the old is
3 wire.

Can I use the new one in place of the 3 wire fan ?

New fan does draws twice the amperage of the old, so I quess it would be
louder. Andy
The alignment tab on a four pin fan, is the same
size as a three pin fan. So that a three pin can be
connected to a four pin motherboard header.

Pin Signal Function
1 GND
2 +12V
3 RPM Fan sends out two pulses per rotation, to the computer
4 PWM Fan accepts 5V level, pulse width modulation for speed
control.
Computer sends 25KHz signal, where the pulse width varies
with
desired speed setting.

Pin Signal Function
1 GND
2 +12V
3 RPM Fan sends out two pulses per rotation, to the computer

If the PWM pin is left "floating", it is pulled high by the fan.
A permanently high signal on PWM, is like a 25KHz signal where
the "logic 1" width is 100%. The floating pin ends up being a
request for "maximum speed". The 25KHz was chosen, to be out
of the audible human hearing range.

*******

Fans fit roughly into four CFM categories. Low, medium, high, ultra.
I have an ultra here, at around 110CFM and maybe one ampere of current.
A low or medium might be in the 35CFM range, maybe 100-200ma current,
and those are relatively quiet. You can't really sit next to the 110CFM
fan, which is 120mm outside dimensions, and 37.5mm deep. That fan
has to be turned down, to be useful.

A CPU fan doesn't have to be a monster, if the heatsink fins have a
relatively large surface area, and the fin spacing is such that
air moves easily (low arrestance). Then, a lower CFM fan can be used,
or a medium fan can be speed controlled to good effect.

You tell by the CPU temperature, while a 100% loading program
(Prime95) is running, as to whether you got the right fan. By
matching the CFM rating and size dimensions, that should get
you into the right ballpark. If you don't have a CFM to work with,
then current draw will give a rough idea.

The only computer designs that use the ultra idea, are things
like single-fan Dells, where one fan cools the CPU and computer
case. Some of those have a high rating, which under normal
circumstances is never used. The fan in that case, stays turned
down, unless the machine goes nuts. And then the vacuum cleaner
noise of the machine, alerts you a failure is occurring on your
machine.

Paul

Interesting! I'm assuming they use PWM to control the motor speed since
it's more cost effective than simply supplying a variable DC voltage (like
out of a D-A converter) to the fan motor(?)



Yes, it's $1 cheaper.

Paul

On Fri, 07 Mar 2014 20:21:28 -0800, Todd wrote:

On 03/07/2014 07:56 PM, Andy wrote:
I got the wrong fan. It is one for cooling the computer chassis instead of the cpu cooling fan.

The new one has a 4 wire connector while the old is 3 wire.

Can I use the new one in place of the 3 wire fan ?

New fan does draws twice the amperage of the old, so I quess it would be louder.

Andy


HI Andy,

You don't have to use the fourth pin. That is
the RPM control used for very slow operation.

If you take a close look at the following adapter, or
actually use it, it will tell you the rpm pin:

http://www.startech.com/Cables/Compu...-Cable~TX3EXT6


Since this is so close to a question in the back of my head, I hope you
won't mind if I hitch a ride.

I repaired a Dell Dimension 4400 for a friend and it had a 3-wire fan
and a 3-wire connector for it on the mobo. but the Dell expert told
me that a computer that old wouldn't report fan speed to the software,
and indeed when I installed software, it showed a fan speed of zero,
even when I know it was running.

So why the 3-wire fan? Did they stop making 2-wire fans, or made so
few they were more expensive?

Don't anyone waste time doing research on this. It's only a question of
curiosity.

Where did you find this fan? I need ball bearing case
fans with 4 pins (motherboard connectors).

-T

micky wrote:
On Fri, 07 Mar 2014 20:21:28 -0800, Todd wrote:

On 03/07/2014 07:56 PM, Andy wrote:
I got the wrong fan. It is one for cooling the computer chassis instead of the cpu cooling fan.

The new one has a 4 wire connector while the old is 3 wire.

Can I use the new one in place of the 3 wire fan ?

New fan does draws twice the amperage of the old, so I quess it would be louder.

Andy

HI Andy,

You don't have to use the fourth pin. That is
the RPM control used for very slow operation.

If you take a close look at the following adapter, or
actually use it, it will tell you the rpm pin:

http://www.startech.com/Cables/Compu...-Cable~TX3EXT6


Since this is so close to a question in the back of my head, I hope you
won't mind if I hitch a ride.

I repaired a Dell Dimension 4400 for a friend and it had a 3-wire fan
and a 3-wire connector for it on the mobo. but the Dell expert told
me that a computer that old wouldn't report fan speed to the software,
and indeed when I installed software, it showed a fan speed of zero,
even when I know it was running.

So why the 3-wire fan? Did they stop making 2-wire fans, or made so
few they were more expensive?

Don't anyone waste time doing research on this. It's only a question of
curiosity.

Where did you find this fan? I need ball bearing case
fans with 4 pins (motherboard connectors).

-T


They make two kinds of fan status signals.

So-called "computer fans" use the two pulse per revolution (RPM)
signal on the third wire.

Fans for cooling other kinds of equipment (instrumentation boxes),
they're not monitored necessarily by something with a display.
And those use a LR (Locked Rotor) status signal. The line state
is either high or low, depending on whether the rotor is jammed
at zero RPM, or the rotor is still spinning.

If the Dell had one of those fans, then there's be "nothing
to count" :-)

It's more normal for a chassis fan to have two wires, have
no RPM wire, and report zero RPM. That's the kind I've got.

Lots of SuperI/O chips have fan monitor channels. The chips
may vary as to how many channels they've got. Some motherboards
have as many as five of the three pin headers, and monitoring
is not available for all of them. Then, the count of monitored
channels in the BIOS, doesn't match the number of headers,
and you have to figure out which one is not monitored.

Paul

In message , Paul
writes:
Bill in Co wrote:
Paul wrote:
[]
4 PWM Fan accepts 5V level, pulse width modulation for speed
control.
Computer sends 25KHz signal, where the pulse width
varies with
desired speed setting.

But this is a control signal - the motor mainly runs from the 12V
supply, so does it actually matter whether the control signal is in the
audible range or not - or are you saying it just directly controls the
12V (through a power transistor or something)?
[]
Interesting! I'm assuming they use PWM to control the motor speed
since it's more cost effective than simply supplying a variable DC
voltage (like out of a D-A converter) to the fan motor(?)

Yes, it's $1 cheaper.

Paul

(-: Two reasons: one, electric motors don't respond linearly to varying
d. c. supply voltage, they tend to stall and/or at least run
inefficiently at low (my old - 1970s, probably older technology, model
train set used to regularly overheat the transformer, which safety
cutout until it cooled off, at low speeds); two, even if they did, it'd
be inefficient: if you supply 6V from a 12V supply via a power
transistor (or linear power stage out of a DAC), you're dropping 6V
across the load (motor), and 6V across the power device, which thus gets
hot. The power device in a PWM system is either fully on - in which case
there's negligible voltage across the device, it's all across the load -
or fully off, in which case no power is flowing anyway. There's a lot
more to it really, especially with inductive loads like motors, but
that's the gist of why PWM is so popular. (Though as I've indicated
above I suspect there may be cases here where PWM is only signalling,
not directly controlling.)
--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Madness takes its toll. Please have exact change
[via Penny Mayes )]

J. P. Gilliver (John) wrote:
In message , Paul
writes:
Bill in Co wrote:
Paul wrote:
[]
4 PWM Fan accepts 5V level, pulse width modulation for speed
control.
Computer sends 25KHz signal, where the pulse width
varies with
desired speed setting.

But this is a control signal - the motor mainly runs from the 12V
supply, so does it actually matter whether the control signal is in the
audible range or not - or are you saying it just directly controls the
12V (through a power transistor or something)?
[]
Interesting! I'm assuming they use PWM to control the motor speed
since it's more cost effective than simply supplying a variable DC
voltage (like out of a D-A converter) to the fan motor(?)

Yes, it's $1 cheaper.

Paul

(-: Two reasons: one, electric motors don't respond linearly to varying
d. c. supply voltage, they tend to stall and/or at least run
inefficiently at low (my old - 1970s, probably older technology, model
train set used to regularly overheat the transformer, which safety
cutout until it cooled off, at low speeds); two, even if they did, it'd
be inefficient: if you supply 6V from a 12V supply via a power
transistor (or linear power stage out of a DAC), you're dropping 6V
across the load (motor), and 6V across the power device, which thus gets
hot. The power device in a PWM system is either fully on - in which case
there's negligible voltage across the device, it's all across the load -
or fully off, in which case no power is flowing anyway. There's a lot
more to it really, especially with inductive loads like motors, but
that's the gist of why PWM is so popular. (Though as I've indicated
above I suspect there may be cases here where PWM is only signalling,
not directly controlling.)

I figured that operating the circuitry between c/o and saturation would be
more efficient, and that's definitely an advantage for high current loads.

However, I don't think running a small dc motor with reduced voltage is such
a problem (for stall outs), but that IS one thing you can't do with an AC
motor, or it will stall, which makes controlling ac motors a bit of a PIA.
:-) (Changing frequency to change the ac motor speed isn't the easiest
thing to do. :-)

In message , Bill in Co
writes:
[]
However, I don't think running a small dc motor with reduced voltage is such

Our experience varies. It may depend on the type of motor, whether a
simple brush-commutating one, or one with electronics actually inside
it. For the simple type, I think it does tend to stall if fed less than
a percentage of it's nominal supply voltage (whereas PWM _at_ the supply
voltage gives it in effect knocks big enough for it to get over many
sticky spots), and the electronic types probably don't line it either.

a problem (for stall outs), but that IS one thing you can't do with an AC
motor, or it will stall, which makes controlling ac motors a bit of a PIA.
:-) (Changing frequency to change the ac motor speed isn't the easiest
thing to do. :-)


--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Madness takes its toll. Please have exact change
[via Penny Mayes )]

J. P. Gilliver (John) wrote:
In message , Bill in Co
writes:
[]
However, I don't think running a small dc motor with reduced voltage is
such a problem (for stall-outs) [sic]

Our experience varies. It may depend on the type of motor, whether a
simple brush-commutating one, or one with electronics actually inside
it. For the simple type, I think it does tend to stall if fed less than
a percentage of it's nominal supply voltage (whereas PWM _at_ the supply
voltage gives it in effect knocks big enough for it to get over many
sticky spots), and the electronic types probably don't line it either.

a problem (for stall outs), but that IS one thing you can't do with an AC
motor, or it will stall, which makes controlling ac motors a bit of a
PIA.
:-) (Changing frequency to change the ac motor speed isn't the easiest
thing to do. :-)

It probably also depends on how low a speed you want and how heavy the load
is. Granted, I didn't load down the motor very much :-) I am also
thinking of just the simple brush motor types (the "electronic ones" are too
new for me :-)