26 Jun
2012
26 Jun
'12
9:42 p.m.
So in laymans terms you vary the speed by narrowing or widening the pulses delivered on
the positive lead (duty cycle). A steady 12vdc and you get max air flow/speed? The tach
lead is registered by something on the mobo, and otherwise has nothing to do with the
rotation of the blades?
------------------------------
On Tue, Jun 26, 2012 8:42 PM PDT Dave McGuire wrote:
On 06/26/2012 11:27 PM, Chris Tofu wrote:
black and red leads are labeled, white is not.
Speed control? From
large Sun servers OOPS! Sorry. But can someone inform regardless, and
what kind of signal/circuit is required to control the speed, if
thats the case. Cant successfully look it up now, its this stupid
phone.
Put them BACK in the large Sun servers if they're anything remotely
recent.
It's almost definitely a tachometer lead. They're typically
controlled via PWM, with a closed-loop servo for speed control and
OS-accessible speed monitoring.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA 
26 Jun
26 Jun
9:53 p.m.
On 06/27/2012 12:42 AM, Chris Tofu wrote:
So in laymans terms you vary the speed by narrowing or
widening the
pulses delivered on the positive lead (duty cycle). A steady 12vdc
and you get max air flow/speed? The tach lead is registered by
something on the mobo, and otherwise has nothing to do with the
rotation of the blades?
This is usually the case, yes. Visualize PWM...perform an integration
over several cycles, to take the area under the curve. The energy
delivered to the load varies with the pulse width. But the switching
devices (usually MOSFETs) are going from cutoff to saturation and back,
never operating in active mode for more than a few microseconds at a
time, and so dissipating very little power, thus running VERY efficiently.
Many Sun systems drive their fans this way. Actually many high-end
server systems from many manufacturers drive their fans this way, as
does some PC hardware. It's very common.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
10:54 p.m.
On Wed, 27 Jun 2012, Dave McGuire wrote:
On 06/27/2012 12:42 AM, Chris Tofu wrote:
4-wire fans have also become very common, both in servers and consumer
PCs. These have a dedicated pwm control lead in addition to the power and
tach leads. The majority of these fans use red/black/blue/green for their
leads, but other color combinations exist too.
With 3-wire fans, one very common scheme is red/black/yellow, but some
manufacturers use white instead of yellow for the tach feedback. Other
times white can be pwm control, so you often need the manufacturer's
datasheet to know for certain what you have. Even then, sometimes the
datasheet can be wrong...I have some ADDA fans that had the wrong part
number put on them at the factory.
Some of the ADDA and Mechatronics fans I have in my machines use a white
lead for the tach. With the ADDA fans in particular, if you happen to
accidentally ground the tach lead, it will fry the fet transistor in the
fan itself.
Where some of these fans get to be really fun is when you have 8-wire
dual-motor units. These have two fans in-line that rotate in opposite
directions (but with airflow in the same direction), with each fan having
a different blade pitch/configuration. These type of fans are designed to
allow a very wide range of airflow changes depending on the mode of each
fan motor. The downside is that they tend to be incredibly loud due to air
turbulence if both fans are running at full speed.
So in laymans terms you vary the speed by
narrowing or widening the
pulses delivered on the positive lead (duty cycle). A steady 12vdc and
you get max air flow/speed? The tach lead is registered by something on
the mobo, and otherwise has nothing to do with the rotation of the
blades?
This is usually the case, yes. Visualize PWM...perform an integration
over several cycles, to take the area under the curve. The energy
delivered to the load varies with the pulse width. But the switching
devices (usually MOSFETs) are going from cutoff to saturation and back,
never operating in active mode for more than a few microseconds at a
time, and so dissipating very little power, thus running VERY
efficiently.
Many Sun systems drive their fans this way. Actually many high-end
server systems from many manufacturers drive their fans this way, as
does some PC hardware. It's very common. 
27 Jun
27 Jun
1:24 a.m.
On 27/06/2012 05:53, Dave McGuire wrote:
On 06/27/2012 12:42 AM, Chris Tofu wrote:
This months Everyday Practical Electronics
http://www.epemag3.com/
features a PIC based Fan Speed controller (see front cover on link) that
discusses this but I am sure the same info is available freely on the
net elsewehere. If you are stuck I can scan the apposite portion of the
article..
--
Dave Wade G4UGM
Illegitimi Non Carborundum
So in laymans terms you vary the speed by
narrowing or widening the
pulses delivered on the positive lead (duty cycle). A steady 12vdc
and you get max air flow/speed? The tach lead is registered by
something on the mobo, and otherwise has nothing to do with the
rotation of the blades?
This is usually the case, yes. Visualize PWM...perform
an integration
over several cycles, to take the area under the curve. The energy
delivered to the load varies with the pulse width. But the switching
devices (usually MOSFETs) are going from cutoff to saturation and back,
never operating in active mode for more than a few microseconds at a
time, and so dissipating very little power, thus running VERY efficiently.
Many Sun systems drive their fans this way. Actually many high-end
server systems from many manufacturers drive their fans this way, as
does some PC hardware. It's very common.
-Dave
4566
days inactive
4566
days old
3 comments
4 participants
participants (4)
-
dave.g4ugm＠gmail.com -
mcguire＠neurotica.com -
rampaginggreenhulk＠yahoo.com -
tothwolf＠concentric.net