White and Yellow wires in NMB-MAT fan

[mxmone]

Hello!

I can't handle this may self so maybe someone in forum may know the answer. Even uncle Google don't know how to help me

I have an old fan installed in Power Supply inside my NetApp DS14mk2 at home This fan (two fans) is very noisy so i want replace it with quite one (two of them). I bought a PWM fan and I've try to connect wires to power supply board but board doesn't have a PWM standard (maybe it is too old, who knows).

My point is what are white and yellow cables for? They are (I think) responsible for rotating speed but in what standard?
Black is GND and red is +12V.
And what a hell is: "Lead Wire : UL1007, AWG22" in datasheet?
Fan is: NMB-MAT, model: BL4447-04W-B49
Don't have datasheet? Here you go: http://www.nmbtc.com/pdf/catalogs/Fan_a ... g_Full.pdf (page 42)
Picture? Here: http://www.pchub.com/uph/laptop/640-635 ... r-Fan.html (not the same sticker but color wires are the same)

Please help!

P.S. Sorry for my english my friends

[Karlsweldt]

Check this link concerning small fan power leads.
The color code may not be the same among brands.. but the power connection is 'standardized'. As noted on the page, you can use a 3-wire fan plug on a 4-pin connector, and also the other way around. The plug and socket are polarized to avoid damage. A 3-wire fan has only the power leads and a sensor lead to a "Hall Sensor" which monitors fan RPM. The #4 wire is a control lead to a speed regulator on the fan motor itself.
The "Hall Sensor" is no more than a tiny magnetic sensor, like a relay coil.. as a magnet on the motor armature passes by it, this causes a tiny electrical pulse. Same as used on floppy drive motors, and in most automobile ignition timing setups.. as well as for the anti-lock braking systems!

[mxmone]

Nope. I've try with 3-pin and 4-pin PWM configurations and nothing. Except when i was trying to connect white wire (from NMB) to green or blue (on my PWM) the fan STOPS !!! And i don't see any speed control in any configuration. Now that?

Of course my NetApp screaming FAN ERROR!!

And more. When i use only ONE fan (original NMB) and i'm connecting a second white wire to first white wire - it works with no error! Even if there is no second fan at all.

[Karlsweldt]

Do the voltage and current specs match between fans? Some fan circuits have a "soft start" feature, no more than a thermistor that lowers its resistance to current flow as temperature rises. If the current flow is not identical between old and new fans, then problems as to monitor alarms can arise.
With a typical computer or NAS setup, ensure the CPU fan is connected to its proper header.. and other fans to their respective headers. For case fans, best if connected directly to the PSU. This avoids loading the +12 volt supply to the motherboard circuits. As to "fan error", look at the specs page for the old fan. It may have had two tiny magnets to energize the Hall sensor, while the new fan may have only one. Half the RPM count would be a result!
If using a 3-wire fan on a 4-pin connector, and you are certain the fan is working properly, then set the BIOS page concerning fan for that specific fan's speed to "ignore". Giving up performance of cooling to quieter operation has inherent problems. The best fan for computer use is a "squirrel-cage" type. Highest air movement, lowest noise. All small fans will create a ruckus, due to the air stream (wave front) leaving the blades and hitting the support struts in pulses. No support struts in the air stream path of "squirrel-cage" designs!

[mxmone]

OK. This is not a computer with motherboard and BIOS. I'm telling here about Power Supply Unit. The fans are in PSU. There are 2 PSU in whole case. Of couse there is a small mainboard but i can't change here any BIOS options. This is a disk array. PSU has his own logic, and board have 3 LEDs. After system stats up, every thing is OK. But after about 8-10s, the fan LED lights up in PSU and whole case beeps an error.
This is NOT PWM. But what is that? Most important thing is what a white wire do? Between black and white wire is about 3V and 1mA on working NMB-MAT fan. But on white wire it self its only 0.5mA.
Specs are in first post on page 42 in datasheet but i don't see there any useful informations.

This is inside the NMB-MAT fan:

[evasive]

What is the reason you didn't replace the fans with new ones of the very same model? Trying to save money? Trying to silence an appliance for living room use?

PWM is for steering the fan, what you are needing is monitoring the fan... More than likely the yellow or white wire is sending back some sort of pulse signal to let the PSU know the fan is still running. The question is if that is an active signal or a passive signal and if passive if it's resistive or capacative sensor based. You explained this isn't a standard PC, why try using a standard PC component then?

[mxmone]

Well, this an old PSU. I don't have access to replacement QUIET fan!!!
Now, i know what a white wire do. He is responsible for locked rotor:
"A locked rotor is also a type of “sensor output” that measures when the fan has completely stopped or “locked”. It sends a signal, frequently called an alarm signal, at either high or low voltage when the rotor locks. If the fan starts spinning again, the alarm signal condition will go away. This is known as a non-latching locked rotor signal, which is standard with NMB fans."

Yellow one is tach output (sensor output):
"A tach output or “sensor output” as shown in the catalog indicates the speed of the fan at different operating levels. Its purpose is to identify when the fan drops below a certain RPM, and to identify a potential problem with airflow. A tach output fan will always have at least 3 wire leads.
The standard tach out feature is an open collector, 2 pulses per revolution, square wave."

So i don't think any at this forum can handle a transform this signals to standard PWM fan Sorry for trouble...

[evasive]

Yep, time to find an electronics forum for that. Still, the question remains why you want to use superquiet fans in something that usually found in server rooms only where sound level doesn't matter...

[mxmone]

I want to use it at home for tests But the noise is too loud. Thats 4 fans... even in slow mode they are too loud

[Karlsweldt]

Here is a link about the specs for PWM fans.. http://www.formfactors.org/developer%5C ... Public.pdf
and the specs page link for the DS14.. http://www.netapp.com/us/products/stora ... specs.html (about 1/2 way down the page).
And also a link about NMB fan specs.. http://www.nmbtc.com/dc-fans/engineerin ... tions.html . The page also has related links to "white papers" and other fan specs, such as sensors and speed control.
That PSU in your unit is rated at less than 50 dB sound level, normally.. so either the original fans have been replaced at some time, or they have impaired air flow which can raise noise levels.
Measuring the output from those fan sensors is not accurate unless using an oscilloscope or frequency counter. The output is a rapid DC pulse, roughly a square wave form. A DC measurement would only be an "average" level.
While some models of the NetApp line do have a BIOS feature, yours may not.. but would still rely on a ROM program for operation. And a very basic mainboard setup, such as the Arduino® micro boards.
Have not worked with network setups in years, but remember the hassles with working on the old HP Netserver Pro systems.. the "beasties" that howl! The only way to access the BIOS on them was with a special setup/service diskette.