Re-wiring PWM fan

[streetfighter 2]

So I recently pulled an awesome looking PWM fan from an old (and busted) LGA775 based Dell Dimension desktop.

Fan Specs (from retail version):

Model | 4715KL-04W-B56-PWM
Voltage | 12VDC
Current | 1.30A
Fan Size | 120mm x 38mm
RPM | 3600 RPM
Air Flow | 129.90 CFM
Noise Level | 50.0 dBA
Power Connector | 4-pin PWM

Here's the one I have, the Dell OEM version, which has a fantastically useless proprietary Dell connector. :shadedshu

So here's my problem:
I know I have to hack off the useless Dell proprietary connector, but what I want to know is can I power the +12V and GND from a 4-pin molex and then just wire the fan's sense and PWM to the mobo's PWM fan header. Here's the proposed wiring:

I know some people might think this is stupid, but I don't want to be pushing 1.3A through the CPU fan header as I'm already using all the other fan headers (on a very very very cheap motherboard ). I also don't plan on running the fan at full speed (50dBA ) 24/7 but I do want to control it manually in speedfan. Also I can't think of any reason why this wouldn't work.

 

[mlee49]

Sure, do it.

Repost your results.

 

[streetfighter 2]

I got around to wiring it up. Gonna give it a try in a second.

I'm not looking forward to modifying the fan bracket to go from a 25mm wide to a 38mm wide fan. Oy!

EDIT: Tests were successful. It works great! It's incredibly loud at 100% as expected. I accidentally dropped it during testing and it cut the tips of my fingers up pretty good . I've now got it running but with two fan guards Both SpeedFan and the BIOS have no problems controlling it.

 

[The Von Matrices]

It will work fine. I'm using a PWM splitter (see link below) in my PC to control three fans with one PWM signal, and it's wired identically to what you say.

http://www.frozencpu.com/products/10350/cab-183/Akasa_PWM_Splitter_-_Smart_Fan_Cable_AK-CB002.html

 

[streetfighter 2]

Unfortunately for me being a catastrophic idiot, the fan doesn't fit ; It hits one of my RAM heatsinks. I'm going to sell the fan to one of my friends who has a Scythe Mugen 2 Rev B. Combined with the ridiculous fan I salvaged, I'm betting it will beat the NH-D14, but it might be louder .

It's kinda amazing that Dell was using such high quality fans in their desktops, but since it's the only fan in the box, I guess it makes sense.

Oh and a warning to people as careless as me: 3700 RPM makes quick work of fingers.

It will work fine. I'm using a PWM splitter (see link below) in my PC to control three fans with one PWM signal, and it's wired identically to what you say.

http://www.frozencpu.com/products/10..._AK-CB002.html

It works so amazingly well I'm kinda curious why the top end PWM fans don't come with adapters similar to the one you linked and the one I built.

Fans like the one I salvaged already cost around $30; it wouldn't cost them much more to throw in a nice PWM adapter because the fans can eat more power than some mobo headers provide. IIRC including power adapters was pretty typical for fans boxed with heatsinks prior to the standardization of PWM support.

 

[mlee49]

I'm with ya, PWM support should be on all fans and more motherboard manufactures need to provide more than 1/2 4pin headers.

 

[streetfighter 2]

I'm with ya, PWM support should be on all fans and more motherboard manufactures need to provide more than 1/2 4pin headers.

Currently newegg is only selling seven 120mm PWM fans , whereas they're selling around 160 regular 120mm fans. Heatsinks are really driving the demand for PWM fans and I have no doubt they will eventually dominate.

 

[The Von Matrices]

Currently newegg is only selling seven 120mm PWM fans , whereas they're selling around 160 regular 120mm fans. Heatsinks are really driving the demand for PWM fans and I have no doubt they will eventually dominate.

I wish there were more PWM models as well, but I don't see it occurring on any large scale anytime soon due to their increased price over traditional fans. I had to go to frozencpu to find the one PWM fan with white LEDs (Enermax cluster), while there a about a dozen of non-PWM white LED fans on Newegg. The great thing about PWM LED fans is that the fan speed and the brightness of the LEDs are uncorrelated on a PWM fan whereas on a traditional fan controlling the fan speed by lowering the voltage also dims the LEDs.

Regarding the motherboard headers, while I would like to see more PWM headers, I think it's more important that the manufacturers implement proper fan control in the BIOS for all headers before we start talking about adding more PWM headers. It's hard to find a board that allows control of more than the CPU fan and one additional header. Still, I don't think it's much of as issue as long as you have a splitter like the one I mentioned because I am not concerned about making cooling zones in the case with each fan at a different speed.

 

[entropy13]

Currently newegg is only selling seven 120mm PWM fans , whereas they're selling around 160 regular 120mm fans. Heatsinks are really driving the demand for PWM fans and I have no doubt they will eventually dominate.

Shame you don't have Deepcool in Newegg. They sell the Ice Blade PWM fans here (they're the fans that come with the Ice Blade Pro), and they include a 4-pin Y-splitter for free

 

[Bo$$]

I got around to wiring it up. Gonna give it a try in a second.

I'm not looking forward to modifying the fan bracket to go from a 25mm wide to a 38mm wide fan. Oy!

EDIT: Tests were successful. It works great! It's incredibly loud at 100% as expected. I accidentally dropped it during testing and it cut the tips of my fingers up pretty good . I've now got it running but with two fan guards Both SpeedFan and the BIOS have no problems controlling it.

your avvy is perfect

 

[bnborg]

I see nine 120mm PWM fans at NewEgg now (plus four out of stock).

Artic Cooling makes gangable PWM fans. Similar wiring to OP.

I like PWM fans but I REALLY detest two wire fans with only a molex connector.

 

[SaiZo]

Hey, I actually connected a potentiometer between the blue and yellow wire on my fan. Other than that, it is like you have - connected with a molex plug. But I assume, this one you are using is for the CPU??

 

[bnborg]

Hey, I actually connected a potentiometer between the blue and yellow wire on my fan. Other than that, it is like you have - connected with a molex plug. But I assume, this one you are using is for the CPU??

Me? I am using them for case fans. But just because your motherboard has a four pin sysfan connector it doesn't mean it supports PWM. The instructions with the AC gangable PWM fans show how to gang it to your CPU fan.

 

[streetfighter 2]

Rosewill sells a cheaper version of the PWM splitter that The Von Matrices linked.
Rosewill 12" PWM Splitter Model RCW-FPS-401

Hey, I actually connected a potentiometer between the blue and yellow wire on my fan.

You put a pot in between the +12V and PWM control wire? Why would you do that?

If it's running in parallel with the fan I can't see it doing anything but generating heat and possibly damaging the PWM circuit.

 

[The Von Matrices]

Rosewill sells a cheaper version of the PWM splitter that The Von Matrices linked.
Rosewill 12" PWM Splitter Model RCW-FPS-401

Wow, thanks for the find. If I ever need another, that will save me a little.

 

[The Von Matrices]

Hey, I actually connected a potentiometer between the blue and yellow wire on my fan. Other than that, it is like you have - connected with a molex plug. But I assume, this one you are using is for the CPU??

That configuration makes no sense. Usually, the yellow wire is an RPM sensor, and putting a potentiometer on that would just skew the RPM readings.

If the yellow wire is a +12V PWM line, that makes no sense either. The point of a PWM fan is to send 12V pulses of varying width to the fan thus controlling its current directly rather than traditional control with relies on adjusting voltage to control current. In this situation, your potentiometer is acting as a resistor between the +12V PWM and +12V constant lines, which means it either does nothing (when PWM pulse is +12V) or acts as a resistor wasting heat (when PWM pulse is 0V), but does nothing to control the voltage and thus the fan speed.

 

[mlee49]

That configuration makes no sense. Usually, the yellow wire is an RPM sensor, and putting a potentiometer on that would just skew the RPM readings.

If the yellow wire is a +12V PWM line, that makes no sense either. The point of a PWM fan is to send 12V pulses of varying length to the fan thus controlling its current directly rather than traditional control with relies on adjusting voltage to control current. In this situation, your potentiometer is acting as a resistor between the +12V PWM and +12V constant lines, which means it either does nothing (when PWM pulse is +12V) or acts as a resistor wasting heat (when PWM pulse is 0V), but does nothing to control the voltage and thus the fan speed.

Actually, it's widths. Otherwise it would be called pulse length modulation or PLM.

 

[SaiZo]

That configuration makes no sense. Usually, the yellow wire is an RPM sensor, and putting a potentiometer on that would just skew the RPM readings.

If the yellow wire is a +12V PWM line, that makes no sense either. The point of a PWM fan is to send 12V pulses of varying width to the fan thus controlling its current directly rather than traditional control with relies on adjusting voltage to control current. In this situation, your potentiometer is acting as a resistor between the +12V PWM and +12V constant lines, which means it either does nothing (when PWM pulse is +12V) or acts as a resistor wasting heat (when PWM pulse is 0V), but does nothing to control the voltage and thus the fan speed.

Did you not understand what I did type? I said, it is connected between the blue and yellow. And it actually works. If you do not believe me, test it yourself.

 

[The Von Matrices]

Did you not understand what I did type? I said, it is connected between the blue and yellow. And it actually works. If you do not believe me, test it yourself.

I don't doubt that the fan spins, but does the potentiometer actually control anything?