PWM fans

[kinku]

Hi All,
I have an old LX 195 hP media smart server which was working fine until recently.I opened the case and found out that the cooling fan is not working.It is a 70mm AVC fan with 0.17 Amp and 4 pin(AVC DA07015B12L).
What is the importance of current ratings in PWM fans? What if I get a higher current(wattage) rated PWM fan ,will that damage my motherboard by any chance?
Will the PWM fans will adjust for the current they get from motherboard or will they try to spin using their full rated current when needed and damage the fan driver,in my case in the motherboard?
I can not find the same fan except at (http://www.cwc-group.com/70mm.html) about whom I do not know anything about?And surprisingly they listed the AVC DA07015B12L in their website after me checking with them about another fan model will match mine(sneaky)...
Please help........

 

[Bobnova]

Too high a draw will fry the motherboard.
The PWM function is generally based on CPU temperature rather than current draw.

Most motherboards can do at least half an amp, there ought to be a spec for yours somewhere.
Alternatively you can power the fan directly off 12 V from the PSU and just run the PWM and TACH lines to the motherboard.

 

[kinku]

Thank you Bob.That was a fast reply and raised my hopes.
Please have a close look of LX195 motherboard here
http://i280.photobucket.com/albums/kk189/donavonwest/homeserverhacks/hp-lx195-mainboard.jpg
It is from a manufacturer called Boxster!!!
So do you think if I use a 0.25-0.35 Amp fan it should be ok?
I thought about getting a direct power connection to fan with leaving PWM and tach sensor input intact as a last resort.Do you think it will be ok to do that?

 

[ehume]

I thought about getting a direct power connection to fan with leaving PWM and tach sensor input intact as a last resort.Do you think it will be ok to do that?

Not only will it be all right, it will solve your problem and allow you to use any PWM fan, regardless of how much power you draw. I use this one. A prettier one is this. They are "PWM Splitters" and power more than one fan, but you can use one to power a single fan.

To recap: 12v and ground wires come from the PSU. The PWM control signal comes from the motherboard. The RPM reporting signal goes from one of the fans (usually labeled "CPU") to the motherboard. Your motherboard is safe that way.

 

[kinku]

Thank you ehume.I wish my motherboard has a regular power supply.But unfortunately this board has power-supply similar to one in Laptop. Please refer to the picture.So I have to solder the power supply out from the DC terminal socket. That is why I am looking at higher power rated fans.
Thank you for your suggestions.

 

[Bobnova]

Looked up some pics, you're right, no spare power to be had there!

Based on the only trace I can see headed towards the fan connector, I would guess that a 0.4a-0.5a fan will be OK, especially as it likely won't be running full blast all the time, and the fan blows air right on it's power feed trace). That's not a guarantee, as I don't know what feeds that trace.

This fan, if the specs are accurate (decent chance, I'd say) should be about perfect:
http://www.ebay.com/itm/Superred-CH...?pt=US_CPU_Fans_Heatsinks&hash=item1c2c905851
it's coming from China, which isn't exactly a good thing, but if it's even close to what it says it is it'll do the job nicely.

Newegg has this character, which should do nicely as well:
http://www.newegg.com/Product/Product.aspx?Item=N82E16835119150

 

[kinku]

Thank you Bob.I do not know anything about how a 4 wire fan works. But from the discussion it seems like the fan derives it powersupply directly from powersource and the PWM and tach reading regulate the speed. So the power of the fan is not through a driver component(like a transistor or mosfet ),is that what you meant?
If that is the case I think a 100 milliamp more than original rating should not hurt the board since it is just taken from main powersupply and tracings should be able to handle milliamp range currents. Please correct me if I am wrong.
If that is the case what about Delta AFB0712HHB (http://www.delta.com.tw/product/cp/dcfans/download/pdf/AFB/AFB70x70x15mm.pdf)?
Thanks again for all your help and time to look through the motherboard schematics.

 

[Bobnova]

PWM fans take power from the 12v pin in the 4p connector and ground to the GND pin just like a normal fan, as well as reporting the RPM on the tach pin just like normal fans.
The difference is that the fourth wire sends digital data to the fan, telling it what percentage of full speed to run.
A normal fan fed 12v will run full speed at all times, a PWM fan fed 12v and receiving a 50% PWM signal will run at 50% speed.
This is nice, as it's far easier (and more efficient) to send a very small (5ma max) PWM signal than it is to reduce the voltage to the fan.

Essentially it's a speed control, much like the heat fan in your car. Most modern heater fans use PWM control actually. No charge for that bit of trivia


If the above was confusing, it's probably due to the Vodka
There's another, good, explanation to be found here: http://www.overclockers.com/pwm-fan-controller

 

[kinku]

So this is what I understand.
PWM signal is what the motherboard controls but PWM is not the driving powersource for the fan.Fan takes the power from main source(a trace directly from power-supply and to ground like any other fan), but how fast it need to spin is decided by the control circuit in motherboard,through tach sensor input and PWM circuitory output to fan.
So as long as I am not connecting power hungry fan of say 0.5 amp the tracings in my motherboard should be able to carry the current needed to drive the fan even at 100% PWM signal driving the fan.Is it correct?
Thanks for the help and please correct me if I am wrong
I am getting little greedy and thinking of getting that 0.3 Amp fan from newegg with a better CFM than original one.

 

[inVain]

So this is what I understand.
PWM signal is what the motherboard controls but PWM is not the driving powersource for the fan.Fan takes the power from main source(a trace directly from power-supply and to ground like any other fan), but how fast it need to spin is decided by the control circuit in motherboard,through tach sensor input and PWM circuitory output to fan.
So as long as I am not connecting power hungry fan of say 0.5 amp the tracings in my motherboard should be able to carry the current needed to drive the fan even at 100% PWM signal driving the fan.Is it correct?
Thanks for the help and please correct me if I am wrong
I am getting little greedy and thinking of getting that 0.3 Amp fan from newegg with a better CFM than original one.

you got that right
just a minor correction though, the tach doesn't do anything with the controlling, it's only reporting the fan's RPM

and you can always feed the power directly from the PSU to the fan to secure your mobo header (but I'm afraid you need to do some extra wiring job to do)

 

[kinku]

Thank you all.

Thank you all.I should have asked you all a long time ago, instead of looking for an exact power and current rated fan across the globe!!!!
I appreciate all your help.
Long live overclockers community.

 

[kinku]

update

I bought Delta AFB0712HHB from ebay and used the speed fan to see how it work.
Please see the Delta fan specs here
http://www.delta.com.tw/product/cp/dcfans/download/pdf/AFB/AFB70x70x15mm.pdf
The fan seems working at 1950-1965RPM all the time. Maximum rated RPM is 4300.My CPU temperature is 45-48 Degree Celsius all the time. I am happy with it since it was at 65-66 before (when original fan was not working and the case in open position).
Is there anyway to bring CPU temperature to 35 ,by increasing RPM.I tried playing with speedfan 4.46 without much luck.
I can not see bios since LX195 has only a KVM port in motherboard and need an expensive adapter to see bios.
Hope someone can help.

 

[inVain]

first of all,
the delta wiring code typically is:
black = gnd
red = 12v
BLUE = tach
YELLOW = PWM

I strongly recommend to try short the yellow wire to supply gnd to see speed change and confirm it's the PWM wire

the minimum CPU temperature will be dependent to your room temperature and your case airflow.

the minimum RPM during operating will be dictated by your bios setting or your setting via speedfan (which unfortunately makes me out of luck, since I'm sucks on speedfan )

 

[inVain]

whoa....
that color scheme is sooo intel flavor

yes, it is what we expected.
since gnd mean 0 duty cycle, which equal to the lowest pwm position.
the fan should slow down to it's minimum speed.
so the pwm line for your fan is the blue wire.
you've the the right wire mate

and what kind of noise did you hear?
AFAIK, clicking is the most typical noise for this kind of pwm fans,
but since I played with some denkis, I've got some stranger noise like whining, and growling.
in this case, the fan appears to be need a special frequency on the pwm line.

but most of my deltas seem to be happy with any given frequency that I feed them
or maybe my deltas just can't spin low enough to let me hear another noise beside the "wooosssshhh"


edit:
for some reasons the post arrangement seems to be wrong on my PC,
just in case that this post made some confusion, this is the post that I intend to replying

Thank you.But my delta Fan has black(1),yellow(2) ,green (3) and blue (4) in that order from right to left(pin configuration in bracket).
What change should I expect if I short ground to PWM wire. I did try shorting black to blue and the span speed slowed down subjectively with some noise .
Is that what expected?

 

[kinku]

Thank you.But my delta Fan has black(1),yellow(2) ,green (3) and blue (4) in that order from right to left(pin configuration in bracket).
What change should I expect if I short ground to PWM wire. I did try shorting black to blue and the span speed slowed down subjectively with some noise .
Is that what expected?

 

[kinku]

I did something creative while awaiting your reply. I took the fan off from cabinet and the let the PC start and through remote access. I checked the speed of fan using speedfan.
I was excited to see as the temperature of CPU started shooting up(I used Prime 95 to stress) the fan spun up to 3800 RPM and up( then I turned the server off) and I am sure now it works.In fact I think it works better than original AVC since the temperature is steady and reading 33degree celsius for core HDD and 50 for CPU using speedfan. CPUID showing only 45 degree celsius.Don't know which one is accurate?
InVain the noise was slight click when I shorted but not otherwise fan was still spinning and seems half of original speed (subjective)hopefully it works for a long time.Thank you for your whole heart-ed support

 

[inVain]

my pleasure....
glad to see you made it and love your fan

yea, the software reading sometimes get and rather confusing.
personally, I rather the CPUID (h/w monitor) since it shows me the reading directly from the cpu and also via the mobo sensor

the noise seems typical to me, but then again....
the noise is really such a subjective matter