Can I fix this fan this way?

[jojo]

Hi all....

I have a poweredge 1300 (old one) that has a dead fan.
I purchased a new fan (dc 12v) and installed it into the
dell specific casing.

The worry I have is that the new fan has 2 wires (black and red) coming from
it.
The old one had 3. (black, red and white)

The guy at radio shack (only thing available to me) said the white wire was
probably a sensor
for the server. He said to simply splice the old fan plug red/red
black/black to the
new fan and plug it in, letting the white wire do nothing.

I have done that and it makes perfect sense, but I wanted to get other
opinions of if this is safe and will
work before I do it.

Thanks,
jojo

 

[kony]

Hi all....

I have a poweredge 1300 (old one) that has a dead fan.
I purchased a new fan (dc 12v) and installed it into the
dell specific casing.

The worry I have is that the new fan has 2 wires (black and red) coming from
it.
The old one had 3. (black, red and white)

The guy at radio shack (only thing available to me) said the white wire was
probably a sensor
for the server. He said to simply splice the old fan plug red/red
black/black to the
new fan and plug it in, letting the white wire do nothing.

I have done that and it makes perfect sense, but I wanted to get other
opinions of if this is safe and will
work before I do it.

Yes the Radio Shack guy was right, usually. It is by far
most common to have a fan arranged this way but with a
proprietary system they could wire them any way they like.
They "should" use a different pin header and socket for
proprietary wiring but anything is possible.

What you could do is take a couple of jumper leads (pair of
wires with insulation stripped off each end) and connect the
12V and ground from a power supply to that red and black
wire- at which point the fan should turn.

The other consideration is if the system had a fan-failure
alarm or auto-shutdown feature. If so, it would probably
(but not necessarily) use the sensor lead to detect the fan
failures. If this is the case then the issue becomes
whether the bios has a setting to disable the warning or
shutdown feature.

Radio Shack is a quick source for a fan if you've no other
suppliers nearby but IMO, their fans aren't the greatest
quality nor prices. Ideally a fan for your purpose would've
been sourced online.

 

[jojo]

Thank you for answering my questions!

Yes the Radio Shack guy was right, usually. It is by far
most common to have a fan arranged this way but with a
proprietary system they could wire them any way they like.
They "should" use a different pin header and socket for
proprietary wiring but anything is possible.

What is worse case if it's backward?

What you could do is take a couple of jumper leads (pair of
wires with insulation stripped off each end) and connect the
12V and ground from a power supply to that red and black
wire- at which point the fan should turn.

Don't have any way to test this....other than the server itself.

The other consideration is if the system had a fan-failure
alarm or auto-shutdown feature. If so, it would probably
(but not necessarily) use the sensor lead to detect the fan
failures. If this is the case then the issue becomes
whether the bios has a setting to disable the warning or
shutdown feature.

won't the fan work, but the system continue to tell me it does not?

Radio Shack is a quick source for a fan if you've no other
suppliers nearby but IMO, their fans aren't the greatest
quality nor prices. Ideally a fan for your purpose would've
been sourced online.

I simply could not figure out what to order online by myself. This is an old
unit
and dell does not seem to have a part specifically for it anymore...

Thanks,
jojo

 

[kony]

Thank you for answering my questions!


What is worse case if it's backward?

There is not only "backward", 3 pins is more than 2
combinations.

If the power leads were reversed the fan simply wouldn't
spin... that is, "most" fans are polarity protected and I
will assume the Radio Shack fan is also. In that case no
problem, just reverse the leads. If the power lead were
connected to the sensor pin, should just give an invalid RPM
reading.

The sensor pin on the other hand, I don't know for certain
but suspect it could be damaging to have 5V grounding
through it, if you connected the fan ground lead to the
sensor pin by mistake.

The simplest solution is to simply use a multimeter to check
voltage (for the power pin) and continuity (between the
suspected ground pin and system ground).

Don't have any way to test this....other than the server itself.

Well there you are, it has a power supply in it and every
molex 4 pin connector has 12V and ground on it.

won't the fan work, but the system continue to tell me it does not?

yes the fan would spin, but as mentioned previously, if the
system doesn't "know" it's spinning it will matter what
provisions there were for a failed fan.... failed in the
logical sense that the motherboard has no reason to believe
it's spinning without an RPM or rotor-lock signal going back
on the 3rd pin.

I simply could not figure out what to order online by myself. This is an old
unit
and dell does not seem to have a part specifically for it anymore...

Fans are spec'd by diameter, thickness, voltage and current
(and the resultant RPM of that voltage and current).

For example, if you had a 92mm x 25mm fan powered by 12V
(typically the old fan would state it on the label) you
would simply choose a popular brand of fan with same
connector and the 3rd lead for the RPM... previously I
mentioned rotor-lock but if forced to guess between the two
a non-industrial computer will usually have RPM rather than
rotor-lock feature and a PC almost always the RPM.

 

[jojo]

well, I spliced and plugged it in.
works fine, although as we discussed, the sensor does not
acknowledge that the fan is there.
Anything I should worry about at this point?
jojo

 

[kony]

well, I spliced and plugged it in.
works fine, although as we discussed, the sensor does not
acknowledge that the fan is there.
Anything I should worry about at this point?
jojo

No, so long as the fan doesn't draw more current than the
board can provide long term... which is usually in the
neighborhood of 6W for all fan headers combined but in
practice can vary from one header to the next and is usually
a little higher. I don't think Radio Shack even carries
fans using this much current in a PC/server size so you
should be ok.

 

[jojo]

Thanks for all of your input on this!
I was desperate and nervous!
jojo