Best to run case fans off the MB or PS?

[craftech]

I am wondering whether there is any argument for or against using the fan pin headers on the motherboard for the case fans or whether it is less strain on the motherboard to avoid that and connect them directly to the power supply. I have done both for years and could care less about monitoring the case fan speed, but I thought I'd ask the question of people who build a lot more rigs than I do.

Thanks for the input.

John

 

[mb103051]

anytime you can run fans from the p/s directly, thats a better idea..it takes some of the load off the motherboard..its not real uncommon to see fan headers burnt on boards where the fan has malfunctioned ...
I always run a fan controller where i have control of all my fans all the time....

 

[kiriakos]

Any fan 80 - 90mm up to 1600 rpm or 120-140mm up to 1000 rpm ,
can safely connected to motherboard headers ...

This is the absolute safe limits ,
anything above in rpm VS dimensions,
it must be connected to molex plugs.

 

[MagnusTheBrewer]

I think it's more a matter of cable management preference than any 'strain' on the MB.

 

[fffblackmage]

Any fan 80 - 90mm up to 1600 rpm or 120-140mm up to 1000 rpm ,
can safely connected to motherboard headers ...

This is the absolute safe limits ,
anything above in rpm VS dimensions,
it must be connected to molex plugs.

This is the first time I've ever heard of something like that...

 

[craftech]

It seems that there is no clear answer. Fans do tend to go after awhile, especially sleeve bearing fans; but I thought it was just mechanical problems. Didn't know the motors sometimes burn up putting a strain on the circuits.

John

 

[AstroGuardian]

Any fan 80 - 90mm up to 1600 rpm or 120-140mm up to 1000 rpm ,
can safely connected to motherboard headers ...

This is the absolute safe limits ,
anything above in rpm VS dimensions,
it must be connected to molex plugs.

This is one of those "I wanna look smart" who fail to do the job.

Fans consume so little power that it's strain on the motherboard is insignificant. For example a standard 12cm fan works @ 4-14 volts consuming 0,1amps of current which is in worst case scenario 12v x 0,1a = 1,2 watts.

If this amount of power consumption is not a significant strain on the mobo.

 

[craftech]

This is one of those "I wanna look smart" who fail to do the job.

Fans consume so little power that it's strain on the motherboard is insignificant. For example a standard 12cm fan works @ 4-14 volts consuming 0,1amps of current which is in worst case scenario 12v x 0,1a = 1,2 watts.

If this amount of power consumption is not a significant strain on the mobo.

But what about the poster above who stated?:

..its not real uncommon to see fan headers burnt on boards where the fan has malfunctioned ...

If it malfunctions and is connected to the MB, would it not make sense that it could burn the fan headers as opposed to a molex plug that can easily be replaced and is in fact thicker and stronger to begin with?

John

 

[AstroGuardian]

But what about the poster above who stated?:

..its not real uncommon to see fan headers burnt on boards where the fan has malfunctioned ...

If it malfunctions and is connected to the MB, would it not make sense that it could burn the fan headers as opposed to a molex plug that can easily be replaced and is in fact thicker and stronger to begin with?

John

Well yes of course. In theory it can happen for fan headers to burn even though I haven't seen such case within hundreds of computers i have assembled and serviced.
But you see, brand computers often don't have an extra free molex connector and you will have to attach it to the HDD's or OD's power cable which carries the same risk if the fan fails. I am not exactly sure but i think i have read about poor quality fans that can lower the current's quality coming to the HDD or OD.
If you do actually have a spare molex cable and you don't need the fan's RPM displayed then the molex connected fan is better option but IMO only for a small margin.

In most cases when the fan starts to fail it first starts to vibrate and make noise. So it's the first sign when the fan starts to fail and needs to be changed. By doing that you ensure that your fan will not burn your mobo headers and stuff.... right?

The thread poster can also check another topic: whether the connected fan causes current ripple or similar problems. I am sure i read something like that long time ago

 

[craftech]

This is interesting. I looked up the technical specs for my primary motherboard (Intel D975XBX2) and it stated the following in terms of current allowance for the fan headers:

Table 34. Fan Header Current Capability
Fan Connector Maximum Available Current
Processor fan 3.0 A
Front chassis fan 1.5 A
Rear chassis fan 1.5 A
Auxiliary rear fan 3.0 A
MCH fan (optional) 1.5 A

http://downloadmirror.intel.com/15065/eng/D975XBX2_TechProdSpec.pdf

Page 75.

John

 

[AstroGuardian]

Well as you can see a normal 12cm fan is only 1/15 of specs. So as i said, the strain to the mobo is insignificant.

 

[craftech]

Well as you can see a normal 12cm fan is only 1/15 of specs. So as i said, the strain to the mobo is insignificant.

Indeed. I see that now. Thanks a lot for all your input on this.

Regards,

John

 

[::2dfx::]

Any fan 80 - 90mm up to 1600 rpm or 120-140mm up to 1000 rpm ,
can safely connected to motherboard headers ...

This is the absolute safe limits ,
anything above in rpm VS dimensions,
it must be connected to molex plugs.

This is erroneous. It purely depends on the current limits/header that your motherboard provides. Consult your motherboard's manual for exact values, then consult the fan itself to see how much current it draws.