• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

Question Could a Fan Hub fry my motherboard ?

Apr 7, 2024
15
2
15
Hey guys
I bought an Arctic Fan Hub but I saw a comment fromsomeone saying that connecting sata and fan header (which are the hub connectors ) at the same time could cause problems.
is that true?
 
Hey guys
I bought an Arctic Fan Hub however I saw a comment of someone saying that connecting sata and fan header (which are the hub connectors ) at the same time could cause problems
is that true?
A hub can fry your mobo, just like anything else can. The arctic one you linked gets power from the PSU though, the 4pins are just used for the pwm signal. If it's a newly built pc did you install the cpu in the socket properly?
 
Last edited:
You do NOT have a problem, so go ahead and use that Hub.

There are old stories and misinformation about these things, so I'll try to clarify a little. At the root is the general rule that any standard mobo fan header can supply 12 VDC power at up to 1.0 A total load for all devices attached to that one header.

There are two different types of device to connect several fans to a single header, but makers and sellers very often use their labels as if they are the same and interchangeable, and they are NOT. My use of the terms is that a SPLITTER is the simpler device that connects all its fans in parallel to the header. So it has one input (female) connection to a mobo header, and two or more outputs (male with pins) to plug in fans. ALL of the fan power must come from the header, so that 1.0 A limit is important. A HUB is similar but differs in one very important item. It also has a THIRD connection to a SATA or 4-pin Molex output from the PSU, and ALL fan power comes from the PSU. It sends NO power from the host header to any fan, so the header Amp limit does not apply. (There is a different limit here, imposed by the contacts in the connection to the PSU. For a SATA connector source the limit is 4.5 A total load; for a Molex source it is higher. Most fans consume less than 0.25 A max, so this is not a practical limit - that's about 18 or 20 fans!) The Hub does get the PWM control signal from the host header to share to all its fans, but the circuits involved draw VERY little current and will NOT overload the header.

Note that you cannot tell which device is which by simple physical appearance. Either of these device types MAY appear as a collection of cable "arms", as a small circuit board with edge connectors, or as a closed box with ports recessed inside holes. What IS distinctive is the presence or absence of that THIRD connection to a PSU power output.

So one major source of mixed-up info happens if the real device is a SPLITTER but you treat it as if it were a HUB that gets power elsewhere. This CAN lead to overloading the mobo header.

An unlikely problem NOW is poor design of what IS a Hub. In some early designs the power line from the host header (Pin #2) WAS connected to the proper external power line from the PSU. These days that is never done. But on those old designs, if the host header was operating in the old Voltage Control Mode and trying to send out less than 12 VDC to run the fan slower, that could cause a problem for the header.

There is a limit on HUB use that is NOT a danger. The way a HUB works is that it gets a fixed power supply of 12 VDC from the PSU to send to all fans via Pin #2. Then it also gets the PWM signal from the host header and sends that, too, to all fans via Pin #4. This is the way that PWM fans work and they use the PWM signal to set their own speed. But that also means that you can NOT control the speed of an older 3-pin fan with this set of signals. Such a fan would always run full speed when powered by a HUB. It means also that the host header MUST be configured to operate in PWM Mode so that it DOES send out the PWM signal on Pin #4.

You CAN connect several older 3-pin fans to a single header that is operating in Voltage Control Mode (aka DC Mode) using a SPLITTER, because the power sent out by that device is whatever the Voltage is from the header Pin #2, and in that MODE of control the header VARIES the Voltage to change fan speed. In this case the 1.0 A limit appplies.

A side note. All fans generate a speed signal (consisting of a series of pulses 5 VDC high, 2 pulses per evolution) sent back to the host header via Pin #3. The header counts these to display a speed, but does NOT use that info to control speed. However, is DOES use that to monitor the fan for FAILURE - no speed signal. The header can deal with a speed pulse train arriving from only ONE fan. So any Splitter or Hub will send back the speed of only ONE fan and ignore all the others on it. You will never "see" the speeds of those "other" fans anywhere, and they can NOT be monitored for failure. On a unit that looks like a collection of cable "arms", only one male output will have all 4 pins; all other outputs will be missing Pin #3 so they cannot send back that speed signal. On other physical designs ONE output will be marked in some way to indicate this one WILL send back the speed signal. So in using these devices, make sure the one send-speed-back output DOES have a fan plugged in.
 
Last edited: