5 Replies Latest reply on Sep 23, 2016 11:16 AM by Intel Corporation

    CPU fan runs faster than front/rear

    paramountain

      On a DH67CL, I connected a 3-pin fan to the motherboard CPU fan port. It runs at full speed (1300 rpm) according to BIOS. Then I connected different 3-pin fans to the front and rear fan ports and both fans ran at 800 rpm, much less than their top speed of 1200. Then I disconnected front and rear fans from the motherboard and connected them via a Molex connector and they ran faster (one has a slight whine which becomes annoying at top speed, so it's easy to determine if it's running flat-out). Is it normal for 3-pin fans connected to the CPU fan port to run at full speed while the other ports throttle them back?

       

      In case anyone cares, the reason I am asking this question now is that I moved the system to a new case. Before I had the front/rear fans connected via Molex, but this time I thought I'd try the motherboard ports, something I had not done before.

        • 1. Re: CPU fan runs faster than front/rear
          N.Scott.Pearson

          First, some background...

           

          • The chassis fan headers provide support for both 3-pin and 4-pin fans. When the board powers on, the fan control circuit for each header detects which type of fan is installed. If it is a 3-pin fan, the circuit controls the speed of the fan by varying the voltage provided to the fan. If it is a 4-pin fan, on the other hand, it includes built-in support for controlling the fan speed on its own and the circuit thus continually provides 12V. In this case, the 4th pin provides a pulse-width modulated (PWM) signal whose duty cycle tells the fan how fast to spin (the duty cycle of the signal indicates the percentage of full speed at which to operate).
          • The CPU fan header does not have a circuit that support 3-pin fans; it only has support for 4-pin fans. If you plug a 3-pin fan into this header, it will continually operate at full speed. Even though the 4th pin may be specifying that the fan should lower its speed, the 3-pin fan is not receiving this signal (nor does it have a capability to do anything with it anyway) and it is continually getting a full 12V signal, so at full speed it runs.
          • The motherboard contains an environmental monitoring and fan speed control capability. In Intel Desktop Boards previous to the 6 Series, this capability was provided by a separate IC. In the 6 Series and later boards, this capability is embedded into the Super I/O (SIO) IC (the same IC that provides support for such features as Consumer I/R, Serial Port(s), CMOS/Clock, Watchdog Timer, POST Code display (on select boards), etc. and etc.). The environmental monitoring capability exposes temperature, voltage and fan speed sensors. The fan speed control capability monitors these temperatures and, based upon programmed algorithms (thresholds, etc.), controls the speed of each fan.

           

          Ok, you should now understand why your 3-pin fan is running at full speed if plugged into the CPU header and that you should really be using a 4-pin fan in this header. At the same time, you should also understand why it is better to plug the fans into the motherboard rather than just into the power supply. The motherboard's fan speed control subsystem will minimize the system's overall acoustic signature by slowing fans when thermal conditions allow.

           

          Ok, that said, are there legitimate reasons for connecting the fans to the power supply and running them as full speed? Yes, there are. First of all, if you are overclocking the processor, its rate of temperature change will be higher than normal and the standard algorithms used by the fan speed control subsystem may not keep up with it, thereby allowing the processor to reach critical temperatures from time to time (even if only momentarily). Some people avoid this by locking the CPU fan at full speed. Secondly, when it comes to chassis fans, there are some pretty crappy 3-wire fans out there that, when their speed is changed, respond poorly acoustically. The best thing to do with them is throw them away and replace them with better ones (get 4-wire chassis fans, they are clearly superior). There is also the issue of psycho-acoustics. When fan speeds change, this is noticeable. The faster they change, the more noticeable it is. If it is noticeable, it can become irritating. For this reason, sensitive people like to lock their fans at full speed. Yes, this is louder, but, because the sound is not changing, you eventually don't notice it anymore.

           

          That's a good introduction and enough for now...

          ...S

          • 2. Re: CPU fan runs faster than front/rear
            paramountain

            Very interesting. I could tell empirically that there was a difference, but I did not know why. Thanks.

             

            This system has a (now discontinued) Noctua NH-C12P SE14 cooler with a 3-pin NF-P14 140mm fan. It is strange that Noctua would have included a 3-pin fan with a cooler, but many of Noctua's products are geared toward overclockers. It spins at 1300 rpm, but is quiet so I never minded it running flat-out. Temperatures stay in the 30s unless the processor is being hammered.

            • 3. Re: CPU fan runs faster than front/rear
              Intel Corporation
              This message was posted on behalf of Intel Corporation

              N.Scott.Pearson thank you very much for your kind input.

               

              paramountain, please refer to the previous post, I believe that it answers to your question; however, if you need further assistance let us know.

               

              Regards,

              Amy.

              • 4. Re: CPU fan runs faster than front/rear
                paramountain

                Yes, Scott answered everything in his usual thorough manner.

                • 5. Re: CPU fan runs faster than front/rear
                  Intel Corporation
                  This message was posted on behalf of Intel Corporation

                  Yes, indeed. 

                   

                  If there is something else we can help you with, please let us know.

                   

                  Regards,

                  Amy.