1
\$\begingroup\$

I have scoured this website and the rest of the internet looking for a solution to my particular issue, but as of yet I have not found an answer. If this is a duplicate question I am sincerely sorry, and I would appreciate being pointed in the correct direction.

I have an only Intel CPU heatsink with the stock Nidec fan that was discarded during a previous computer building project. I now have a new project where I require additional airflow so I decided to try and "re-purpose" the unused CPU heat sink fan for a cheap fix to my solution.

Everything I have read on the internet says to buy a 12V power supply, connect positive to positive and ground to ground and you should be good to go, but this has so far not worked for me.

I currently am using this Nidec CPU Fan and have wired it up to this Power Supply. Following this 4-Wire Fan Description (link 1 in comments) I have connected the yellow fan power wire to the red wire on the power supply and the black fan ground to the black power supply wire. The green tach and blue PWM wires have been left unconnected to anything and are simply hanging free.

When I connect the power supply to an outlet the fan "stutters" forward a single pulse like it is overcoming a single pole on the motor. I would estimate the turn is something like 30 degrees of a full revolution, but I have not measured this accurately. After the initial "stutter" or "pulse" in the right direction the fan stops moving. After waiting a few seconds the fan again "stutters" forward another 30 or so degrees before stopping again.

This behavior and the "simplicity" with which everyone else seems to be making these fans work leads me to believe that either my power supply is insufficient, or I need to feed a signal to the PWM wire. I found this YouTube video (link 2 in comments) where a guy just throws a big cap across what appears to be the positive and ground terminals of his power source and the fan runs. However, I am very unfamiliar with circuitry and his set up so I am not sure exactly what his solution is and if it will work for me problem.

This YouTube video (link 3 in comments) shows a pretty decent break down of how PWM and 4 pin fans work but I don't have a lot of fancy gear to hook up and generate the power signals he is using in the video.

My goal in the end is to keep my solution cheap! I really want to avoid building 555 timer circuit to control fan speed or anything crazy. I have access to a few capacitors from my father so I may be able to build a very basic circuit if a cap is all I need. However, if running this fan requires something more expensive/complex I will likely go buy a cheap 2 or 3 pin fan to replace the one I am trying to use, but again, looking for a near zero cost solution here.

Thank you for reading! Any help is very much appreciated!

\$\endgroup\$
8
  • \$\begingroup\$ link 1: pavouk.org/hw/fan/en_fan4wire.html link 2: youtube.com/watch?v=1nA-uE_AogQ link 3: youtube.com/watch?v=gKHww3qJbs8 \$\endgroup\$
    – APOLLO457
    Commented Sep 3, 2015 at 14:14
  • \$\begingroup\$ The PWM line should be internally pulled up by the fan, so if you don't connect it, it should run at full speed. Are you sure it's not grounded? The PSU seems big enough and your wiring is correct but the fan may draw more than its rated 0.28A when it starts, so a big cap might help. \$\endgroup\$
    – Roger Rowland
    Commented Sep 3, 2015 at 14:22
  • \$\begingroup\$ Are you sure you haven't accidentally killed the fan? I've recently connected a fan in reverse (killing it) and afterwards it would show a quite similar behaviour as the one you describe. Can you connect the fan to a mainboard (there should be a chassis fan header or something) and try if it works there? \$\endgroup\$
    – Arsenal
    Commented Sep 3, 2015 at 14:26
  • \$\begingroup\$ Roger Rowland - What do you mean by am I "sure it's not grounded?" Like am I sure i connected the ground wires? Or am I sure the power supply is working? The PS is new, and the little light on it comes on when it is plugged in. I don't have a volt meter or anything handy to measure anything, but I may need to take it over to my father's house to borrow his and test it. Any suggestions on what to test? \$\endgroup\$
    – APOLLO457
    Commented Sep 3, 2015 at 15:07
  • \$\begingroup\$ Arsenal - I am sure that I did not kill it with incorrect wiring, but not sure if the fan is still working properly. I hacked off the 4 pin connector all the way back to the connector to maximize the wire length connecting to the Power Supply I bought so I'm not sure how to go about reconnecting it to the MoBo header unless I go buy a pin puller or something... However, the fan has been sitting in the intel CPU box it came in for the last 3 years since I built my PC, so the only thing that could have killed it is maybe age? \$\endgroup\$
    – APOLLO457
    Commented Sep 3, 2015 at 15:16

1 Answer 1

Reset to default
1
\$\begingroup\$

Can you measure the voltage on the 12V line and see if your PSU is dropping to a low voltage and then resetting and trying again. This may be due to too much or too little load.

Certain types of switched mode power supplies need a load on their main output(s) to operate correctly, perhaps yours does and placing a more significant load will cause it to generate a stable voltage.

The PWM wire might prefer 5V rather than 12V though if you measure it you may find that it is indeed pulled up already to 5V or 12V with circuitry in the fan.

You could test the fan on a working PC set-up carefully using a old style diskdrive power connection, just stuff the ground and 12V wire ends into the correct holes (black and yellow) and hold them in place with your fingers or tape when you turn the PC on.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.