Nu_NRG sent me an email referencing his post above, and asked a question of me. I cite the relevant portion here:
As you can understand, and probably you already have came to the above conclusion on your own earlier, there may be a way to fit another fan into the PSU. My candidate would be SCYTHE S-Flex SFF21G (1900 RPM, 3-4 V starting voltage â€“ based on a 1600 RPM model review). Perhaps you have tried fitting the mentioned fan inside that PSU already, if not maybe you or someone from your surrounding can give an answer whether the SCYTHE fan will start spinning inside the PSU if it would be fed from RED, BLACK, YELLOW wires leaving the white wire untouched.
The reason Iâ€™m asking you this is because I donâ€™t have S-FLEX fans available here in Kiev and currently I have a chance to order one from Moscow, Russia, but I donâ€™t want to waste 17 dollars to find out that the experiment would be a failure and that I may even burn down the PSU circuitry if I would try to feed a 3-pin fan in such a way, but most likely is that the later shouldnâ€™t be true.
Anyway, if you are reading this then I have gotten your attention and I hope that youâ€™ll give me some sort of response.
Now my PC is almost completely at the 13 db level where I can clearly hear the motor ticking after eliminating the bearing noise, this ticking can be heard on your 30 cm recording of the MODU 82+ fan that you have in the end of the review article.
Iâ€™m not sure 100% whatâ€™s the source of that noise, but I believe that this has to do with a magnetic field switching between the coils. If Iâ€™m wrong and you have some sort of a solution for this noise then please let me know.
Thanks again for your time, and I canâ€™t wait to hear an answer from you!!!
I felt his query deserved a reply. Here's what I wrote:
Thanks for your email.
If you are a native Ukrainian, then I commend you on your excellent written English! Better than most native speakers!
I also must commend you on the research you've done. You've taught me things I did not know, as I have not experimented with replacing the fan in the PSU.
From the description you quote, it seems that one pair of wires from the PSU to the fan has steady 12V. The other pair has variable voltage to control speed. You should be able to use a multimeter to determine which pair is which, using the original fan.
I hope what I write is clear enough -- measure the voltages at the terminals to the fan with the PSU running. WARNING -- There is some risk of electrical shock if you are not careful.
Actually there's about a 10 minute pause between this and the last thing I wrote. I spent the time to locate a Pro 82+ 425W and opened it up, turned it on and measured those fan voltages. I was curious for myself and didn't want to tell you to risk something I did not try myself. I attached the black lead from the multimeter, set to <20VDC range. Then I used the red probe to touch the bit of metal exposed at the 4-pin terminal where the wires from the fan meet the PCB. Here's what I found, at idle:
yellow wire -- 12V
black -- 0V (ground)
white -- 0.25V -- this has to be RPM sense wire
red -- 2.6V -- this is the variable voltage that drives the fan motor.
SO.... assuming that 2.6V is normal for all the Enermax 82+ PSUs, the 1900rpm Sflex fan is not an ideal match. It won't start up at turn-on because the voltage is too low. It probably will start up as the PSU heats up; the main question would be whether the heat build up before this point is small enough so the fan can cool the PSU down to keep the drive voltage from continuing upwards. Chances are... probably.
Does all that make sense? You should probably try measuring the fan voltage on your PSU. Maybe it's not quite the same as what I measured.
After all the above, I got under my desk to check on the Modu82+ 625W that's powering my main system right now. I've been aware of some mild fan chuffing noise for the last loittle while. Too low to really be audible from my desk but definitely audible from up close. It's mostly the fan in the Enermax. Very faint, like I said. Now that I know what the start voltage is, however, and which wires to connect, I have a good option for replacement, some time in the future if/when this fan gets worse -- one of the Sanyo Denki fans used in the Seasonic M12D, which actually starts at that low voltage. Seasonic sent me a couple of spares to analyze.