Results 1 to 3 of 3

Thread: PWM Fan hub Test and Review

  1. #1
    Super Moderator
    Join Date
    May 2013
    Location
    United Kingdom
    Posts
    714

    PWM Fan hub Test and Review

    A electronics mate of mine borrowed my PWM fan hub for testing. Her are is what he as found out so far.:




    Phanteks PWM Fan Hub Review
    (PWM control converted to variable voltage fan hub)
    By Tealc


    The Phanteks PWM hub is a small device that allows the user to control a number of 3 pin fans using your computer’s native CPU PWM control signal, or valid PWM signal you choose to feed into it.


    [IMG ALT=""]http://www.overclock.net/content/type/61/id/2910039/width/700/height/1000[/IMG]


    The hub measures approximately 45 x 40 x 15mm so is small enough to be tucked away behind a motherboard. The fans exit around the outside of the hub and this increases the footprint considerably but adds no further height.


    The unit can be powered in two ways:-
    • Via computer PWM header using the standard 1 Amp header.
    • Via special 2 pin connector and cable assembly to SATA connector for more power.




    Both can be connected as they have a shared connection on the PCB. To fully disconnect PWM header power the wires would need to be broken, or terminals pulled.


    How the device operates:
    A PWM signal is taken from the single 4-pin connector and converted to DC power which then feeds each of the six 3-pin headers.


    How it works, the circuit:
    A NPN transistor takes the PWM signal and sends it to a push pull NPN/PNP transistor pair where it is level shifted to a 12v PWM, this in turn feeds into a P-channel Mosfet which handles the switching of the 12v fan load. There is also an inductor and a capacitor on the output which does the smoothing and provides, more or less a DC voltage signal. A Schottky diode sits across the fan load and prevents any significant reverse transient spikes that could potentially destroy the Mosfet. There are a number of other resistors and diodes.


    Interestingly there are a number of positions unpopulated, R5 (resistor), C2 (capacitor), D3 (diode), and perhaps most interesting of all on the front are spaces for three resistors and three LEDs. These are wired directly to the 12v source and the resistors are there to limit current. I feel this is an opportunity missed by Phanteks as they really would be better on the fan output and get brighter as output voltage increases.


    The benefits.

    • Simple to use. Plug your fans in, provide power and PWM and the fans will adjust according to CPU temperature.
    • Easy to tuck out of sight.
    • Can be mounted on the chassis.
    • Available with some cases fitted and also available in a neat plastic box.
    • Energy efficient.






    The limitations, drawbacks and potential issues.
    There is no adjustment that can be made to the output directly, all must be done in software, or by changing the fan load to effect a desired speed.


    The hub is set up to work with an ideal load, of around 6 or 7 fans, which is around 15W. Too much above or below this load compromises the way the hub operates. With just one fan you will get 9v -12v. At a much higher load you’ll get 3v-10.5v for the same duty cycle. See table and graph for detail.


    [IMG ALT=""]http://www.overclock.net/content/type/61/id/2910040/width/700/height/1000[/IMG]




    Different fans will run at different speeds depending on their individual characteristic and there is no easy way to get feedback from the hub to the motherboard headers, only the white fan header feeds back.


    At higher loads it was noticed that some voltage spikes occurred with the switching of the circuit, this may cause premature failure of the circuitry and produce extra heat. The image below shows a nice smooth trace with 4 fans, then the lower shows some spikes with 11 fans.


    [IMG ALT=""]http://www.overclock.net/content/type/61/id/2910041/width/700/height/1000[/IMG]

    The Mosfet package can get very hot when a high load of fans is attached. With a load of 24W and 12 fans the Mosfet was too hot to touch after a few seconds at 50% duty cycle. This may cause premature failure. At low duty cycle the temperature was reduced. It may be possible to heatsink the mosfet, however it is a very small package measuring only 7mm x 5mm.


    Without care it will be possible to overload your PWM fan header. Ideally 12v or ground should be disconnected from CPU_Fan lead if using external power source but this is not done automatically.


    Not really suitable for 4 pin fans.


    No LEDs.


    Testing

    • For the purposes of this test I used a collection of 3 pin 12v fans I’ve accumulated over the years.
    • I used a Gould OS300 oscilloscope to view output at the fan header.
    • I used a 555 timer based 22000kHz variable PWM circuit to simulate the PWM.
    • I also used a linear regulated 12v power supply which outputs a solid 12.3v no matter what load I throw at it.
    • I measured current using a Mastech MS8268 DMM and voltage using Isotech IDM600 DMM.
    • I estimated 20%, 50% PWM using the scale on the oscilloscope and actual max PWM is within a 1 or 2% of full 100% PWM.




    Conclusion
    I found the Phanteks PWM hub behaviour just as I’d expect from such a simple device. I was a little disappointed with the heat generated at higher loads but feel that multiple hubs could be employed to get around this issue. The circuit itself wouldn’t load down the PWM signal too much so it’s likely that 2 or 3 of these things would work just fine. Ideally Phanteks could have used a mosfet on the input, but to keep costs down they went with an NPN transistor.


    I’d really like to see a version of this with temperature related LEDs and an adjustment to compensate for the amount of fans.


    Hub supplied by doyll
    Published layout by doyll
    =i7 980X @ 3.55GHz =PH-TC14PE w/2x TY-147A fans =Crucial Ballistix 3x4GB =GA-X58A-UD5 =ENGTX580 DirectCU II =e quiet! Straight Power E10 500W =Enthoo Evolv w/2x PH-F140XP case intake fans;

  2. #2
    seems all the fans attached to this controller must run at the speed of the 1 white connector, so pwm type control is not possible. why would an experienced user want all fans run at one speed. and there is this crazy heat problem with the controller. nonesense. disapponts just like this case dissappoints. We cannot put a slim rad and fans inside the case. not enough room . fans when used with a liquid cool radiator must only go in roof. so led fans cant be seen. nonesense. plastic tabs in roof dont allow a 360 rad in roof, more nonesence. Emthoo lux case is so disappointing.
    Last edited by file1man; 09-26-2017 at 03:45 AM.

  3. #3
    Super Moderator
    Join Date
    May 2013
    Location
    United Kingdom
    Posts
    714
    Quote Originally Posted by file1man View Post
    seems all the fans attached to this controller must run at the speed of the 1 white connector, so pwm type control is not possible. why would an experienced user want all fans run at one speed. and there is this crazy heat problem with the controller. nonesense. disapponts just like this case dissappoints. We cannot put a slim rad and fans inside the case. not enough room . fans when used with a liquid cool radiator must only go in roof. so led fans cant be seen. nonesense. plastic tabs in roof dont allow a 360 rad in roof, more nonesence. Emthoo lux case is so disappointing.
    Sorry you are disappointed with your Enthoo Luxe. What plastic tabs don't allow 360 radiator in top?
    =i7 980X @ 3.55GHz =PH-TC14PE w/2x TY-147A fans =Crucial Ballistix 3x4GB =GA-X58A-UD5 =ENGTX580 DirectCU II =e quiet! Straight Power E10 500W =Enthoo Evolv w/2x PH-F140XP case intake fans;

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •