SYS_FAN2 auto control? Argghhh

I just put together a system with a GA-MA770-UDP3. Works fine, but I got 2 cheap-o 120mm Rosewill fans, and they're louder than I anticipated.

Smart System Fan is enabled in BIOS, but only the front (SYS_FAN1) is responding with lower RPM (incidentally, this is the system fan connector with 4 pins). SYS_FAN2 however is still going full blast... what can I do, other than go and buy a quieter fan for the rear exhaust? I tried simply disconnecting it but the internal temps (ambient and core) increased by 5-7C right off the bat... makes me uncomfortable, I'd really like to have both fans at least spinning.

Any advice/experience appreciated....
22 answers Last reply
More about sys_fan2 auto control argghhh
  1. First, Gigabyte doesn't seem to have a model # of GA-MA770-UDP3
    Even on the Taiwan site:
    http://www.gigabyte.com.tw/Search/Search_List.aspx?Keyword=GA-MA770-&SearchType=Spec

    There are two types of fan speed control, PWM and voltage. In a 3-pin connector, the 3rd wire doesn't control the fan speed, it just sends the RPM to the board to monitoring.

    I don't think Smart control will reduce your fan speed. It's designed for PWM (4-pin) control which you don't have. If your G board is like mine and sounds like it, with the "front" fan 4 pin and the other connector (on mine NB fan) is only 3 pin, you'd need a variable speed controller to set the read fan speed - I think! And the Smart System fan in BIOS is likely only referring to the front fan.

    SpeedFan software may be able to adjust the voltage on that fan for you - doesn't work on my computer so I have no experience. Worth giving a try. And remember, lots of nice fans out there for $10 and shipping - with control switches. Heck, I got really cute 3-led 80mms with controls for only 5.99 each. Just tack one on to your next order to save the shipping.
  2. Only CPU_FAN and SYS_FAN1 are modulated - the other two simply provide twelve volts. If you'll post your fan's pin count (three or four), first thing iin the morning, I'll put up a wiring diagram for an easily made splitter that will allow both to be controlled by SYS_FAN1, while still enabling separate monitoring by the BIOS in case of failure...
  3. bilbat said:
    Only CPU_FAN and SYS_FAN1 are modulated - the other two simply provide twelve volts. If you'll post your fan's pin count (three or four), first thing iin the morning, I'll put up a wiring diagram for an easily made splitter that will allow both to be controlled by SYS_FAN1, while still enabling separate monitoring by the BIOS in case of failure...



    Sorry for model error -- it's GA-MA770-UD3P.

    SYS_FAN1 is 4-pin, SYS_FAN2 is 3-pin.

    The odd part is that both fans are identical (Rosewill RFA-120) and have only 3-pin connectors.... so I have no idea how Smart Fan could throttle the front fan using that fourth pin. Nevertheless when I was tooling around and disabled Smart Fan, lo and behold the front fan went up to 100%. I don't understand.

    Regardless -- a wiring diagram would be great, so long as it's easy enough for a non-electrician to follow and I ultimately don't risk frying my mobo in the process...


    /edit -- I did try using Speedfan, but there was no preset configuration for my model and frankly I'm not inclined to push it with hardware I don't understand inside and out.
  4. Bilbat's saying that forget the 4-pin stuff or PWM. The board only attempts to modulate the front/SysFan regardless. It won't try to change the other connector. It would be able to adjust the front connector even if you used a 2-pin fan!

    Remember, the 3rd pin is only to monitor and report the "estimated" rpm.

    Bilbat knows much, much more about anything electrical than I do.

    All the wiring changes can be made with fingernail clippers and electrical tape if need be. No need to be fancy unless you plan on pulling on the wires a lot later! Just make sure you identify that red is positive and black is negative and which one is pin 1 and pin 2 on the motherboard. The manual header area shows this info in detail.

    --------

    Bilbat, plz also post the diagram for the other Fan wiring you have - I can't find the one you once posted about converting the fan wiring. Thanks!
  5. Thank you gentlemen.... I eagerly await the diagram!
  6. While I monkey around in AutoCAD to make the diagram, I'll give you guys some stuff to 'chew on'... The two headers work differently, and that's how the SYS_FAN can operate both three and four pin fans... First, a quick PWM (pulse width modulation) primer (if you already know this stuff, skip it :??: ):

    Standard PWM (i.e., the CPU_FAN header) is done by varying the 'duty cycle' (which the Brits, I think, have named better - they call it the 'mark to space ratio') of a low (~5V) voltage 20-25 KHz square wave generator on pin four - which then is used to switch 'gating' transistors on the fan itself, which switch the twelve volt supply (on pin two) to the fan, correspondingly, to control the fan speed. This diagram was made to explain a different recurrent problem, but will suffice... The Vth level illustrates the 'threshold voltage' at which the on-fan 12V 'gating' transistors fire - at the left, are the speed percentages achieved by switching as shown... You can see pretty plainly how it works. SYS_FAN (and this is odd - on the Intel socketed boards, it's SYS_FAN2, instead of SYS_FAN1 as with this board - I'll have to scan some manuals and see whether this is standard across all the AMDs) 'locks on' the 'command signal' (providing a steady, unswitched five volts on pin four), and then PWMs the twelve volt signal on pin two - so, you see how it can work with either kind of fan: if you have a three pin, the 'average' voltage on pin two is simply varying to control the fan; if you have a four pin fan, its on-fan 'switches' are locked on by the steady five volts on pin four, and the 'average' voltage on pin two is still simply varying to control the fan! Neat idea on GB's part!

    Mongo, here is the older diagram, which shows how to 'split' both - the pot on the three pins can be ignored (omitted), as I simply use it to 'balance' fan speeds between the northbridge and southbridge, as really tiny fans spin outrageously fast!
  7. I'm glad you're keeping it simple! Actually, I understand the explanation, but not the upper diagram. Thanks for the repost of the lower diagram.

    Remember wnatw, that's not the diagram you're waiting for!
  8. Heh, no worries. I don't cool my RAM anyway... :D
  9. You can use a few of these to get the parts - 'bare' fan header parts are hard to come by:
    http://www.newegg.com/Product/Product.aspx?Item=N82E16812119148
  10. bilbat said:
    You can use a few of these to get the parts - 'bare' fan header parts are hard to come by:
    http://www.newegg.com/Product/Product.aspx?Item=N82E16812119148
    http://img230.imageshack.us/img230/3707/0113g.jpg



    fantastic, i'll see what i can do with what i have. not to ask a ridiculously stupid question, but if I don't have enough fan cables to make everything connect, could i use.... speaker wire, properly shielded?

    regardless, thanks a lot bilbat/mongox. people like you make the internet good.
  11. Always welcome!
    Any wire will work, shielding is not an issue for these low frequency signals... The reason I suggest using fan splitter cables is that, if you want to be able to replace individual fans, you need the connectors anyhow, and that way you have the correct, consistent color coding. My policy is, if I've gotta take a soldering iron to it anyway, may as well do it right, and tidily, at the same time :sol: :

    If you don't care about having to solder again to replace a fan, you can omit the middle connectors, thus:

    and simply make do with the two connectors that came with the fans...

    If you're not the 'soldering type', one thing you will need to become familiar with is 'heat shrink tubing'. You can, again, get away without this, and simply wrap your soldered connections with electrical tape - but that's just, well, messy! (...too much Monk, I think!) Shrink can be found in the electrical department of any hardware store, and does just what it's called - shrinks... You cut a little length of it, and slide it onto the wire before soldering (dammit, I forgot the shrink, gotta unsolder and resolder!), some distance up the wire away from the solder joint, (so the heat of soldering doesn't partially shrink it where it sits!), then, when you're done making the joint, you slide it over the soldered segment so it covers the 'bare spot', and briefly hold a lighter under the tubing - the heat makes the tubing shrink tight onto the wire, insulating your joint... You might want to practice a little on some similar wire, so you see what size to use, and how to do it without setting fire to your shrink or wires!
  12. He he, I actually have one of those commercial heat guns for some project I was working on. A nice contract is good for filling up your tool chest! It's basically like a really ugly metal Conair hair dryer - except it puts out enough Watts to melt your hair. A quick wave of this baby over heat-shrink tubing works wonders - and no burnt plastic!

    If you do use electrical tape, use real elec tape like 3M or the fun colored stuff at Radio Shack.

    Thanks for all the great work bilbat - when I re-post these I'll properly attrib. You need to start putting a sig of some sort on your creations too!

    Oh a great source of wiring AND connectors is any old computer you or friends have gathering dust. Strip out the wiring and fans too, along with the front panel connectors, leds, switches, etc... If you can avoid removing the tiny copper ends of the wires from the connectors, do so, they are fragile and hard to slip back in properly. ie, splice the wires to the wires, don't try to attach directly to the connector.
  13. Yeah - I have a heat gun too, handy gizmo (great for 'quick-curing' epoxies, too...), but A, I'm too damned lazy to use it most of the time, and B, it's permanently ensconced next to our snow-blower, which has an electric-drive worm-gear thingie for push-button control of the auger direction, that freezes up in any kind of inclement weather - I mean, when the ^&%$ else would you be using it?
    Quote:
    Oh a great source of wiring AND connectors is any old computer you or friends have gathering dust. Strip out the wiring and fans too, along with the front panel connectors, leds, switches, etc... If you can avoid removing the tiny copper ends of the wires from the connectors, do so, they are fragile and hard to slip back in properly. ie, splice the wires to the wires, don't try to attach directly to the connector.

    Definitely good advice - I always clip 'em back to about one to two inches, in 'staggered' fashion, leaving about a quarter or three-eighths difference between their lengths, so when you solder leads to 'em, the half inch of shrink covering the joint is staggered - a sharp 'point' of errant solder won't so easily puncture an adjacent 'shrunk' joint and cause a short... One of these days when I'm dictating, I'll do a post on all my 'connector secrets': how to release and back pins out of every common computer connector, without ruining the pin for reinsertion :sol:
  14. Alright. I spliced according to the diagram, plugged everything in, and it seemed to work -- Speedfan displayed both speeds as roughly the same @ significantly less than 100%

    Then the short occurred. In my excitement I neglected to cover all the splices with electrical tape, and yes, there was a tiny spark as one of them hit the case and everything turned off. Panic set in for ten seconds, then I cycled the PSU master switch and everything turned on again, including the fans. However, both fans are now running at 100%. I made sure Smart System Fan was still enabled in BIOS, but... well, at the end of the day the thing is louder than it was before. Ironic.

    Any suggestions?
  15. Ackk! I know this is a bad time to quote myself, but "my policy is, if I've gotta take a soldering iron to it anyway, may as well do it right, and tidily, at the same time"; I'm a really old fart, and I've been doing this a really long time - given the combination, I've learned that being in a hurry is never a good thing!

    A few things to try - first, pull the wiring & go over it carefully. Then, toggle SmartFan to disabled, save & exit, reboot, power down. Power down at the PSU switch. Leave it off for about five minutes. Power back up, reset SmartFan to enabled, save & exit, and try it again... I'm not familiar with the exact hardware on the AMD boards, but, typically, if you 'blow' the output transistors that the LPCIO chip (which controls the fans) uses to run them, they blow 'open', and the fan won't run at all...
  16. Ok, I'll try it. Feel like an idiot. :pfff:

    On the other hand I suppose it could be much worse...
  17. Well, I tried the process you outlined, and no luck. I guess I can either spend $$ to get some fans that run superquiet at 100%, or drop $ on a fan speed controller...

    Rest assured that your solution/schematic DID in fact work, albeit for 10 seconds! It was nice while it lasted. :)

    Thanks again.
  18. If you take a look here:
    http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&DEPA=0&Order=BESTMATCH&Description=scythe+slipstream+120mm
    you'll see that the same fan is offered in a wide range of speeds, which, of course, means differing air movement cpacities as well as sound levels. These come from 500 RPM, 24 CFM, 7.5 dBA models, all the way up to 1900 RPM, 110 CFM, 37 dBA 'vacuum cleaner motors'; a little hunting will find a number of manufacturers who do pretty much the same thing. I have three different speeds of the same fan on the various intakes to my case, just to balance out the airflow between the main case volume itself, and the enclosed, 'drawered' drive cage - and then I use an ESA-based controller to further fine-tune the airflow:


    (out of view are the intake 'sucking' air through the top of the drive cage [which is inside the regular drive rack], and the pair of Arctic 'cage' type exhausts, similar to the one at the back, in the roof of the case...)
  19. Oh darn darn darn. Sorry you had a short after all this work.

    Since you're buying new fan(s), might look at ones with built-in switches. I got a couple of those, cute 3led ones at 80mm to fit my old case, and have been happy with em. I keep em both at med speed and they make no noise that can be heard. At high, you can hear them rev up like the CPU fan when booting or running Prime95.
  20. Wnatw -

    you might want to wait a bit before doing anything - I have another idea for you. If you're willing to do a bit more wiring, I thought, overnight (isn't it odd how the subconcious mind works :pt1cable: ), of a way to rewire your fans that may take care of your problem... Take me a little while to draw, and the only thing I'm not sure of is whether it will result in the fan 'speed detection' working, as I don't know for sure if either the hall effect sensors on the fans, or the detectors themselves on the board, will 'trigger' at half-voltage, but I think it's worth a try...

    Later,

    Bill
  21. OK - give this one a swat:

    ;) carefully, this time...
    As I mentioned, I don't know if the fan speed detection will work - may work for only one fan, may work for both, may work for neither - have to look in PC Health to find out; but it will run both fans at half speed, which is probably close to where you wanna be...

    Bill
  22. bilbat, I'm gonna email you when I decide to play with putting some attic fans in! The one in the house should be easy, since the IRS will pay some of the cost and Elec company a fair bit more.

    But an odd idea I had was putting an couple of inexpensive solar-powered fans in the lakehouse. This place normally sits empty and closed up in the Texas heat. About 2/5s of the small stilt house has an attic and I need to a unit there to replace the cyclone fan that's noisy and I often cover then hate climbing on roof to un-cover.

    But the interesting idea I had was for a solar fan in the main area - has no attic space and high ceiling. I turn off all electricity when house is empty except the fridge. I've noticed that even with no visitors, the elec use jumps way up during the summer, because the fridge works over-time to stay cool with indoor temps over 110F I'm sure. So, a solar fan at peak of roof would draw out that hot air when no one's there, reducing the requirements of the fridge. I figure could pay for the fan in a couple years. Sound reasonable? Think it would do any good even without a louver to allow air in?

    No A/C in the house now, but I could put a shutoff switch to disable when I do add A/C. Even with louver, could rig up way to close em.
Ask a new question

Read More

Gigabyte Rosewill Fan Motherboards Product