I would like to run multiple fans from a motherboard with only one variable voltage three pin output. I already know the output will not supply the required current. I've looked all over but I'm not sure exactly what search terms to use. Is there a product that will accept a higher current 12v supply and just use the fan output voltage to regulate the voltage to the fans? I would like the speed regulation without drawing the current from the motherboard.
Thanks for the quick reply. I did run across these solutions when I looked, but I didn't think they could slave one fan's voltage to another eg duplicate a signal voltage. I don't have any sensors, all I have is the output voltage that would normally run the fans. So when the signal voltage is 7.8V I would like the voltage from the controller to be 7.8V as well. Will these controllers do that for me? I will research further mayself as well.
Edit: I would like an out of sight package that can be tucked away in the case...Out of sight, out of mind.
Sorry, what you want I've never heard of, although I am sure one could be made. Maybe even one can be found to purchase - I just don't know where.
Likewise your last idea is viable, but I've never heard of such a unit. It would probably be more expensive than a really good fan controller.
BUT I can tell you a few ideas won't work.
1. Your first thought is correct. It is common opinion that you can wire TWO fans together in parallel from one mobo port and the current available will be sufficient for the worst-case condition - start up. But more than two is pushing it, and you want "multiple".
2. The item uther39 linked is a good one, but does not fit your situation. It is used only for 4-pin fans. It is set up to draw all the fans' power from a 4-pin Molex connector direct from the PSU which has LOTS of current available. Then it shares the special PWM signal from a mobo 4-pin fan output among the three fans it controls. ALL of those fans need to be 4-pin so they can use that PWM signal correctly, and they do not load that signal line heavily so it's OK to share the signal.
A last comment: I went to a popular on-line part seller's site and looked through their fan controllers. Not impressed. The less expensive units just have knobs that allow you to set the speed of each fan connected to it, but do not have any way to sense or display temperatures. But at least they work! The fancy ones have sensors included that you mount at points in your system, and actual fan controllers that use measured temps to control fan speeds. However, user feedback on these seems often unflattering - units fail, and many cannot remember their temp settings so they need to be re-adjusted for every fan at every boot-up.
OP, you may be onto something! I'm starting to think that third-party fan controllers are not a good alternative, and we need a good way to power many case fans from one reliable mobo controller.
Paperdoc, I think you know exactly what I'm talking about. The temperature sensor located directly on the cpu would likely react the quickest to any additional heat so I would like to avoid external sensors.
I am surprised that a product does not exist that would transform a voltage level into a PWM signal, at least I haven't been able to find any. I might need to tinker with it and create my own. Until then I'm running full blast on 12V. My issue is not actually multiple fans, it is a blown fan controller. It still outputs correct voltage when hooked to a high impedence load like a voltmeter, but if a fan is connected it cannot supply the current and the voltage drops to 1-2 volts. If I could just use that signal to control a driver for the fan then I could quiet my system.
I am actually quite impressed with the caliber of responses on this site. I actually joined just to ask this question but I think I'll stick around. Although the product I'm looking for doesn't seem to exist, all of the posts were helpful.
Actually, the CPU temp sensor is the best measure of the temp inside the CPU chip, but case temps (more important for northbridge and southbridge chips of the mobo, and for GPU cooling on a graphics card) are better measured with a sensor in the case. For this purpose most mobos use a sensor built into the mobo itself somewhere. In this way they can run two independent temp control loops. One is for the CPU cooler system, and is guided solely by the CPU's internal sensor. The other is for Case Fan or SYS_FAN, and it is guided by the mobo sensor. A few cases have two SYS_FAN outputs, but I'm not clear whether they use separate temp sensors, or share one as their guide.
The unit that uther39 linked to is designed to use the mobo's CPU cooling fan output to control three fans, but as I said they must be 4-pin fans, and the mobo output port must be 4-pin type to provide the PWM signal. Normally, this wiring unit powers the CPU cooler on one of its outputs, and the wires also convey the CPU cooler fan speed signal back to the mobo. The other two outputs are for other fans, and they simply will mimic the CPU fan's actions. They could be case fans, in which case the case cooling is being driven by the CPU's cooling needs, as you suggest. Another way to use this wiring accessory, though, is to connect its leads intended for the mobo's CPU_FAN output port to a 4-pin SYS_FAN port and then run all three of its outputs to case fans. In that situation they all would be guided by the the case cooling controller and the mobo's temp sensor. As you say, you can't use it that way because you don't have a 4-pin SYS_FAN output port on your mobo.
That wiring device is relatively cheap to produce because it does not have to reproduce the PWM signal - it merely lets up to three fans share it in parallel, and that works because each 4-pin fan motor's internal controller chip loads that signal line only minimally. However, to build a circuit that takes one 0-12VDC signal and produces three or four such signals, each capable of supplying a fan motor with up to a couple of amps (for start-up) takes much more in design and components, and probably would cost as much as existing third-party fan controller modules. So it appears nobody has decided to try to crack the market with such a device.
As regards the quality of the posts here, you are right. There are many people who are real experts on particular things, and lots of people with experience in your exact problem almost every time. I like to go where there are people who know more than me so I can learn and get help, so I stick around here, too.
...That wiring device is relatively cheap to produce because it does not have to reproduce the PWM signal - it merely lets up to three fans share it in parallel, and that works because each 4-pin fan motor's internal controller chip loads that signal line only minimally. However, to build a circuit that takes one 0-12VDC signal and produces three or four such signals, each capable of supplying a fan motor with up to a couple of amps (for start-up) takes much more in design and components, and probably would cost as much as existing third-party fan controller modules. So it appears nobody has decided to try to crack the market with such a device...
I was actually thinking of somthing a little simpler than creating the high current PWM required for the fans. I will attempt to create a circuit that will take the 6-12V signal and will generate a 5V PWM signal for some 4pin fans to use. The high current can come from the PSU. I've got all of the electronic components in my basement, just need to look up the specs for the PWM signal and put in less overtime at work...so it will take a while. Too busy!
Interesting read about the temperature sensor achitecture. At this point all I know is that I have a signal that appears to be reliable and that right now my cooling is set to overkill and sounds like a jet taking off. Oh well at least I know the cooling is working.
The PWM signal likely does not need to be high-current - it just CONTROLS the current to each 4-pin motor, but it does not supply that current. That is why it is reasonable to run several 4-pin motors that share the PWM signal but draw power from a source like the 4-pin Molex power output connector from a PSU. See my post in another thread here:
What you suggest, in fact, could be used to convert a 3-pin mobo output port into a 4-pin one, quite possibly with a higher-current power source (Molex from PSU) that allows multiple fans to be run from one controlled port.