Problem powering new case fans

[46andtool]

Okay so I installed a new vid card and then I attached 2 Rosewill 120mm case fans as intakes onto my side panel. I hooked one up to my motherboards sysfan3 connector, and the other fan using an adapter to my PSU.

The fan hooked up to sysfan3 started up no problem on initial boot-up but the second fan that was running off the PSU didn't spin at all. I thought okay I must have done something wrong when I decided to power it off my PSU. My system began loading windows and it began to hang, so I took my side panel off and now the CPU fan had stopped spinning! It would hitch like it was about to start spinning but then would stop for a few seconds..then hitch again.

I immediately unplugged the rosewill fan that was getting power off sysfan3 and sure enough the cpu fan spun right up and behaved normal.

My motherboard is the MSI Z87-G41 http://us.msi.com/product/mb/Z87-G41-PC-Mate.html#/?div=BIOS

And these are the fans I'm working with:

http://www.newegg.com/Product/Produc...82E16835200049

The other two sysfan connections on the mobo are already being used by the front intake and rear exhaust fans that came with my corsair 300r
What is the best way to power these new rosewill side panel fans?

Can someone tell me what the heck I'm doing wrong? Thanks!

 

[dma0991]

What about the fan that you initially connected to SYSFAN3? Does it show similar symptom to the one that you tried with the molex and then with the SYSFAN3 header? If it doesn't, its quite possible that the fan that you tried with both methods is a faulty unit. I would like to know how many amps it requires (you can find that info on the center hub of the fan, 0.??A).

These are 2000 RPM units. I do not recommend running them off the molex connector as they will be very loud since they would run at 100% regardless of load. Best way is to get a Y-splitter and connect these two fans to SYSFAN3 header to regulate its speed. However, depending on the total amperage of the two fans, you may not do this as most motherboard headers are limited to 1A.

 

[46andtool]

What about the fan that you initially connected to SYSFAN3? Does it show similar symptom to the one that you tried with the molex and then with the SYSFAN3 header? If it doesn't, its quite possible that the fan that you tried with both methods is a faulty unit. I would like to know how many amps it requires (you can find that info on the center hub of the fan, 0.??A).

These are 2000 RPM units. I do not recommend running them off the molex connector as they will be very loud since they would run at 100% regardless of load. Best way is to get a Y-splitter and connect these two fans to SYSFAN3 header to regulate its speed. However, depending on the total amperage of the two fans, you may not do this as most motherboard headers are limited to 1A.

What I did was connect 1 fan to sysfan 3 and 1 fan to the PSU. The one connected to the PSU didnt spin at all for some reason, and the one connected to sysfan 3 made the CPU fan stop spinning!

I just checked and the fans are 0.4A.


I'm just concerned that hooking up a fan to sysfan3 caused my cpu fan to stop spinning! thats not good! thanks for your help

 

[dma0991]

What happens if you plug the Rosewill fans to the fan headers that are currently occupied by the front and rear fans? If it works, get two Y-splitter to power all four fans.

 

[46andtool]

What happens if you plug the Rosewill fans to the fan headers that are currently occupied by the front and rear fans? If it works, get two Y-splitter to power all four fans.

I'll test it out in the morning, if it works that just gives me an excuse to order stuff from frozencpu then :whiste:

i wonder why sysfan3 is interfering with the cpu fan?

Could it be possible that running fans off of all 3 sysfan connectors is pulling to much power from the motherboard causing the cpu fan to stop spinning?

I may also daisy chain the fans and try running them off the psu again..might just say screw it and pick up a fan controller.

 

[dma0991]

Could it be possible that running fans off of all 3 sysfan connectors is pulling to much power from the motherboard causing the cpu fan to stop spinning?

Can't say with certainty as there might be more to this problem that can't be seen with forum posts. If you're planning to get a fan controller, consider this as a cheaper alternative. That splitter should have no problems with power as it draws from the PSU's SATA connector while getting the speed control from the motherboard.

 

[Tweakin]

What I did was connect 1 fan to sysfan 3 and 1 fan to the PSU. The one connected to the PSU didnt spin at all for some reason, and the one connected to sysfan 3 made the CPU fan stop spinning!

I just checked and the fans are 0.4A.


I'm just concerned that hooking up a fan to sysfan3 caused my cpu fan to stop spinning! thats not good! thanks for your help

Is the .4A within limits of your board? That seems like a high amount of current for a board bus to handle.

 

[monkeydelmagico]

Is the .4A within limits of your board? That seems like a high amount of current for a board bus to handle.

Agreed. I'd run them straight off the PSU. If they run too loud a fan controller or in-line step down would work too.

 

[BonzaiDuck]

Is the .4A within limits of your board? That seems like a high amount of current for a board bus to handle.

First, I'm thinking if he connected one fan to a PSU modular port designed for that purpose, it may not be a malfunction. The PSU fan itself won't spin below certain operating temperatures. I have a Seasonic "Gold" 80-plus, and I don't think the fan has ever spun up at all on that thing. It runs so cool you can't tell the difference.

It seems that mobo headers are limited to 1A. I'd seen in earlier generations where the spec was "overall" for all headers on certain motherboards, implying that you could exceed one amp for a single port if you didn't use the others. But why test the uncertainty about it.

What the OP describes as the effect of removing a "SYS" or "CHA" fan from the board certainly suggests that he may have exceeded the board-maker's intentions for the total amperage draw. I think I encountered it on my own Z68 mobo when I attempted to hook up a 0.70A fan to CHA_FAN2. There wasn't enough power to start the fan, and it behaved in the way the OP described.

I had also made some posts some weeks ago about the Swiftech 8W-PWM-SPL-ST (you can't get it with a Molex anymore). Depending on the number of PWM fans, you could mod their cables and run the tach wires to your unused 3-pin mobo ports so that you get the monitoring you want, since the Swiftech only monitors one fan.

It becomes a temptation to buy a fan controller, but if you're like me, you want total thermal control. so you 'd spend more money and get a controller with thermal sensors. Then you'd go to some extra trouble just to install those sensors where they'd do some good. What you'd really want is a controller that would read the CPU's own temp sensors and allow you to create fan-curve profiles without installing an analog sensor where it might actually read significant rise in CPU temperatures. That means a bigger outlay for such a controller.

The Swiftech, Akasa and similar PWM solutions can forestall that extra expense. Only problem is that some fans -- especially the large 200mm units -- are still 3-pin. Somebody mentioned that PHanteks will be releasing a 3-pin-to-PWM converter gadget soon. Point being: whatever allows you sufficient monitoring and thermal fan control while taking the power-draw from the PSU instead of the motherboard -- that's a good thing.

 

[46andtool]

Can't say with certainty as there might be more to this problem that can't be seen with forum posts. If you're planning to get a fan controller, consider this as a cheaper alternative. That splitter should have no problems with power as it draws from the PSU's SATA connector while getting the speed control from the motherboard.

That looks good however I don't think my fans are pwm fans, my powers been out since Wednesday from a snowstorm so ive been unable to do any more trouble shooting on the issue im having, but its tax return season so Im thinking about just springing for a good fan controller and a few pwm high cfm fans

 

[46andtool]

Is the .4A within limits of your board? That seems like a high amount of current for a board bus to handle.

Im unsure about that, I checked the manual and it doesn't seem to mention the power draw limit of the sysfan ports

 

[dma0991]

That looks good however I don't think my fans are pwm fans

This diagram shows two wires (PWM and fan speed) connects to the motherboard. It can revert back to voltage control if your fans are only 3 pin.

 

[46andtool]

First, I'm thinking if he connected one fan to a PSU modular port designed for that purpose, it may not be a malfunction. The PSU fan itself won't spin below certain operating temperatures. I have a Seasonic "Gold" 80-plus, and I don't think the fan has ever spun up at all on that thing. It runs so cool you can't tell the difference.

It seems that mobo headers are limited to 1A. I'd seen in earlier generations where the spec was "overall" for all headers on certain motherboards, implying that you could exceed one amp for a single port if you didn't use the others. But why test the uncertainty about it.

What the OP describes as the effect of removing a "SYS" or "CHA" fan from the board certainly suggests that he may have exceeded the board-maker's intentions for the total amperage draw. I think I encountered it on my own Z68 mobo when I attempted to hook up a 0.70A fan to CHA_FAN2. There wasn't enough power to start the fan, and it behaved in the way the OP described.

I had also made some posts some weeks ago about the Swiftech 8W-PWM-SPL-ST (you can't get it with a Molex anymore). Depending on the number of PWM fans, you could mod their cables and run the tach wires to your unused 3-pin mobo ports so that you get the monitoring you want, since the Swiftech only monitors one fan.

It becomes a temptation to buy a fan controller, but if you're like me, you want total thermal control. so you 'd spend more money and get a controller with thermal sensors. Then you'd go to some extra trouble just to install those sensors where they'd do some good. What you'd really want is a controller that would read the CPU's own temp sensors and allow you to create fan-curve profiles without installing an analog sensor where it might actually read significant rise in CPU temperatures. That means a bigger outlay for such a controller.

The Swiftech, Akasa and similar PWM solutions can forestall that extra expense. Only problem is that some fans -- especially the large 200mm units -- are still 3-pin. Somebody mentioned that PHanteks will be releasing a 3-pin-to-PWM converter gadget soon. Point being: whatever allows you sufficient monitoring and thermal fan control while taking the power-draw from the PSU instead of the motherboard -- that's a good thing.

What I did was use a molex to 3pin adapter to run the fan off the psu, im unaware of any ports on the psu specifically designed for fan control although I may be wrong its happened before

What you mention sounds interesting, is there such a thing as a fan controller with thermal sensors or a mobo hookup that uses the mobos built in sensors to control the fans? Cause thatd be amazing

 

[BonzaiDuck]

What I did was use a molex to 3pin adapter to run the fan off the psu, im unaware of any ports on the psu specifically designed for fan control although I may be wrong its happened before

What you mention sounds interesting, is there such a thing as a fan controller with thermal sensors or a mobo hookup that uses the mobos built in sensors to control the fans? Cause thatd be amazing

Even some of the less expensive controllers come with one analog sensor. Someone here recently had settled on an NZXT unit for $30 or so that came with several of them. The difficulty with this: you would have to do the work required (probably removing your CPU cooler) to install a sensor either between the CPU and motherboard -- running the thin wires between the CPU pins), or getting it under the IHS. Using thermal tape to put it on the bottom of the HSF base so that it isn't sandwiched between cooler and IHS doesn't yield temperature differences that would allow for robust control. Some of the controllers (like T-Balancer mentioned below) provide a calibration process that might work with that type of sensor placement.

I had thought that some of the high-end controllers -- like the Aquaero 5 LT -- would read the CPU and motherboard sensors through their USB connection. I'd also described how it was done with the equally expensive M-Cubed T-Balancer "BigNG:" it requires "other" software (MBM) which hasn't been updated since 2005. Otherwise, these high-end controllers come with their own software, but require use of the bundled sensors.

I think the problem the controller-makers have is similar to Comparetti's SpeedFan: they have to keep track of the sensor components the mobo-makers use in different mobo models. So if that's what you're looking for -- say, the Aquaero -- you might want to verify with them what that "intelligent" USB connection to the motherboard actually provides.

With the $10 Swiftech item and similar (Akasa has one), you can thermally control several PWM fans from the CPU_FAN PWM header on any motherboard, but you can only monitor one device. But -- only PWM fans. Someone here had noted how a friend built 3-pin-to-PWM converter devices that allow for PWM thermal control of 3-pin fans, and the circuit diagrams can also be found on the web. Another company, Nanoxia, produced a device for about $10 called the "PWMX", which is simply a wire and barrel-shaped device which does the same thing. I've searched all over to purchase one, but they're not in stock anywhere.

 

[BonzaiDuck]

The only customer review at Frozen CPU says:

"I''ve had this for a week now and it''s great - reads my CPU temperature and adjusts fans for maximum quietness. Perfect if you like playing with your system to get the best results."

The very lengthy PDF "guide" and promotion for the Aquaero is less clear. Obviously, and as I'd discovered some time ago with similar devices, the USB connection between your computer and the onboard microprocessor(s) allows for control of the device through Windows. There is a hint that the device reads BIOS data. This possibility might mean reading the temperature monitors of the motherboard.

Somehow, the PDF also suggests something about "windows software" for "virtual sensor."

These devices come bundled with a lot of analog sensor wires that you'd have to install yourself. But for the most part, if the device could read the motherboard CPU and "mobo" sensors, you wouldn't need to do all that.

As far as I can see it, you might only need a couple temperatures at most to control such a device: the CPU or an average of the core values; a temperature for a graphics card; a temperature from the motherboard, chipset, etc. And frankly, I think you'd have a robust and elegant system if you can thermally control several fans using only the CPU temperature.

So the question is: "How is the CPU temperature to be provided? Reading the motherboard sensor and transmitted through the USB connection? Or installation of an analog sensor?"

 

[BonzaiDuck]

Someone had posted a guide to this Aquaero controller -- look for "NAM'S NEW BASIC GUIDE to the . . . "

Here's an excerpt:
==================================
"

What if you run your PC with just Air Cooling, I hear you ask?.....
Well!!....
This is where the Aquaero/Aquasuite SOFTWARE SENSORS come in very handy.....
​

Here's my Guide to get you up & running with setting up some SOFTWARE SENSORS

......


The Aquaero/Aquasuite Software Sensors are basically Temperature Readings that the Aquasuite gets from your MotherBoard Temp Sensors, Which in turn means your able to Monitor the CPU (induvidual cores), GPU, RAM, Motherboard e.t.c e.t.c Chips Directly through a Third party Software Program that then relays the readings to the Aquaero/Aquasuite.

This then means that your able to use the Monitoring & Controlling capabillities of the Aquaero to have full control over your Fan Air Coolers & also the Case fans e.t.c. I even used the Aquaero for Monitoring & Controlling the CoolerMaster Seidon 120M all in one Kit I did a review for a few weeks back & it worked fantastically at keeping the Seidon Cooler running at it's best with minimal fan speed & still able to cool a well OC'd i7 860...

First of all your going to need that Third party Software Program I mentioned above...

For this the Aquasuite needs either OPEN HARDWARE MONITOR V. 0.5.1 beta or AIDA64 EXTREME EDITION . The "Open Hardware Monitor" (OHM) software is FREE (or you can donate if you wish) & the "Aida64 Extreme Edition" (Aida64) is a PAID for License (look for special deals)..

To begin with you need to Download & Install which ever of the above two you choose to go with (I personally use both, but mostly I use the Aida64 as it has many more functions to it than the OHM).. Either one will display most Temperature Readings that we use for the Aquaero, Namely the CPU's, GPU's, Motherboard & Ram as well as the other, So take your pick.. ...


Once you've Installed either Aida64 or OHM open it so it's running on your desktop & follow the guide below (It's easy) & everything should be tickity booo.... "

==================================

So, the Aquaero LT 5 (and similar) works much like M-Cubed's T-Balancer. It reads the CPU and motherboard sensors through third-party software like HW Monitor or AIDA 64.

IF this is the best any of the manufacturers can do, it's probably good enough. But I'd rather have a device that doesn't need an extra software layer running to get the CPU and motherboard sensor data. I"d rather have the Aquaero's own software able to do that.

But given the proliferation of new CPUs from Intel and AMD, the number of different motherboard makes and models, it would be a more daunting task for a hardware maker to keep up with new mobos, CPUs and sensor chips than it would for Comparetti and his SpeedFan. And SpeedFan never seems up-to-date enough.

So you'd run HWMonitor or AIDA as well as the "Aquasite" software or whatever it's called.