Is it normal for case fan speeds to fluctuate wildly?

[Pseudoscience888]

I just built a gaming system. The case came with two fans which have two-way speed switches which I left at the low setting for the first couple of weeks. Playing Skyrim Special Edition, the system seemed to be keeping itself quite cool with those fans running between 600 and 800rpm.

I just installed Elder Scrolls Online and right away noticed the system starting to really heat up for the first time. I flipped the case fan switches to the high setting, and I've noticed that the speeds are now fluctuating wildly between around 400rpm and around 15000rpm, all the way up to 18000. This is happening with the computer at idle. The CPU fan does not fluctuate like this.

I looked up the specs of the Twocool fans and they are rated at 12000rpm. So this wild fluctuating, and the system telling me the fans are running 5000rpm faster than they are rated, is making me worry something is wrong. Or is this normal? Thanks.

Build:
Motherboard: Gigabyte Z170XP-SLI
CPU: Intel i5-6500 with stock cooler
GPU: Gigabyte Geforce GTX 1060 6GB
Memory: 16GB Corsair DDR4
SSD: Samsung 840 Pro 128, and Toshiba Q-Series 128
Case: Antec P100
PSU: Corsair RM650i

 

[JackNaylorPE]

Huh ... 18000 rpm ? Your case would fly across the room ... those do not represent accurate numbers... or you have some extra zeros in your typed numbers

http://antec.com/product.php?id=705982&pid=56&lan=nz

A convenient switch allows users to easily choose either silent 600 RPM or maximum 1200 RPM cooling,

Where are you reading these numbers ?

 

[Pseudoscience888]

I am using Corsair Link to monitor my system. My Corsair PSU has a USB connection to the motherboard, and the Corsair Link software is made to monitor the PSU, but also displays the rest of my system information.

After you posted this, I referred back to Corsair Link and realized the case fan readout is switching between RPM for the lower numbers and RPR for the higher number. For example, it reads 422rpm, then immediately changes toe say 17000rpr. So this apparently changes the conversation from a wildly fluctuating fan speed to what is RPR? Which I am going to go Google now.

Huh ... 18000 rpm ? Your case would fly across the room ... those do not represent accurate numbers... or you have some extra zeros in your typed numbers

http://antec.com/product.php?id=705982&pid=56&lan=nz

A convenient switch allows users to easily choose either silent 600 RPM or maximum 1200 RPM cooling,

Where are you reading these numbers ?

 

[Paperdoc]

My guess is that all your case fan speed readings are bogus.

As you dscribe the system, your two case fans were pre-installed and are controlled by some Lo / Hi switch on the case. MOST syystems like that have NO connection between those fans, their control switch, and the mobo. Only the mobo can read and display fan speeds. With no connections to those case fans, the mobo cannot read them at all.

Now, the Corsair Link software can get the speeds the mobo has determined for the fans attached to its headers. The one that the mobo DOES control and read properly is teh CPU cooler, and that reading is very likely correct. But since the mobo CHA_FAN headers have nothing attached to them, any speed they display (and feed to Corsair Link) is false.

 

[JackNaylorPE]

I am using Corsair Link to monitor my system. My Corsair PSU has a USB connection to the motherboard, and the Corsair Link software is made to monitor the PSU, but also displays the rest of my system information.

After you posted this, I referred back to Corsair Link and realized the case fan readout is switching between RPM for the lower numbers and RPR for the higher number. For example, it reads 422rpm, then immediately changes toe say 17000rpr. So this apparently changes the conversation from a wildly fluctuating fan speed to what is RPR? Which I am going to go Google now.

RPR = Resident project representative is what comes to mind for me (it's an engineering term)

http://acronyms.thefreedictionary.com/RPR

 

[Pseudoscience888]

Interesting. Now, all of my fans are connected to the motherboard. The case fan switches are wires coming off the fans themselves with the switches at the end. Are you saying that the motherboard can't properly read fan speeds on the case fans? I will add that the case fans have three-pin connectors. Would it be different if they had four-pin connectors? I've been thinking about adding a couple of Noctua fans, connecting those to the motherboard, and driving the original Antec fans directly from the PSU. I believe Noctua fans have four-pin connectors, so would doing this give the motherboard accurate readings and control of the Noctua fans, and would the Corsair Link software also give me accurate reading and control of the Antec fans directly connected to the Corsair PSU? Or do fans need to have four-pin connectors for this regardless of whether they're powered by the mobo or the PSU?

My guess is that all your case fan speed readings are bogus.

As you dscribe the system, your two case fans were pre-installed and are controlled by some Lo / Hi switch on the case. MOST syystems like that have NO connection between those fans, their control switch, and the mobo. Only the mobo can read and display fan speeds. With no connections to those case fans, the mobo cannot read them at all.

Now, the Corsair Link software can get the speeds the mobo has determined for the fans attached to its headers. The one that the mobo DOES control and read properly is teh CPU cooler, and that reading is very likely correct. But since the mobo CHA_FAN headers have nothing attached to them, any speed they display (and feed to Corsair Link) is false.

 

[JackNaylorPE]

These switches must be for 3 pin DCV fans and they are confusing the MoBo sensor. The way speed control works on DCV fans is the higher the voltage, the faster they go. With switches, "back in the day" the switches just had "doohickeys" (it's technical term) that dropped the 12v feed from the PSU to 5, 7 and 12v for each of the 3 positions.

In the modern era, I can't imagine what the case designer was thinking of these are intended to work of the MoBo. If they were set up to work off the PSU, and now you are effectively modifying the design to work offf the MoBo, well... don't do that Gotta pick one method or the other.

 

[Paperdoc]

OK, now I have more questions. But first, a few "answers".

You say your fans are by Antec and they have 3-pin connectors on them, plus a cable that ends in a 2-position switch for speed control. I have similar Antec fans from some time ago but they 3 possible speeds and have 4-pin Molex connectors on them, so they cannot plug into a mobo fan header. Now, the "standard" 3-pin fan connector is female with 3 holes in a connector about ½" wide, and ridges up one side of it. It fits neatly onto any mobo SYS_FAN or CHA_FAN header, and the groove system ensures it only goes on one way. Is that what you have?

If that is the connector you have, then the speed of these fans is controlled entirely by voltage. Power is fed to the fan on Pin #2 (Red wire) of the connector, with Pin #1 (Black wire) as Ground. For control, the mobo header varies the voltage supplied on Pin #2, from about 5 VDC min up to 12 VDC. The Yellow wire on Pin #3 takes a pulse signal (2 pulses per revolution) generated in the motor back to the mobo header for reading. The mobo does NOT use that speed signal to accomplish control, although often it does monitor that for possible failure of the fan to work. It appears that your fan also has an additional non-standard feature in its special switch on the end of a cable. At the "Low" setting, all that does is insert a resistor into the power supplied from Pin #2 to reduce the voltage going to the fan motor. In its original design, this was the only way to change the speed of these fans, because they were fed from a fixed 12 VDC source from a Molex connector. But when they are used with a mobo fan header that actually does use Voltage Control Mode, you should always leave that fan selector switch set to "High" so that it does nothing to the supply voltage coming from the mobo.

SOME mobos with 4-pin fan headers use only PWM Mode to control fans. That system always supplies the fixed 12 VDC on Pin #2, and then sends the new PWM signal to the fan on Pin #4. Inside the fan is a chip that uses the PWM signal to modify the flow of power from Pin #2 to the motor, thus changing its speed. But a 3-pin fan cannot use that signal, so it can only run full speed if plugged into a true 4-pin fan header using PWM Mode. In that case, the extra switch on your fan could be useful. Just let me alert you about this, though: 4 pins on a fan header do not guarantee use of PWM Mode. There are some 4-pin headers that allow you to select in BIOS Setup which method it will use for fan speed control, and there are also many mobos today that use 4-pin headers but only use Voltage Control Mode (the system for 3-pin fans) on those headers.

IF your fans do have the "standard" 3-pin connector on them with 3 wires coded as above, then very likely your mobo headers CAN read their speeds properly, and hence Corsair Link should be able to display them. In that case, how you set the special added switch on the end of the cable will not interfere with the ability to read speed. BUT it certainly will change the speed, and it might interfere in this way: if you set it to "Low" and the mobo also tries to slow down the fan with a lower voltage, the fan may receive such a low voltage that it stalls. If that happens, the mobo might detect that and raise the voltage to re-start the fan, then lower it again to slow it down. Repeat cycle. THAT is why I said you should only set the switch to "High" if your mobo is actually able to slow the fan down by itself.

Now my questions: since you say that now all those fans are connected to the mobo headers, do their speeds actually change as the system workload changes? Does Corsair Link now display reasonable speed readings for them?

You have a lot of confusion about 3- and 4-pin fans. BOTH types can report their speeds to the mobo. The two types differ in the WAY that their speed is controlled. To do that, the mobo header must use different signals sent out on its pins. And lots of confusion - NOT yours - is caused by the way that mobo makers have used odd designs for their headers' control systems.

Like many makers, Noctua makes both 3-pin and 4-pin fans. They are not all the same.

If you tell us your exact mobo maker and model number, we can look up its fan control system and advise. Tell us also how many case vent fans you have, and tell us about what they are doing now while connected to mobo headers. Also tell us this: are you using any Splitters to connect more than one fan to a single mobo header?