[SOLVED] How to control 3 pin and 4 pin fans from same splitter/hub?

[BetosBurrito]

Hello! I have a mini ITX board (ASUS ROG Strix B450i) with one case fan header, but have two 4 pin 120mm fans and two 3 pin 120mm fans. I'm trying to find a solution that can control all four fans from a single header that doesn't cause the 3 pin fans to go 100%. Most of the 4 pin PWM fan hubs I've looked at don't have control options for the 3 pin fans.

I'm thinking a 4 way splitter (total amps of the four fans equals .68) off the one header might work best, but would there be a problem with mixing the fans on the same splitter?

 

[Darkbreeze]

Yes, because those 4 pin PWM fans and your 3 pin voltage controlled fans have completely different operational configurations. Three pin fans on a PWM header, even with a splitter, will always go full speed because they do not have PWM capability. You need additional hardware to control the signal. You will want to look at something like the NZXT Grid+ which is hardware capable of controlling BOTH 3 and 4 pin fans on the same hub, and needs only a single PWM header from the motherboard. Power is supplied via 4 pin Molex.

https://www.nzxt.com/products/grid-plus-v3

Although many boards these days are capable of independently being switched in the BIOS for both 3 and 4 pin fans, there are no single header capable of controlling both simultaneously.

 

[BetosBurrito]

Thanks for the tips! I did see in the BIOS being able to change the header to voltage, and in experimenting with a two way splitter I tried both a 4 pin and 3pin fan and that seemed to work in controlling both fans. I could be dead wrong and this is an incorrect/unhealthy method, and I'm happy to hear other thoughts.

 

[Darkbreeze]

Yes, you can control the 4 pin PWM fan through voltage controls, but it is bad for the PWM hardware inside the fan and the fan motor, which is designed for a straight 12v signal that is pulsed, hence the Pulse width modulation (PWM) specification. 3 pin will only run at full speed on PWM circuit, because it cannot pulse the 12v signal to modify speed. PWM will burn out the hardware at low voltages, which is what you will see on the voltage circuit with it set to voltage controlled when it is at any speed other than full speed. Only at full speed will you see at, or close to, 12v, with the voltage speed control enabled.

 

[CountMike]

Just switch to DC control in the BIOS, not as precise as PWM but will still be controllable.

 

[BetosBurrito]

Yes, you can control the 4 pin PWM fan through voltage controls, but it is bad for the PWM hardware inside the fan and the fan motor, which is designed for a straight 12v signal that is pulsed, hence the Pulse width modulation (PWM) specification. 3 pin will only run at full speed on PWM circuit, because it cannot pulse the 12v signal to modify speed. PWM will burn out the hardware at low voltages, which is what you will see on the voltage circuit with it set to voltage controlled when it is at any speed other than full speed. Only at full speed will you see at, or close to, 12v, with the voltage speed control enabled.

Thanks for the detailed explanation! Another option I was thinking about is having the two 3 pin fans split off the CHA_FAN header and then have the two 4 pin fans and the CPU fan split off of the CPU_FAN header. Does that sound like a viable option?

There is also the AIO_PUMP header, but I haven't had much luck being able to control any fan off of it.

 

[Darkbreeze]

Just switch to DC control in the BIOS, not as precise as PWM but will still be controllable.

He can't run BOTH PWM and voltage controlled 3 pin fans off the same header of the motherboard, no matter WHAT the header control method is set to. I already outlined this, if you didn't read it. One header cannot handle BOTH types of fans. There IS hardware controllers that can do it though.

You do not want the PWM fans running off the CPU_FAN header unless you don't mind them CONSTANTLY ramping up and down like the CPU cooler does. The thermal diode for the CPU fan header is much more reactive than a header for the case fans being controlled by feedback from one of the motherboard thermal sensors. The AIO pump header is likely full time 100% operation, but if it's not, you should be able to set that header to PWM and then possibly change the thermal sensor it uses to one of the motherboard headers instead of the CPU thermal diode. I'd have to look at the user manual or elsewhere to determine this for sure. You could go into the bios and look in the settings to see if there are selectable settings for which thermal sensor each header uses. Do NOT change the CPU_FAN header to anything other than it's default selection. Other headers can be changed based on need if the option exists on your board.

 

[BetosBurrito]

He can't run BOTH PWM and voltage controlled 3 pin fans off the same header of the motherboard, no matter WHAT the header control method is set to. I already outlined this, if you didn't read it. One header cannot handle BOTH types of fans. There IS hardware controllers that can do it though.

You do not want the PWM fans running off the CPU_FAN header unless you don't mind them CONSTANTLY ramping up and down like the CPU cooler does. The thermal diode for the CPU fan header is much more reactive than a header for the case fans being controlled by feedback from one of the motherboard thermal sensors. The AIO pump header is likely full time 100% operation, but if it's not, you should be able to set that header to PWM and then possibly change the thermal sensor it uses to one of the motherboard headers instead of the CPU thermal diode. I'd have to look at the user manual or elsewhere to determine this for sure. You could go into the bios and look in the settings to see if there are selectable settings for which thermal sensor each header uses. Do NOT change the CPU_FAN header to anything other than it's default selection. Other headers can be changed based on need if the option exists on your board.

These explanations are a gold mine. How about this setup? The two 3 pin fans off a splitter on CHA_FAN (set to DC), the two 4 pin fans off a splitter on the AIO_PUMP (poking around in BIOS I can set it to PWM, I'll have to look into the thermal sensor though), and then the CPU fan off the CPU_FAN. Hope that doesn't break anything.

 

[Darkbreeze]

IF, IF you can set the AIO pump header to something other than 100% speed AND if you can change the thermal sensor that it uses to make speed adjustments so that it is not using the CPU thermal diode, then it will work perfectly fine. If you can't, then it would still work temporarily but I promise you that you will get very, VERY annoyed with those fans going up and down, up and down, quickly and to very high speed, every time the CPU comes under a load.

It IS an option though. I have, in the past, set the rear exhaust fan to run in sync with the CPU cooler fan, and it was annoying, but it definitely lowered temps a degree or two and assisted the CPU fan with efficiency as it tends to immediately remove anything coming through the heatsink if you have a tower cooler. Especially if you have one with two fans. But it's annoying, and having a rear exhaust and a top rear exhaust, set to work normally, works just as well and isn't annoying like that is.

 

[CountMike]

He can't run BOTH PWM and voltage controlled 3 pin fans off the same header of the motherboard, no matter WHAT the header control method is set to. I already outlined this, if you didn't read it. One header cannot handle BOTH types of fans. There IS hardware controllers that can do it though.

You do not want the PWM fans running off the CPU_FAN header unless you don't mind them CONSTANTLY ramping up and down like the CPU cooler does. The thermal diode for the CPU fan header is much more reactive than a header for the case fans being controlled by feedback from one of the motherboard thermal sensors. The AIO pump header is likely full time 100% operation, but if it's not, you should be able to set that header to PWM and then possibly change the thermal sensor it uses to one of the motherboard headers instead of the CPU thermal diode. I'd have to look at the user manual or elsewhere to determine this for sure. You could go into the bios and look in the settings to see if there are selectable settings for which thermal sensor each header uses. Do NOT change the CPU_FAN header to anything other than it's default selection. Other headers can be changed based on need if the option exists on your board.

He can't run BOTH PWM and voltage controlled 3 pin fans off the same header of the motherboard, no matter WHAT the header control method is set to. I already outlined this, if you didn't read it. One header cannot handle BOTH types of fans. There IS hardware controllers that can do it though.

You do not want the PWM fans running off the CPU_FAN header unless you don't mind them CONSTANTLY ramping up and down like the CPU cooler does. The thermal diode for the CPU fan header is much more reactive than a header for the case fans being controlled by feedback from one of the motherboard thermal sensors. The AIO pump header is likely full time 100% operation, but if it's not, you should be able to set that header to PWM and then possibly change the thermal sensor it uses to one of the motherboard headers instead of the CPU thermal diode. I'd have to look at the user manual or elsewhere to determine this for sure. You could go into the bios and look in the settings to see if there are selectable settings for which thermal sensor each header uses. Do NOT change the CPU_FAN header to anything other than it's default selection. Other headers can be changed based on need if the option exists on your board.

You must have some peculiar fans or misread what I wrote. I'm doing it right now. Two top 3pin fans and one 4 pin at the back running of same 4pin PWM CHA_Fan header in DC control mode .

 

[Paperdoc]

My understanding of these two fan designs differs somewhat from some of the posts above. The speed control of 3-pin fans is done solely by changing the voltage to the motor supplied on Pin #2, from 12 VDC for max speed down to about 5 VDC min for lowest speed. A lower voltage can stall the fan, and it won't re-start until that voltage is raised significantly. For a 4-pin fan, the voltage supply on Pin #2 is always the full 12 VDC, and it is NOT pulsed. (This type of "PWM fan" used in computer cases is NOT the same as DC motors commonly used in other applications with PWM motor drives. In those, the actual DC supply from drive to motor is pulsed. That is the reason that the makers of computer "PWM fans" tell you NOT to use them with a standard PWM drive system from other designs that actually would send pulsed power to the fan.) Within the computer PWM fan case there is a small chip that uses the PWM signal supplied on Pin #4 to modify the flow of current from the constant 12 VDC supply line through the motor windings, thus altering the average magnetic field force and changing the rotor speed.

The design of the newer 4-pin PWM fan system has a couple of backward compatibility features to ensure some success if you mix the two systems. First, the connectors and electrical signals are as similar as possible so that you CAN plug either fan type into either header type. If you connect a 3-pin fan to a 4-pin header that is using the new PWM Mode for control, the fan will receive a fixed 12 VDC supply from Pin #2 and will not receive the new PWM signal becasue it has no Pin #4. Besides, that fan has no special chip and could not modify current from the DC supply, anyway. So the fan will run full speed all the time with no harm other than it might wear out a little faster. You get good cooling - maybe too much - and higher noise than if the fan were able to slow down. The other mis-match is plugging a 4-pin fan into a 3-pin header that can only operate in Voltage Control Mode (aka DC Mode). That fan would receive no PWM signal becasue there is none and there is not even a 4th pin on the mobo header for it. It does receive from Pin #2 a DC supply that varies from 12 to 5 VDC, and that can NOT be modified by the chip in the casing because that chip gets no PWM signal. So the motor windings receive a NON-pulsed voltage which varies, and that makes it behave exactly as a 3-pin fan motor would. Its speed IS controlled by these signals, with no harm to the motor. The main impacts of this mis-match is that you lose two advanteges of the PWM design: they can be run at lower minimum speeds without stalling, and they can be started at a lower speed.

OP's original proposal is the correct way, I believe. That was to use SPLITTERS (not a HUB) to connect all four case fans to a single mobo case fan header, and ensure that the header is configured to use only Voltage Control Mode (aka DC Mode). That is the ONLY Mode that can control the speed of 3-pin fans. The backwards compatibility features inherent on the newer 4-pin fan design ensure that those fans also will be controlled by this power system. OP appears to have given correct consideration to an important limit in this proposal. Almost any fan header can supply up to 1.0 A max current to the fan(s) connected to it, and OP says the total load of the 4 fans is 0.68 A. As a hint, finding a four-output fan Splitter can be difficult, although I have seen ones that look like small printed circuit boards that can do this. The other simple option is to buy three 2-output Splitters and plug two of them into the outputs of the third. The resulting "stack" converts a single mobo fan header into four outputs.

 

[Darkbreeze]

That's basically, exactly what I said. 3 pin fans will go full speed. That is not desirable in most cases unless completely unavoidable. Full speed is better than no fan.

Your assumption that there is no harm to the motor is senseless and I don't see any evidence to the contrary. If you give a motor designed for ONLY a specific range of voltage, a voltage that is outside that range, you WILL cause damage to the motor. THAT is irrefutable. No person with substantial electronics knowledge would try to argue that it's ok to feed a voltage to a motor, of ANY kind, that it was not expressly designed to accomodate.

So, it is NOT recommendable to run a PWM fan on a 3 pin voltage controlled header that will drop voltages considerably out of range. I've seen hundreds of motors, and not just on PWM cooling fans, burn up from prolonged use at voltages they were not designed to be used at, not to mention the damage that can occur to a motor trying to start back up with too low of voltage or start back up after stalling. Ever hear a motor that is unable to start and all you hear is the motor pulling but not running. Yeah, that's not good either.

There WILL be some variance allowed, because there are even low noise adapters that come with many of these fans dropping the voltage to around 7-8v in order to reduce maximum speeds (And thus, noise levels. This is mainly a gimmicky marketing tactic but has some real world benefits to a small niche sector) but when you go below those voltage levels then you are getting into a range the motor was not designed for. Certainly there are all kinds of motors, and some of them are designed for variable (Very low to fairly high) voltages. I have never seen a PWM PC fan that was that way though. Running any motor (Not just on these fans) outside the specifications it was designed for is a bad idea, regardless of who says otherwise.

If you don't mind replacing fans semi-regularly, sure, go for it. I'd rather not.

 

[BetosBurrito]

Just as an update for my own particular use case, i was able to set the AIO_PUMP header to PWM and control the fan speed from BIOS. Since it seems to work fine, I have the 3 pin fans split off the CHA_FAN header and the 4 pin fans off the AIO_PUMP fan (with the CPU fan on the CPU_FAN header). The only quirk I've noticed so far is that the 4 pin fan will run high when the computer first starts up, but normalizes afterwards.

 

[Darkbreeze]

PWM fans and headers running fast at initial startup is by design. Often these PWM fans don't like to start well with low speeds, they will stall or bog down. So they ramp up the voltage for a second or two to get them going and then cut down to the control speed. Totally normal, and a good thing for the PWM motor.

Glad it worked out with that other header. That configuration should work fine for you.

 

[kcstrom]

I was very curious if running a 4pin fan at lower voltage would actually damage it as I have a mixture of fans and one header, so I looked up the spec sheet for the fan I was considering doing this with: https://www.arctic.ac/us_en/f12-pwm-pst.html . This clearly shows that it is intended to run between 5 volts and 12 volts. You can see a chart there that shows the fan speed for different voltages and pwm combinations. If you're controlling with voltage, it's 100% pwm and whatever voltage is being applied. Some fans may not be made to do that, so check the specific fans specifications if you want to know for sure.

 

[Darkbreeze]

Just because the fan is capable of it does not mean that a non-standard voltage won't harm or cause problems with the PWM controller itself if it is not designed to operate at a voltage other than 12v. Some might be, some might not be. On motherboards that offer the option to switch between DC and PWM mode, it's unlikely to be a problem since it's already capable of variable rate control in DC mode anyhow, but then you really wouldn't NEED to do that anyway, since it could work either way.

For boards that can't do that, who knows. It probably depends on the design of the PWM controller on that board.

 

[mamasan2000]

Fan controllers used to be normal back in the day, motherboards didn't have many headers. My fan controller has lasted for probably 15 years now.
Some examples: https://www.thetechlounge.com/best-fan-controller/
I don't know anything about the new ones but it is an easy and relatively cheap solution.

I know @Darkbreeze already mentioned those in the beginning of this thread.
But now you know those things are called Fan Controllers and can search and research on your own.