[SOLVED] Will have a 4pin and a 3pin case fan on the same header through splitter, should I use DC or PWM mode to control them both?

cloudropis_1

Honorable
Dec 16, 2016
63
2
10,535
0
I have a gigabyte b450m ds3h mobo, which only has one 4pin header for case fans, so I bought a splitter to run my two 3pin case fans on it. One of them is about to kick the dust and I'm about to buy an arctic p12 to replace it, meaning that I'd have my old 3pin and the new 4pin on the same splitter.

Can I control the curve of both considering one need PWM and the other DC? If yes, do I need to put the header in PWM mode or DC mode? From my understanding, in PWM mode I'd have full bios control for the 4pin fan but the 3pin would be at 100% all the time, so I guess DC will let me control both (although with less granularity for the 4pin compared to PWM mode)?
 

DRagor

Illustrious
3 pin fan can't be regulated via PWM - it will run 100% all time. However, 4 pin fan CAN be regulated via DC. Sure they don't like it, but it is possible. Splitter changes nothing on the matter. Using splitter only means you have no signal from second (and next) fans so mobo can't detect any malfunction or stall, but it does not change how are they controlled. You just need to make sure 3 pin fan is connected to control wire of the splitter (the one with more pins) and set fan curve such way that 4 pin fan never stalls.
 
You can't use 3pin and 4pin fans on same header (with splitter) and get them regulated.
All fans on same header have to be of same type.
Connect 3pin fans to one header (no more than 3 fans on same header), 4pin fan to a different header.

If your board has ony single fan header, then you'll have to replace all your fans.
 
Last edited:
Reactions: cloudropis_1

cloudropis_1

Honorable
Dec 16, 2016
63
2
10,535
0
You can't use 3pin and 4pin fans on same header (with splitter) and get them regulated.
All fans on same header have to be of same type.
Connect 3pin fans to one header (no more than 3 fans on same header), 4pin fan to a different header.

If your board has ony single fan header, then you'll have to replace all your fans.
Well that's unfortunate, guess I'll have to replace both.
 

DRagor

Illustrious
3 pin fan can't be regulated via PWM - it will run 100% all time. However, 4 pin fan CAN be regulated via DC. Sure they don't like it, but it is possible. Splitter changes nothing on the matter. Using splitter only means you have no signal from second (and next) fans so mobo can't detect any malfunction or stall, but it does not change how are they controlled. You just need to make sure 3 pin fan is connected to control wire of the splitter (the one with more pins) and set fan curve such way that 4 pin fan never stalls.
 

Paperdoc

Polypheme
Ambassador
SkyNetRising got it wrong. DRagor is right.

The only way to control the speed of an older (3-pin) fan is to vary the voltage supplied to it on Pin #2. This method of speed control is called Voltage Control Mode or DC Mode. The new 4-pin PWM fan style was designed with some backwards compatibility features to make their introduction into the market easier. One of these it this: if you connect one to a mobo header that is using the older Voltage Control Mode, it gets NO PWM signal from Pin #4 so its special internal chip can NOT modify the power supplied from Pin #2. In that situation the only thing the motor can do is run according to the voltage it does receive. Thus its speed IS controlled and you CAN mix this new 4-pin fan on the SAME circuit as an older 3-pin fan as long as the signals to them are Voltage Control Mode. This method is not quite ideal for use with a PWM-style fan from a technical standpoint, but it does work as designed.

A small second point. A mobo fan header receives from its fan via Pin #3 the speed signal, consisting of 5 VDC pulses, 2 pulses per revolution, and it counts those to determine fan speed. When you connect two or more fans together to one header you can NOT have all of those speed pulse train signals arriving at the header. Their overlap causes huge confusion and errors in speed measurement. So any Splitter or Hub will send back to its host header the speed signal from only ONE fan, and ignore all the rest. The easiest way to do this with a simple Splitter is to omit Pin #3 from the male output connector for every output EXCEPT one. That way only one fan can send back a speed signal. When you do this, it really does NOT matter which fan's speed is sent back. That info is NOT used by the header for speed control. It IS used to monitor the fans for FAILURE (no speed signal), but of course that failure monitoring can be done ONLY for the one fan whose signal is returned though the Splitter. So from time to time YOU should check that all fans are still working.
 

cloudropis_1

Honorable
Dec 16, 2016
63
2
10,535
0
SkyNetRising got it wrong. DRagor is right.

The only way to control the speed of an older (3-pin) fan is to vary the voltage supplied to it on Pin #2. This method of speed control is called Voltage Control Mode or DC Mode. The new 4-pin PWM fan style was designed with some backwards compatibility features to make their introduction into the market easier. One of these it this: if you connect one to a mobo header that is using the older Voltage Control Mode, it gets NO PWM signal from Pin #4 so its special internal chip can NOT modify the power supplied from Pin #2. In that situation the only thing the motor can do is run according to the voltage it does receive. Thus its speed IS controlled and you CAN mix this new 4-pin fan on the SAME circuit as an older 3-pin fan as long as the signals to them are Voltage Control Mode. This method is not quite ideal for use with a PWM-style fan from a technical standpoint, but it does work as designed.

A small second point. A mobo fan header receives from its fan via Pin #3 the speed signal, consisting of 5 VDC pulses, 2 pulses per revolution, and it counts those to determine fan speed. When you connect two or more fans together to one header you can NOT have all of those speed pulse train signals arriving at the header. Their overlap causes huge confusion and errors in speed measurement. So any Splitter or Hub will send back to its host header the speed signal from only ONE fan, and ignore all the rest. The easiest way to do this with a simple Splitter is to omit Pin #3 from the male output connector for every output EXCEPT one. That way only one fan can send back a speed signal. When you do this, it really does NOT matter which fan's speed is sent back. That info is NOT used by the header for speed control. It IS used to monitor the fans for FAILURE (no speed signal), but of course that failure monitoring can be done ONLY for the one fan whose signal is returned though the Splitter. So from time to time YOU should check that all fans are still working.
DRagor said "a 3pin fan can't be regulated through PWM, it will run at 100% all the time" but, if I understood your post right, it can be regulated through DC mode (rather than PWM) by controlling voltage. So:

3pin on DC mode, it's controlled through voltage in the second pin. 3pin on PWM mode, it runs 100% all the time.
4pin on DC mode, it doesn't use the PWM pin and it will default to regulating through voltage, like a 3pin, which is acceptable but not optimal. 4pin on PWM, works as intended by regulating through the PWM pin.

Did I get it right? My only worry is to not have a case fan running at 100% all the time, so if I understood correctly I can just put the header on DC mode and rest assured it won't, while losing some precision on the 4pin fan which I can tolerate
 

DRagor

Illustrious
Did I get it right?
Correct
My only worry is to not have a case fan running at 100% all the time, so if I understood correctly I can just put the header on DC mode and rest assured it won't, while losing some precision on the 4pin fan which I can tolerate
Correct. It may also shorten the life span of PWM fan but that should not be a big deal.
 
Reactions: Paperdoc

ASK THE COMMUNITY