Can a proper PWM Splitter control 2 Fans with different RPMs?

[Satyajit Mishra]

I have a proper PWM splitter i.e. one of the plug has 4 pins & another one doesn't have the 3rd pin (which carries the RPM signal to the motherboard). Suppose the plug with 4 pins is "Master" & the plug with 3 pins is "Slave". I hear that the RPM of the Fan connected to the Master plug is synchronized with RPM of the Fan connected to the Slave plug.
I have 2 Fans; one has 2000 RPM (Max) & the other has 2500 RPM (Max).
In this scenario the 2500 RPM Fan is connected to the Master & 2000 RPM one is connected to the Slave.
If my motherboard were to ramp up the RPM of the Master fan to 100% i.e. 2500 RPM, what will happen to the Slave fan (as the slave fans max RPM is 2000)?

Reference Links:
https://images10.newegg.com/ProductImage/12-162-026-02.jpg
http://cdn.overclock.net/c/cb/cb05a4a6_closeupviewofthe2maleconnectors.jpeg
https://forums.tweaktown.com/attachments/asrock/7171d1409068284-mb-wont-2-cpu-fans-spin-pwmsplittersmall-png
https://forums.tweaktown.com/asrock/58619-mb-wont-2-cpu-fans-spin.html
http://www.overclock.net/t/1263922/quick-question-on-connecting-two-pwm-fans-to-the-same-motherboard-header

 

[CountMike]

RPM detection pin is missing so it would not confuse sensors in case of 2 different fans. Both are controlled by same channel so there's no way controller can control them separately. Sensor will detect RPMs on that fan that is connected to 4pin and both will get same PVM signal.

 

[Satyajit Mishra]

RPM detection pin is missing so it would not confuse sensors in case of 2 different fans. Both are controlled by same channel so there's no way controller can control them separately. Sensor will detect RPMs on that fan that is connected to 4pin and both will get same PVM signal.

I understand that. But I want to know what happens whens 2 fans with different RPMS are used in this scenario? If the slave fan's rpm is lower than the master fan, what will happen? Or does the pwm controls RPM in terms of percentage i.e. master fan's 100% is 2500 RPM & slave fan's 100% RPM is at 200.

 

[CountMike]

To understand that you have to understand how fans are controlled. 3pin fans are controlled by voltage applied to fan thru middle wire. PWM stands for Pulse Width Modulation and controls fan speed by sending pulse thru 4th wire, pulses of various lengths and frequency. Faster and longer pulses make a fan go faster. All the while voltage on +12 pin stays even.
Slave fan speed is not recorded by sensor so it can't determine its speed And will get same pulses as the "Master" fan so no speed correction can be applied.
All the voltages and pulses come from one header on the MB and can't be separately adjusted.

 

[Satyajit Mishra]

To understand that you have to understand how fans are controlled. 3pin fans are controlled by voltage applied to fan thru middle wire. PWM stands for Pulse Width Modulation and controls fan speed by sending pulse thru 4th wire, pulses of various lengths and frequency. Faster and longer pulses make a fan go faster. All the while voltage on +12 pin stays even.
Slave fan speed is not recorded by sensor so it can't determine its speed And will get same pulses as the "Master" fan so no speed correction can be applied.
All the voltages and pulses come from one header on the MB and can't be separately adjusted.

I understand the concept. So PWM will determine RPM of the slave fan according to the master fan in terms of percentage i.e. if the master fan is running at 80% RPM, then the slave fan will also run at 80% RPM.

 

[CountMike]

Exactly.

 

[Satyajit Mishra]

Exactly.

Good to know. Thanks for clearing that up.

 

[CountMike]

BTW, same would happen if they are voltage controlled. Both would get voltage determined by master's speed (RPM).

 

[Paperdoc]

It might help if you understand this. A mobo automatic control system is totally focused on controlling the TEMPERATURE at a sensor - for case ventilation fans connected to a mobo SYS_FAN or CHA_FAN header, this is a sensor built into the mobo. To do that it manipulates the source of cooling for that area, which means the speed of the fans on that header. If the measured temperature is too high, it will send out a signal that increases fan speed, and will do that in further increases as long as the temp is too high. When temp is just right it will hold the current setting. When temp decreases, it will slow the fan down. It uses ONLY the measured temperature to decide what to do. It does NOT pay any attention at all to the speed of the fan as reported on Pin #3 of the header. The actual fan speed is NOT required or used to control the fan.

The mechanism of how fan speed is changed differs between the two fan designs. For 3-pin fans, the mobo header changes the DC voltage supplied to the fan on Pin #2 from 12 VDC max down to about 5 VDC min. (Less voltage may cause a fan to stall and be hard to re-start.) A 4-pin fan gets a constant +12 VDC supply from Pin #2, but it also gets the new PWM signal from Pin #4. Inside this fan type there is a small chip that uses the PWM signal to modify the DC supply flowing through the motor windings, thus changing its speed. But these are just details of mechanism - the "deciding" part of fan control is as above - based solely on temperature.

So, why have fan speed reported? Well, for interest / curiosity of users. Also, SOME users prefer to do fan control manually using third-party fan controller modules, but for the most part they base they decisions on the fan speeds indicated, because they don't have easy access to measured temperatures. But perhaps more importantly, the mobo uses those speed signals for a different purpose - failure detection. If a fan completely fails to report its speed (or, in some mobos, if the speed reported falls below some minimum setting), the mobo will send out an alarm to the user and often will try to solve the problem by sending the fan a "full speed" signal to try to re-start it. In the specific case of the CPU cooler (its temperature sensor is a different one actually inside the CPU chip), SOME mobos will take stronger action: they will send out on-screen and "beep signal" warnings, and shortly thereafter will shut down the entire system without even waiting for the temperature as measured inside the CPU to go too high. This is all to be sure the CPU does not overheat rapidly and get damaged. So the speed signal from the fans has important uses; it just has NO ROLE in controlling the fan speed.

After all that, back to your real question. With two different fans connected by a Splitter to one mobo header which is using PWM Mode to control its fan(s), the mobo does NOT know there are two fans out there. It is providing only one signal at the header pins. And it bases that signal entirely on the temperature it is trying to control. So it is up to the fans to do whatever they do given that BOTH of those fans will receive the same signal. In fact they will NOT run at the same speed. VERY roughly they will be trying to run at the same percentage of their maximum speed (because they both get the same PWM signal), but that won't be precise. And the mobo does NOT care what those two speeds are, and does not even consider that factor in deciding what signals to send out on its header. So, the facts that the fans are different, and that the Splitter does not even send the speed signal of the second fan to the mobo, do not matter at all.