Daisy chaining mixed 4-pin fans

[Shrek]

When daisy chaining 4-pin fans, only one tachometer pin is active (to avoid mixed signals), but what if one was to daisy chain mixed fans, say 120mm and 140mm, should one take the tachometer signal from the smaller or larger fan?

I have a case that takes a 120mm exhaust fan and on top can carry two 140mm fans, all to be controlled by a single 4 pin connector on the motherboard.

 

[eidairaman1]

When one daisy chains 4-pin fans, only one has the tachometer pin active (to avoid mixed signals), but what if one was to daisy chain mixed fans, say 120mm and 140mm, should one take the tachometer signal from the smaller or larger fan?

I have a case that takes a 120mm exhaust fan and on top can carry two 140mm fans, all to be controlled by a single 4 pin connector on the motherboard.

Whats the max rpms for the 120 and the 140 mm?

 

[Shrek]

2200 RPM for the 120 mm

1500 RPM for the 140 mm

 

[nomdeplume]

Whats the max rpms for the 120 and the 140 mm?

He suggests a set of these were recently purchased. Yet to further an explanation of how "daisy-chaining" will be handled.

Short of it this doesn't work well if at all on a lower end board. BIOS will express it's disregards in one of a few highly suggestive manners.

 

[Shrek]

Will the BIOS even know? it is getting a speed reading from one fan alone.

 

[Dr. Dro]

Will the BIOS even know? it is getting a speed reading from one fan alone.

No, and yes, it is. The problem with fans that while the voltage supply from the motherboard is the same, the speed that each motor will run at different speed.

Assuming a linear response to voltage supply, similar blade design and air throughput with a 12 volt input, a theoretical fan A runs at 2000 RPM and fan B at 2200 RPM, at 8 volts fan A will run at 1320 and fan B at 1450, this will obviously result in mismatched noise levels and possibly, a pressure imbalance. It's best to get similar fans that work as a set if it can be helped, or use a fan controller that allows for individual setting for each port.

 

[nomdeplume]

BIOS handles the fan controller even if programming it is being overruled by third party software. The issue is cascading electrical responses the board doesn't expect will only extrapolate with mixed fans. Honestly, lower end boards are questionable powering and feeding a signal into even two identical fans drawing less than .5A peak.

Legacy board designed to barely support intake/cpu/exhaust are not a good platform to control amount of fans modern cases will mount.

 

[eidairaman1]

2200 RPM for the 120 mm

1500 RPM for the 140 mm

Well to reach those faster speeds requires different input voltages.

You migh consider a fan controller that can handle those fans

 

[LittleBro]

PWM control will work fine, that's independent of fan size. As for RPMs, this may or may not work properly.
I did similar thing years ago and the reporting results from BIOS were sometimes correct, sometimes badly inaccurate.
Even if you have "same" fans, their RPM will not be the same. Every manufacturer states kind of limits for RPMs (+/- speed % tolerances).
Thus, daisy chaining RPM signal from multiple fans of "same" type will not yield the most accurate results.

Pin for RPM speed does not work as 1-wire bus, fans are not addressable.
To get most accurate results, pick one fan and remove RPM pin connection from others.

 

[Shrek]

To get most accurate results, pick one fan and remove RPM pin connection from others.

That is what the daisy chain cables do

 

[pavle]

How much current can that motherboard connector carry?
It doesn't matter if you have different fans even on 4th pin - they will rotate at different speed (at the same PWM input) if all have different max_speed.

 

[Shrek]

Good question, the manual does not say.

Well to reach those faster speeds requires different input voltages.

You migh consider a fan controller that can handle those fans

I believe that with 4-pin fans the input voltage stays the same but the PWM changes, so the BIOS will be driving them all at the same percentage.

So, I wonder what the BIOS does with the tachometer signal? is that just to know if the fan is still running?

 

[_roman_]

That is what the daisy chain cables do

Assuming you use a PWM connector with enough AMPS on teh mainboard.

the first fan should be a 4 pin pwm fan. All other fans are invisible to the mainboard because of the extension cable you mentioned. So when a fan is dead or have bad connection it will just not spin. You will only see it visiually because there is no fan speed signal. Of course the first fan is visible to the mainboard, that includes also the fan speed signal.

4 fans hsould be fine - but you better check if e.g. 1 amps * 12V DC = 12W DC ... is bigger as all 4 fans combined (check amp ratings or Watt ratings before)

--

I recommend using two pwm ports.
2 fans with extension cable for the bigger fans / 2 fans with extension cable for the smaller fans.

I believe that with 4-pin fans the input voltage stays the same but the PWM changes, so the BIOS will be driving them all at the same percentage.

pwm is a different operating mode. small logic determines the fan speed with the pwm signal. pwm is some sort of signal - in short it shows how many percentage a fan should run. please check the web for a detail explanation.

dc is a different operating mode. =direct current voltage determines the fan speed.

 

[LittleBro]

Good question, the manual does not say.



I believe that with 4-pin fans the input voltage stays the same but the PWM changes, so the BIOS will be driving them all at the same percentage.

So, I wonder what the BIOS does with the tachometer signal? is that just to know if the fan is still running?

Shame they don't state max. current draw per fan header. At least ASUS clearly states it's 1A per fan header.

IMHO, even Gigabyte does at least 500mA per fan header, so 4 fans should be pretty much okay. I mean Arcitc P14 fan uses draws maybe 0.12A at 12V at full speed (1700 RPM).
Still, you can go with SATA/MOLEX to multiple fan 4-pin adapter to be perfectly safe. Akasa sells some, they read RPM only from one 4-pin which makes perfect sense.

BIOS/board and/or fan itself might have PWM signal limitation, e.g. minimum and maximum PWM speed.
My current board can't spin fans at lower than 20% PWM (you can set minimum of 20% PWM) but my fans support 0db mode when PWM is <5%, so they never turn off.

BIOS reads output from IC on board which calculates RPM based on pulses from sensor of a fan. Noctua states that the tachometer signal consists of 2 pulses (high/low) per one revolution. Logically, if you let 4 fans at different speed to report pulses through the same wire, it will be a mess, because period of pulse will vary. IC will have troubles to identify start and end of a pulse. That means signal will be barely readable or not readable at all.

 

[nomdeplume]

My educated guess is 2x 140mm will sporadically work for indeterminate amounts of time before mobo errors out or they stop. Three unmatched fans on a single header will be a crapshoot even in DC mode. Probably a few soft fails solved by a reboot and who knows from there on a board that has no aspirations towards...

Theoretical load behaviors are trumped by intentionally implemented points of early failure that allow for salvage value to remain high. At least your prebuilt uses aftermarket mobo that can be replaced from innumerous sources.