Issues w/ NZXT Fan Controller

[jdcranke07]

I'm trying to use eight Noctua NF-S12B 1200 PWM fans per channel, for a total of 24 fans (8/channel with three quad 120mm rads in push/pull).

The channels work flawlessly with one fan to each. However, that changes when going up to eight.
Channel 2 works well, but all other channels either don't detect when the eight fans are connected or they are partially detected. All fans are run @ 100% when not detected by the controller and they run @ 40% when partially detected. (40% is minimum speed before voltage gets too low and shuts off fans). I cannot change settings in either case.

I've done the math, assuming each header allows up to 1amp & I'm not miscalculating, I should be hitting about 60% usage of total wattage & amperage per channel w/ eight fans on each said channel.

Each fan uses 0.075A & 0.9W @ 12V.

Not pertinent to the issue, but for a visual...This is for two watercooled systems within my DIY rack enclosure. The three 480mm radiators are in a push/pull config & mounted to the top of the enclosure as exhaust.

Any help is greatly appreciated. Also, if you have recommendations on externally powering fans outside of this controller, I'm all ears for suggestions.

Edit: I'm using Phobya 1 PWM to 9 PWM Y-cables to connect the fans to each channel on the controller.

Edit: Fan controller == NZXT Sentry 3

 

[jojesa]

Which NZXT fan controller?

 

[jdcranke07]

Sorry, forgot that important piece of information. It's the NZXT Sentry 3. Updated OP with it.

 

[jojesa]

It looks like fans are turning off because they are not getting enough power.
Start at 100% and go down from there (90%, 80%, etc) to find the percentage required to have all the fans running.

When fans are not detected NZXT Sentry applies full power to the fan header in an attempt to restart it, that is why they will run at 100%

 

[jdcranke07]

It looks like fans are turning off because they are not getting enough power.
Start at 100% and go down from there (90%, 80%, etc) to find the percentage required to have all the fans running.

When fans are not detected NZXT Sentry applies full power to the fan header in an attempt to restart it, that is why they will run at 100%

I don't have a problem w/ the fans turning off. I just have a problem w/ detection of said channel when I go up to 8 fans each channel. I might not have said it clearly, so that's my fault. What I meant is that I never lose power to the fans. I just lose the ability to turn the fan speed up or down on channels 1, 3, 4, 5 when they have 8 fans connected. Channel 2 has no issues w/ 1 or 8 fans. So, I'm not sure how this can be a power issue if NZXT told the true specs of their hardware & if I calculated the amount of power correctly.

From what I know, NZXT claims that each channel gets 15W per channel w/ a total combined max of 75W. This is assuming, based on average motherboard knowledge, that each PWM header on the controller can supply up to 1A. This combined w/ my Noctua fans pulling only 0.075A & 0.9W per fan, theoretically I should be able to power about 13 fans at full capacity per channel. I, however, am only attempting to run 8 fans, which would be about 61% total throughput. That being if I was able to calculate that correctly & I haven't missed a critical factor.

 

[jojesa]

Your calculations are spot on.
Could it be a faulty unit.

 

[Paperdoc]

I tried to find a Phobya Splitter that proves 9 outputs from one input and could not. So I'm guessing you have "stacked" some Splitters to get this. What I was looking for is a detail of design, so maybe YOU can look at your actual units for this.

Any mobo header (or Controller output header) can deal with a fan speed input from only ONE fan. The signal is a series of 2 pulses per motor revolution, and is returned from the motor to the header on Pin #3. So, any modern Splitter or Hub will only send to its host header the speed signal from ONE of its fans, and ignore the rest completely.

This leaves us with two things to check.
1. Some of the earliest Splitter designs ignored this and simply connected ALL leads on ALL output male connectors to the input connector, thus feeding speed pulses from all fans to the header. This caused huge confusion, bad readings, and many erroneous attempts to "control" such fans. So check that your Splitters do not have this failure if design. How? With Splitters there are two common ways NOT to send excess fan speed signals back. On some, the wires between the input female and output male connectors simply do no include any wires for Pin #3 for most of the outputs. But ONE output DOES have that wire in place. The other way us to have all the wires to all the male outputs, but to OMIT the actual male pin for the #3 position in all of the outputs except one. IF your Splitters have one of these means of limiting the speed signal return, you are OK.

2. Now, IF your Splitters are OK, then still another item to check. You have not provided full details, but MAYBE your system (stacked Splitters?) provides more than 8 outputs. Now, for each group of 8 fans on one Splitter system you should trace out through the connection lines exactly which single male output connector CAN feed its fan's speed back to the mobo controller header. Each controller header in use MUST receive a speed signal from one fan. So, if your connections do NOT have a fan plugged into the only output that CAN send a signal back, the controller channel will receive NO speed signal. THAT signal is how the controller channel detects the presence of a fan - does that "fan" (or, in your case, group of fans) have a speed or not? No speed signal = there is no fan there! So, check the way you have plugged in the fans. IF you have any unused output connectors, maybe move a fan to that one and see if things change.

 

[jdcranke07]

I tried to find a Phobya Splitter that proves 9 outputs from one input and could not. So I'm guessing you have "stacked" some Splitters to get this. What I was looking for is a detail of design, so maybe YOU can look at your actual units for this.

Any mobo header (or Controller output header) can deal with a fan speed input from only ONE fan. The signal is a series of 2 pulses per motor revolution, and is returned from the motor to the header on Pin #3. So, any modern Splitter or Hub will only send to its host header the speed signal from ONE of its fans, and ignore the rest completely.

This leaves us with two things to check.
1. Some of the earliest Splitter designs ignored this and simply connected ALL leads on ALL output male connectors to the input connector, thus feeding speed pulses from all fans to the header. This caused huge confusion, bad readings, and many erroneous attempts to "control" such fans. So check that your Splitters do not have this failure if design. How? With Splitters there are two common ways NOT to send excess fan speed signals back. On some, the wires between the input female and output male connectors simply do no include any wires for Pin #3 for most of the outputs. But ONE output DOES have that wire in place. The other way us to have all the wires to all the male outputs, but to OMIT the actual male pin for the #3 position in all of the outputs except one. IF your Splitters have one of these means of limiting the speed signal return, you are OK.

2. Now, IF your Splitters are OK, then still another item to check. You have not provided full details, but MAYBE your system (stacked Splitters?) provides more than 8 outputs. Now, for each group of 8 fans on one Splitter system you should trace out through the connection lines exactly which single male output connector CAN feed its fan's speed back to the mobo controller header. Each controller header in use MUST receive a speed signal from one fan. So, if your connections do NOT have a fan plugged into the only output that CAN send a signal back, the controller channel will receive NO speed signal. THAT signal is how the controller channel detects the presence of a fan - does that "fan" (or, in your case, group of fans) have a speed or not? No speed signal = there is no fan there! So, check the way you have plugged in the fans. IF you have any unused output connectors, maybe move a fan to that one and see if things change.

I'm pretty sure that I'll just need to check for the 2nd possibility since these are literally just a 1PWM male to 9PWM female "Y-Cable"/splitter.

Here's the link for the cables:
https://www.amazon.com/gp/product/B0151CY9T4/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

Update: Checked both y-cables. The one had a full 4-pin that was unoccupied. Plugged a fan on it & it is fixed. Thank you both for your input!

 

[Paperdoc]

Glad that helped, and thanks or Best Solution. Also, thanks for the link - I've never seen a Splitter like that! I note in its photo that ONE output arm has a white paper label around it. I'm guessing THAT is the one with the full four pins, and the label MAY say it is for the CPU fan, but that's not necessary unless you really are using this Splitter to power a CPU cooler device from a mobo CPU_FAN header.