[SOLVED] PWM Fan Controller for mixed size fans 120/140mm

SteelMouse

Distinguished
Dec 21, 2013
49
2
18,545
1
Hello,

I have 9 Noctua iPPC fan's in the case for cooling and air flow. It's a mix of 140's and 120's

120's : NF-F12 industrialPPC-3000 PWM (https://noctua.at/en/nf-f12-industrialppc-3000-pwm)

140's: NF-A14 industrialPPC-3000 PWM (https://noctua.at/en/nf-a14-industrialppc-3000-pwm)

I am planning to get this fan controller:
Deepcool FH-10 10 Port Fan Hub (https://www.deepcool.com/product/dcoolingaccessory/2017-11/12_7223.shtml)

And planning to connect 3 x 140's in the front intake + 2 x 120's in the bottom intake to the fan controller (Deepcool FH-10).

And both the 140's and 120's have a max RPM of 3000. But different Max. Input Power and Max. Input Current.

NF-F12 industrialPPC-3000 PWM:
Rotational speed (+/- 10%) : 3000 RPM
Max. input power : 3,6 W
Max. input current : 0,3 A
Operating voltage : 12 V

NF-A14 industrialPPC-3000 PWM:
Rotational speed (+/- 10%) : 3000 RPM
Max. input power : 6,6 W
Max. input current : 0,55 A
Operating voltage : 12 V

Question 1:

Would this set up of connecting 3 x 140's + 2 x 120's to the same fan controller work?

The fan controller (Deepcool FH-10) can only read one fan speed, so I assume the other ports will get similar speed as the speed set on the motherboard 4-pin fan port that's connected to FAN1 of the fan controller.

Considering both the 140's and 120's have the same max RPM of 3000, would this setup work and can I get all the intake fans to run at the same speed?

If this setup would work, then I would proceed to buy a second fan controller, connect all exhaust fans to it, then control the speed of all exhaust fans with the second fan controller, so all the exhaust fans will run at the same RPM.


Question 2:

Can this fan controller (Deepcool FH-10) handle the Noctua 140's (6.6W Max Input Power / 0.55 A Max Input Current)?

I won't be using a splitters and each port on the fan controller (Deepcool FH-10) will be connected to only one fan.


Question 3:

Trying to get positive air pressure in my case to reduce the dust buildup, that's the reasoning behind this effort.

Maybe 5 intake fans running at 2000 RPM vs 4 exhaust running at 1200 RPM will create a decent enough positive air pressure.

Considering intakes has 3 x 140 + 2 x 120 vs exhaust of 4 x 120, should be possible since the 140's have more airflow than the 120's. But intakes do have the dust filter in front for the fans to deal with.

So if the fan controller will work (Question 1), then does this setup for positive air pressure look good


Question 4:

Is this fan controller a good one (Deepcool FH-10)?

And if possible, kindly suggest me some good fan controllers for controlling 4 pin PWM fans.


Thank you for your time, I appreciate it a lot.
 

Paperdoc

Glorious
Ambassador
#1 You are entirely correct in all of this.

#2 I fully agree, the probability of trouble is near nil. Most of the electrical flow in the PSU output wires is DC, although that can have fluctuations as components' power draws change. It is the CHANGE of current flow that can induce noise signals in adjacent wires. But a fan is rarely affected by such small noise in its power supply.

From the other direction, can current flows inside the fan motor affect the DC power flow and induce noise in the PSU's output wires and feed that to mobo, graphics card, etc? Well to start, the current for a fan is small, so any noise it could generate is small. Beyond that, virtually all power consumer devices in your system have filters on their power input lines to ensure they do NOT "inherit" moise in that fashion. Noise pickup that affects computer components almost always is though low-voltage analog or digital signnal lines with poor shielding, and NOT through power supply lines.
 

Paperdoc

Glorious
Ambassador
Just a little jargon to start: the Deepcool device you cite is called a fan HUB, not a Controller. No biggie. And you have provided a very clear list of your components and clear questions. By the numbers

1. For those five intake fans, YES, this will work well. Just a couple notes. The important thing is not the speeds, it is the air flow rate from the fans. For the same actual fan speeds, the 140 mm ones deliver more air flow than the 120's. We talk about fan speed control, but the truth is it is TEMPERATURE control. That is, the mobo control system uses a TEMPERATURE sensor on the mobo to guide its case fan control. It has a target temp, and it manipulates the power signals it sends out to the fans to change the cooling air flow (and thus the resulting temperature) to achieve that target. Although the mobo gets the speed signal from its fan (in the case of a Splitter or Hub, from only ONE fan in the group) and can display that for you, it does NOT use that info in its fan control work. It really does not care what the speed is. (For another purpose, fan FAILURE detection, it does monitor the speed signal.) The net result is that, for different fan models connected together like this, their speeds MAY be different given identical control signals, but that does not matter at all.

2. Yes, the electrical load is OK. It is not clear from the FH-10 web page what that limit is. It's actually the contacts in the SATA power connectors from the PSU that impose a limit of 4.5 A max current on the 12 VDC supply lines - that's total for the entire load on the Hub: 4.5 A at 12 VDC, or 54 W. Your proposed load comes to 24 W for these five fans.

3. I agree with aiming for a small positive pressure, and you are correct that the intake fans' air flow will be less than specs due to the small airflow resistance of the intake dust filters. But that is a modest reduction. Your overall plan is to add four exhaust fans of the 120mm size, with no filters on their outflow. Now, using max airflow specs and fan counts is only a rough approximation to actual air balance. But five intakes vs 4 exhaust (including 2 of the 5 intakes are larger with more airflow) is VERY likely to produce the air balance you seek if ALL of the fans are running the SAME speed. You do not need to ensure that the exhaust fans are running slower. With this in mind, you could connect all four of the exhaust fans to that same FH-10 Hub. It has the ports you need, and this would put all nine fans under the same control signals. Doing this would add 4*3.6 W max load, or 14.4 W to the Hub, yielding a total load of 38.4 W, well below the upper limit of its ability. So you can do it this way if you wish, and not buy a second Hub. IF you choose to get a second Hub for the exhaust fans, I suggest you configure the two mobo SYS_FAN ports you will use for them identically.

4. I don't have one of those Deepcool Hubs, but it certainly appears to be decent quality from a well-known maker.

Just a final note. As I said, a mobo fan header monitors the fan speed signal it receives for fan FAILURE, and warns you if that happens. But when you use a Splitter or Hub, most of your fans do NOT return signals to be monitored this way. Thus you should check all your fans from time to time to be sure they all are still working.

Late edit:
Karadjgne (below) makes a very good point. Nine fans (all of them very high-flow IndustrialPPC units) is a LOT of fans and air flow!
 
Last edited:

Karadjgne

Titan
Ambassador
Just be cognizant of airflow patterns. 9 fans is a lot of possible cfm, but that's not the whole story. You'll want the air to flow, in then out. Many times fans in opposition will defeat that purpose and you end up with a ton of air being moved in a circulatory pattern, moving the heat around the case instead of out of the case. Adding fans is not always for the betterment to case temps.
 
Reactions: Paperdoc

SteelMouse

Distinguished
Dec 21, 2013
49
2
18,545
1
Just a little jargon to start: the Deepcool device you cite is called a fan HUB, not a Controller. No biggie. And you have provided a very clear list of your components and clear questions. By the numbers

1. For those five intake fans, YES, this will work well. Just a couple notes. The important thing is not the speeds, it is the air flow rate from the fans. For the same actual fan speeds, the 140 mm ones deliver more air flow than the 120's. We talk about fan speed control, but the truth is it is TEMPERATURE control. That is, the mobo control system uses a TEMPERATURE sensor on the mobo to guide its case fan control. It has a target temp, and it manipulates the power signals it sends out to the fans to change the cooling air flow (and thus the resulting temperature) to achieve that target. Although the mobo gets the speed signal from its fan (in the case of a Splitter or Hub, from only ONE fan in the group) and can display that for you, it does NOT use that info in its fan control work. It really does not care what the speed is. (For another purpose, fan FAILURE detection, it does monitor the speed signal.) The net result is that, for different fan models connected together like this, their speeds MAY be different given identical control signals, but that does not matter at all.

2. Yes, the electrical load is OK. It is not clear from the FH-10 web page what that limit is. It's actually the contacts in the SATA power connectors from the PSU that impose a limit of 4.5 A max current on the 12 VDC supply lines - that's total for the entire load on the Hub: 4.5 A at 12 VDC, or 54 W. Your proposed load comes to 24 W for these five fans.

3. I agree with aiming for a small positive pressure, and you are correct that the intake fans' air flow will be less than specs due to the small airflow resistance of the intake dust filters. But that is a modest reduction. Your overall plan is to add four exhaust fans of the 120mm size, with no filters on their outflow. Now, using max airflow specs and fan counts is only a rough approximation to actual air balance. But five intakes vs 4 exhaust (including 2 of the 5 intakes are larger with more airflow) is VERY likely to produce the air balance you seek if ALL of the fans are running the SAME speed. You do not need to ensure that the exhaust fans are running slower. With this in mind, you could connect all four of the exhaust fans to that same FH-10 Hub. It has the ports you need, and this would put all nine fans under the same control signals. Doing this would add 4*3.6 W max load, or 14.4 W to the Hub, yielding a total load of 38.4 W, well below the upper limit of its ability. So you can do it this way if you wish, and not buy a second Hub. IF you choose to get a second Hub for the exhaust fans, I suggest you configure the two mobo SYS_FAN ports you will use for them identically.

4. I don't have one of those Deepcool Hubs, but it certainly appears to be decent quality from a well-known maker.

Just a final note. As I said, a mobo fan header monitors the fan speed signal it receives for fan FAILURE, and warns you if that happens. But when you use a Splitter or Hub, most of your fans do NOT return signals to be monitored this way. Thus you should check all your fans from time to time to be sure they all are still working.

Late edit:
Karadjgne (below) makes a very good point. Nine fans (all of them very high-flow IndustrialPPC units) is a LOT of fans and air flow!
Thank you for the detailed answers and technical underlying information. I wish fan hub manufacturer's were this detailed in their description.

Regarding answer no.2, I also noticed that the Deep Cool page says nothing about the electrical load. No info on max watts / max amps per fan header. I was looking into some Amazon reviews of other fan hubs, some real cheap ones, one user said (paraphrasing), the said cheap fan hubs burned up and ruined their pc, something to the effect of that.

Follow up questions:

Question 2.1: (Moot question, just here for information sake)
Provided I power this fan controller via a dedicated SATA cable that's not shared with any HDD's / SSD's and solely powers the fan hub, then I could use this Deep Cool Fan Hub without worrying about melted wires / burning things up?

Would it be perfectly safe to use the fan hub provided the connection to the PSU is setup as a dedicated one?

(Edit: See info below, Corsair Commander Pro has issues with the Noctua iPPC 3000's, so probably other fan hubs might have potential issues).

Question 2.2: (Moot question, just here for information sake)
From your experience, what's the level of risk one has to take with these fan hubs?
If there's some level of risk, I am thinking of alternate scenarios, might as well mothball one or two fans, figure out a setup purely via motherboard headers for max safety.

Info: (For the benefit of fellow readers that might chance upon this post):
Seems the Corsair Commander Pro has issues with the Noctua iPPC 3000 RPM fans. If the Corsair Commander Pro has issues (with it being the most expensive fan hub), then there isn't much hope with any other fan controllers.

Corsair Commander Pro has issues with Noctua iPPC 3000 RPM fans (both 120 and 140).
Source: https://forum.corsair.com/forums/showthread.php?t=189271

Conclusion:
This leaves out all other brand fan hubs, might have to check out the Noctua NA-FC1.
Will contact Noctua Support about possible setup / suggestions and post back here once I receive the reply.
 
Last edited:

SteelMouse

Distinguished
Dec 21, 2013
49
2
18,545
1
Just be cognizant of airflow patterns. 9 fans is a lot of possible cfm, but that's not the whole story. You'll want the air to flow, in then out. Many times fans in opposition will defeat that purpose and you end up with a ton of air being moved in a circulatory pattern, moving the heat around the case instead of out of the case. Adding fans is not always for the betterment to case temps.
Here's the fans setup diagram and my understanding of the possible air flow pattern:

Fans Setup Diagram and Airflow (airflow according to my understanding):



Cabinet Images:






Things to note:
  1. The top exhaust and the NH.D15S are pretty close, the top exhausts are like 5 cm away from the NH-D15S.
  2. The GPU is pretty close to the NH.D15S, like within 2 cm close.
  3. The Noctua NH.D15S has one default 140 in the middle, then I added a Noctua iPPC 3000 RPM 120mm fan in front as PUSH. (not that it's setup to run at 3000 RPM, just stating the fan spec)
  4. One of the bottom intake (the one in the middle of the case), has PSU cables above it, so slight reducting in airflow, but not much of a concern.
  5. The top rightmost exhaust 120 sits directly above the HDD's + SSD. So should exhaust some heat away from the HDD's + SSD.
Airflow Pattern (my understanding):

Air from Front, Side, Bottom comes into the case.
GPU (MSI RTX 3080) has 3 fans pushing in air from bottom into the card.
PSU is exhausting air up into the case (not that PSU gets hot and puts out a lot of hot air, just stating the PSU fan exhaust direction). PSU exhaust directly goes above into the GPU.
Noctua NH.D15S CPU Cooler sucks in air and pushes it out, then rear exhaust 120 that sits directly behind it gets the air out of the case.
The 3 x top 120's exhaust air out of the case.
 
Last edited:

Paperdoc

Glorious
Ambassador
2.1 It is possible to have poor construction or design cause a problem with a fan Hub, and the "cheap" ones are more suspect. I really doubt Deepcool is in that category, and I have never seen a report of such a problem for their products.

You did not quite catch the detail on power limit. The major power output lines from a PSU for peripherals and other accessories are ones with 4-pin Molex or SATA connectors. Normally each such set of wires has several output connectors along one wire set. The wires themselves are normally sized to handle lots of power - well over 10A for most, because the old Molex connectors with heavy pins could deliver a lot. That connector system was deployed in the very first PC's when HDD units and other peripherals used a lot of power. The limit for SATA outputs, though, is NOT in the wires. It is the metal contacts in each connector, partly because the SATA system originally was designed for a new generation of hard drives and optical drives that all required less power than the older IDE devices, but also required in a few cases voltage sources that 4-pin Molex does not have. So the 4.5 A max current limit is for the contacts in EACH SATA connector. In fact, because that is the source of limit, that limit applies separately to the 12 VDC, the 5 VDC, and another (3.3 VDC I think) set of power supply lines in the SATA system. Thus, there is no need to power the Deepcool FH-10 Hub from a SATA power output that is the ONLY one used on a set of wires from the PSU. The WIRES are not the limit. That's perfectly safe and well within the specs of the SATA lines and the Deepcool hub.

2.2 From what I see here on Tom's very few people have problems with Hubs from any of the major makers, as long as they follow the specified limits. By far the most common problems are for people who do not understand how Hubs and Splitters work, and how that impacts where they should be used. None of that has anything to do with product quality.

I had not seen the stories you found regarding problems using a 3000 RPM IndustrialPPC fan from Noctua on a Corsair Commander Pro. So I have no explanation of the cause, and neither did the posters there. I do recognize that the fan we're talking about certainly uses more power that common fans, and I know its internal design is slightly different, but that should not make a difference IF the port it is connected to can deliver the current (just over ½ A) it requires. A simple HUB like the Deepcool unit does not have any complicated electronics in it the way the Corsair Commander Pro does. A HUB like the FH-10 really only makes the correct connections between input and output ports with fixed wires and metal contacts, so the only factors affecting its ability to provide power are the heaviness of the wires and metal contacts, and the ability of the power source the SATA input) to deliver power to the Hub. Providing wires to carry up to 5 A current around inside the Hub is trivial; constructing pins for the output ports to deliver even up to 1 A on a port (there are 10 ports, to the AVERAGE is only 0.45 A per port) is a done deal - ALL mobo headers can supply up to 1.0 A output current per header. There is no reason for the Noctua NF-A14 industrialPPC 3000 RPM fan to have any problem with power from a decent Hub, which I'm sure the Deepcool FH-10 is.

The Corsair Commander Pro is NOT a simple Hub. It is a complex controller system for several types of devices, and it does not merely relay power and signals from the mobo out to its attached devices. It does a lot of control work itself. Its internal details are more complex than a Hub, so designing it is far from trivial. Corsair designed it for use with their equipment, and may not have allowed for the very much higher power needed by fans like the iPPC line from Noctua.

Just as an aside, there are some people who go for even MORE powerful fans originally intended for use in industrial equipment cabinets and high heat applications, and these can use currents five times (or more) what the iPPC ones use. There is NO way to power and control those fans from a mobo header or any common Hub, but they never were intended for use in a desktop computer case.
 

SteelMouse

Distinguished
Dec 21, 2013
49
2
18,545
1
Thanks for the technical info on the STA cable power delivery, so not a risk to use a single SATA cable with multiple connector to power things up.

On the fan controller, what do you think of the Noctua NA-FC1 (https://noctua.at/en/na-fc1)

Designed by Noctua and Noctua states in the NA-FC1 webpage in multiple areas regarding how NA-FC1 works with Noctua iPPC models.
"This makes it ideal for controlling multiple high-speed fans such as Noctua’s 3000rpm industrialPPC models. "

NA-FC1 Specs: https://noctua.at/en/products/accessories/na-fc1
NA-FC1 Manual: https://noctua.at/pub/media/blfa_files/manual/noctua_na_fc1_manual_en.pdf

Question 1:
Regarding the NA-FC1:
NA-FC1 Max. current : 3 A

It has a 3 way splitter and I plan on connecting 3 x Noctua 140 iPPC's (one fan to each end of the splitter, no further daisy chaining). Considering the Noctua 140 iPPC's power draw, I think should be okay to power up 3 x Noctua 140 iPPC's, just would like a confirmation.

And considering the NA-FC1 draws power from the PSU, so power wise should be ok.

Setup would look like:
MoBo Fan Header + PSU (total 2 connections, one PWM signal input, one power source) to NA-FC1.
Then from NA-FC1 --> 3 way splitter and one Noctua 140 iPPC to each splitter end.

NF-A14 industrialPPC-3000 PWM:
Rotational speed (+/- 10%) : 3000 RPM
Max. input power : 6,6 W
Max. input current : 0,55 A
Operating voltage : 12 V

Quote from Noctua NA-FC1 webpage:

"Power Supply Adapter

As powering multiple higher wattage fans from a single motherboard fan header may damage the motherboard, the NA-FC1 includes an NA-AC4 adaptor that powers the device directly from the power supply unit while simultaneously receiving PWM input from the motherboard and forwarding RPM information. This makes it ideal for controlling multiple high-speed fans such as Noctua’s 3000rpm industrialPPC models."

Question 2:
What do you think of the airflow currents in my setup? Any recommendations / suggestions for improvement is most welcome. (See the reply to @Karadjgne for the air flow setup + case pictures / other related info)

Thank you for your time, I appreciate it a lot.
 

Paperdoc

Glorious
Ambassador
The NA-FC1 is useful if you do not have any mobo means of controlling fan speeds - that's its manual mode. In the mode where it uses the mobo's automatic control signal and simply modifies that, all it really can do is reduce the speed the mobo requests, by reducing the PWM duty Cycle. But that same thing usually can be done by configuring the mobo automatic system, without the extra device. One feature of this unit is the optional "no stop" control button, which will ensure that its output will never take the fan to below 300 RPM. BUT read the caution about that. This feature does NOT work for fans designed to never go that slow - that is, ones that can stall at a higher speed. For that reason, this feature will NOT work with the iPPC fans you have chosen. They both have minimum speeds of 750 to 800 RPM.

The Noctua website suggests this controller is useful for high-power fans like the iPPC's. That is true IF your plan is to use several connected to a single mobo header using a SPLITTER, with which all fan power must come from the header limited to 1.0 A. BUT your plan is NOT that - you plan to use a HUB which draws power from the PSU, and has a limit of 4.5A max total current. That is MORE that the 3.0 A limit of the Noctua device.

In fact, the NA-FC1 is rather an advanced Hub. It does get power from the PSU, and can accept a PWM signal form the mobo. Its has additional features in its abilities to operate without the mobo input, to reduce the PWM signal, and to impose a minimum speed for fans that are designed for a minimum at or lower than 300 RPM.

In these discussions you have not told us what mobo - maker and exact model number - you have, or plan. If you tell us that, we can give you clear details of what the MOBO's automatic fan controls can do. Many of the features of that Noctua Controller are already included on your mobo. Your mobo may even have a better "no stop" feature for your iPPC fans.
 

SteelMouse

Distinguished
Dec 21, 2013
49
2
18,545
1
Thank you for the detailed explanation of the NA_FC1's operation modes.

My motherboard is: Asus X99 E-WS : https://www.asus.com/Commercial-Servers-Workstations/X99E_WS/
Asus X99 E-WS Specs: https://www.asus.com/Commercial-Servers-Workstations/X99E_WS/specifications/
Manual of the Asus X99 E-WS: https://dlcdnets.asus.com/pub/ASUS/mb/Socket2011-R3/X99-E_WS/Manual/E13676_X99-E_WS_UM_V4_WEB.pdf

Fan Headers Info:
CPU_FAN
CPU_OPT
CHA_FAN1
CHA_FAN2
CHA_FAN3
CHA_FAN4

CPU Cooler: Noctua NH-D15S : (https://noctua.at/en/nh-d15s/specification)
Default supplied fan Noctua NF-A15 PWM is in the middle of the heatsink
and a Noctua 120 iPPC 3000 RPM in the front for PUSH.

Note:
1. Even though the CPU_FAN header can run both the NF-A15 PWM and Noctua 120 iPPC 3000 RPM in a single header, these two fans vary in max RPM. NF-A15 PWM's max RPM is 1500 while Noctua 120 iPPC's is 3000 RPM.
And setting a 50% PWM makes the NF-AF15 PWM run at 750 while the Noctua 120 iPPC runs at 1500 RPM when both fans share a single header.
2. Moving the Noctua 120 iPPC to CPU_OPT results in the same end result because in the BIOS, there's no curve setting option for CPU_OPT. It seems CPU_OPT follows the curve set for CPU_FAN. (For ex: Setting 50% in CPU_FAN results in CPU_OPT's 120 iPPC 3000 RPM fan running at 1500 RPM)
And this forces us into the situation of having to run the PUSH fan Noctua 120 iPPC into a different header / fan hub for more precise speed control. And doing so we can run the NF-AF15 PWM at max RPM of 1500.

Fans Setup (Existing):
I already have all the fans except a single 140 iPPC 3000 RPM + two NA_FC1's that I'd have to buy to switch to the planned setup + repurpose one Noctua 120 iPPC 3000 RPM (repurpose it in my another older desktop).
The existing setup is:
All fans are Noctua iPPC 3000 RPM's.
Intake: 2 x 120 in front + 2 x 120 in bottom + 1 x 140 in side.
Exhaust: 1 x 140 in rear + 3 x 120 in top.

Another aspect with the Noctua 140 iPPC's in relation with the motherboard fan header is: while the motherboard fan header can run two 120 iPPC 3000's off a single fan header via a splitter, the 140 iPPC's are rated 0.55 A (Max Input Current) and two of them in a single header would exceed the 1A limit of the header. The fans spin at max speed at startup, so the startup max speed scenario gives us a power load calulation that a single motherboard fan header can't handle. This forces us to use a dedicated fan header for a single 140.

Current fans to motherboard fan header connections:
CPU_FAN: Noctua NH-D15S : NF-A15 PWM
CPU_OPT: Noctua 120 iPPC 3000 RPM (Push fan for the NH-D15S)
CHA_FAN1 : 1 x 140 iPPC 3000 RPM (Side Intake 140)
CHA_FAN2: 2 x 120 iPPC 3000 RPM (Bottom Intake 2 x 120)
CHA_FAN3: 2 x 120 iPPC 3000 RPM (Front Intake 2 x 120)
CHA_FAN4: 1 x 140 iPPC 3000 RPM (Rear Exhaust 140)
Case Fan Hub : 3 x 120 iPPPC 3000 RPM (Top Exhaust 3 x 120)

The case is NZXT Phantom 530 and has a basic fan hub that's voltage control and has 3 pin output headers. And with this fan hub, the current problem is the top exhaust fan's are very loud even in the medium setting. And switching to the low setting doesn't work (doesn't slow the fans down) and at next startup with the low setting, the fans don't spin. This forces us to use the NZXT Phantom 530 fan hub at medium setting and the resulting very loud operating noise levels.

Planned Setup:
Objective: The current setup is almost neutral air pressure and dust accumulation is a big problem. Achieving a positive air pressure will make maintenance lot easier with lesser dust accumulation.
Current temps: Current temps are not a concern at all. Case is very well ventilated even with the existing loadout of the fans, just the top 3 exhausts are making unbearable noise due to lack of precise control by the 3 pin case fan hub.

The planned setup is:
CPU_FAN : Noctua NH-D15S : NF-A15 PWM: Fan curve set to 100% all the time in motherboard BIOS. 1500 Max RPM of this fan is lesser noise than the Noctua iPPC's and workable with respect to noise levels.
CPU_OPT: Empty
CHA_FAN1: ---> NA_FC1 (2 connections, first to motherboard fan header + power from PSU) ---> 3 way splitter supplied with NA_FC1 ---> Noctua 140 iPPC : one fan each on one end of the splitter : So 3 fans in total (all Noctua 140 iPPC 3000 RPM's)
CHA_FAN2: ---> NA_FC1 (2 connections, first to motherboard fan header + power from PSU) ---> 3 way splitter supplied with NA_FC1 ---> Bottom Intake 2 x 120 iPPC 3000 RPM + NH-D15S Push fan Noctua 120 iPPC 3000 RPM
CHA_FAN3: 2 x 120 Noctua iPPC 3000 RPM (Two top exhausts)
CHA_FAN4: 2 x 120 Noctua iPPC 3000 RPM (One top exhaust + One rear exhaust).

In effect, I'd be buying a Noctua 140 iPPC 3000 RPM, putting that in the front intake (taking out one Noctua 120 iPPC 3000 RPM and repurposing it). And swap the other front intake (Noctua 120 iPPC 3000 RPM) with the rear exhaust (Noctua 140 iPPC 3000 RPM). This would result in two x Noctua 140 iPPC 3000 RPM's for the front intake.

And overall end fan intake + exhaust setup would be:
Intake: 3 x Noctua 140 iPPC 3000 RPM's (two front intake + one side intake) + 2 x Noctua 120 iPPC 3000 RPM's
Exhaust: 4 x Noctua 120 iPPC 3000 RPM's (3 in top exhaust + one rear exhaust)
And finer control of the NH-D15's NF-A15 PWM (set to run at full 1500 RPM speed) while it's push fan will be paired with the bottom intakes.
And all the intakes (plus CPU Cooler push fan) will be running around 1800-2000 RPM's (maybe tweaked to 1700 if noise levels still persist, I don't mind the noise usually, just way too much becomes rough).
And all the exhausts will be running around 1200-1400 RPM.
And the NH-D15's NF-AF PWM will be running at max RPM of 1500 all the time.

Overall this would end guaranteeing a good positive air pressure and with daily cleaning of the intake air filters (front + side + bottom) would make maintenance much easier. Currently I have to clear out the dust every month. Maybe if this setup makes dust accumulation lesser, that would be worth it. And would have achieved the maximum possible cooling with air cooled setup that's precisely controllable. It's a combination of inability to run double Noctua 140 iPPC's off the same header + CPU_OPT not having a dedicated fan curve in BIOS + disparate CPU cooler fans RPM capability (1500 vs 3000 RPM) + top exhausts running at unbearable noise + dust accumulation + maintenance pain that necessitates the need to switch to planned setup. And the planned setup would solve the noise problem from the top 3 exhausts + more intake than exhaust (positive air pressure) + resulting lesser dust accumulation + resulting easier lesser maintenance + ability to run NH-D15S NF-A15 PWM at max RPM would make the planned setup worth it in the longer run.

Could have purchased all Noctua iPPC 2000 RPM's and none of these issues would be there, but the extra on-demand buffer RPM from 2000 - 3000 RPM is desirable and is why the need for a detailed fan control setup.
 
Last edited:

Paperdoc

Glorious
Ambassador
Six comments to offer.

1. I can see from this last post that you have no intention of allowing any fans to be automatically controlled according to workload as measured by temperature sensors. You prefer to have total control yourself, and plan for settings that, if anything, will over-cool and avoid any possibility of inadequate cooling.

2. It appears in all cases you have fixed speed settings for each fan header and group. You can accomplish that entirely (with the exception that CPU_OPT is always the same as CPU_FAN, but you don't plan to use that header) by using the Manual option for setting fan speed for each header. See mobo manual p. 3-7 and 3-8. NOTE in doing that, that you can avoid setting the fan speed below the fan's minimum, AND that a further minimum speed limit is not required since the speed is fixed.

3. In the two locations for using Noctua NA-FC1 units, their only function is as a HUB that can provide more power than a mobo header and Splitter can provide. You are not using the minimum speed limiter, and you have no need to reduce the speed signal from the mobo since you can set that at the mobo header directly. For that application you CAN use common fan Hubs, which may cost less.

4, When deciding on fan speeds and balance, I suggest you use the fan's max AIR FLOW rating, rather than the speed. That parameter is a better guide to what the fan really does for your system. Unfortunately, neither measure can tell you the reality when a fan is NOT at 100% speed, so you have to make a rough guess based on presumed linearity between max and min speed ratings.

5. Regarding dust, I recommend that ALL intake fans have dust filters on their intake sides. Although this does reduce their air flow from the stated ratings, the impact is slight. (See next comment for a hint.) Intake dust filters substantially reduce dust accumulation inside the case and make most dust removal a task of cleaning and re-installing the filters, rather than trying to clean then entire case interior.

6. After it's all assembled, set and running, you can get a good indication of actual air flow balance with a simple procedure, although not quantifiable. You need a source of smoke, like a smouldering cigarette or a lit incense stick. Move that smoke source close to any areas of the case with a small crack where air can flow, but not near a fan. Observe the direction and speed of the smoke movement. If it is away from the case, you have a positive pressure inside and air leakage flows inside-to-out. If it drifts away at a modest rate, the balance is about right. If it blows away rapidly, you might want to reduce the air intake fans' speeds.
 

SteelMouse

Distinguished
Dec 21, 2013
49
2
18,545
1
Thank you for the detailed comments.

1. I should have explained the intended fan control RPM's a bit better: I am planning to achieve the RPM's (1700-2000 is the normal temp curve, then above 50/55 Celcius, it will go up a bit). Idea is to have the intake curve higher at all temp points compared to exhaust curve. Ex: Intake curve starts at 50% PWM Duty Cycle @ 20 Celcius, then 60% PWM Duty Cycle @ 30 Celcius, 75% Duty Cycle @ 55 Celcius and the exhaust curve would start at 30% PWM Duty cycle @ 20 Celsius, and respectively lower than the intake curve. So pretty much achieving faster intake RPM's via higher intake PWM curve than the exahaust's PWM curve.

2. Regarding the dust filters, yeah, that has become my routine now, clean up dust filters before computer startup. And all intakes have the dust filters set up.

3. Nice idea on testing the positive air pressure, will try it once the setup is complete.

Question 1(repeated from last post):

What do you think of the airflow currents in my setup? Any recommendations / suggestions for improvement is most welcome. (See the reply to @Karadjgne for the air flow setup + case pictures / other related info)
So far from my understanding, there air should be getting in and out smoothly without any air trapped in circulatory pattern.

Question 2:

On the Nocta NA-FC1, regarding the supplied 3 way splitter:

NA-FC1 : Specs:
Max Current: 3A
(Planning to power it via PSU + get the PWM signal from motherboard 4-pin fan header)
(i.e. the Setup A as stated in the installation manual of NA-FC1)
(NA-FC1 installation manual: https://noctua.at/pub/media/blfa_files/manual/noctua_na_fc1_manual_en.pdf)

NF-F12 industrialPPC-3000 PWM:
Rotational speed (+/- 10%) : 3000 RPM
Max. input power : 3,6 W
Max. input current : 0,3 A
Operating voltage : 12 V

NF-A14 industrialPPC-3000 PWM:
Rotational speed (+/- 10%) : 3000 RPM
Max. input power : 6,6 W
Max. input current : 0,55 A
Operating voltage : 12 V

Fans setup:
1. MoBo Fan Header (CHA_FAN1) + PSU (total 2 connections, one PWM signal input, one power source) to NA-FC1.
Then from NA-FC1 --> 3 way splitter and one Noctua 140 iPPC 3000 RPM to each splitter end.

Quote from Noctua NA-FC1 webpage:

"Power Supply Adapter

As powering multiple higher wattage fans from a single motherboard fan header may damage the motherboard, the NA-FC1 includes an NA-AC4 adaptor that powers the device directly from the power supply unit while simultaneously receiving PWM input from the motherboard and forwarding RPM information. This makes it ideal for controlling multiple high-speed fans such as Noctua’s 3000rpm industrialPPC models."

Can the above fan setups work with the NA-FC1 without overloading the electrical loads / causing wires to melt / fire? Especially the 3 x Noctua 140 iPPC 3000 RPM's connected via 3 way splitter (one fan on each splitter end) to the NA-FC1 which in turn is powered by a dedicated cable to the PSU + gets the PWM signal from the motherboard. The dedicated cable to PSU will not be shared with any HDD/SDD/Other SATA devices.
Usually running 2 x Noctua 140 iPPC 3000 RPM's via a dual splitter (Y Splitter) off a single motherboard fan header will exceed the total 1A limit of the motherboard with each 140 iPPC 3000 RPM having 0.55 A Max Input Current. And with the fans spinning at max speed at startup, 2 x 0.55A will exceed the 1A limit of the motherboard fan header.
Considering this, running 3 x 140 iPPC 3000 RPM's would give the power load of:
0.55 A Max Input Current x 3 = 1.65 A for three 140 iPPC 3000 RPM's against the NA-FC1's Max Current of 3A
In theory, seems like should work, only thing being: Can the 3 way splitter cable included with the NA-FC1 handle 3 x 140 iPPC 3000 RPM's connected to the 3 way splitter (one fan on each splitter end)?
And can the wires in the 3 way splitter handle the electrical load of 3 x 140 iPPC 3000 RPM's without melting / other problems?

Thank you for your detailed answers, I appreciate it a lot.
 

Paperdoc

Glorious
Ambassador
If your plan really is to make fan control automatic based on actual measured temperatures, your mobo can do all of that. See its manual, p. 3-7 and 3-8. When you select Standard Profile at the bottom, it does that using a pre-set fan "curve". But if you choose "Manual" at the far right, then you get to set your own fan curve (p. 38). It appears to allow three points. I don;t know if it allows more. Use the furthest-left point to define the minimum speed for the lowest temperature range. Of course, you still need a couple of Hubs (or the Noctuas) to deal with the power issues, but the CONTROL all can be done with the Manual fan curve option.

A fan HUB (including that Noctua) deos NOT draw any power from the mobo header and hence does not present any load problem for th header. Then the limit on fan load is the capacity of the HUB to provide power, and in this case that turns out to be 3.0 A max. You are tight, your proposed aramgement is no problem. I am sure the Splitter provided with the Noctua unit has wires sufficient for that load. A Hub does get the PWM signal from the mobo header and merely share it out to its fans (normally, but Noctua is different), but that is NOT an overload issue for the mobo header.
 

SteelMouse

Distinguished
Dec 21, 2013
49
2
18,545
1
Thank you for the confirmation on the Noctua Splitter wires. I can get by with a basic fan hub, but quality / lack of documentation / compatibility issues and the Noctua NA-FC1 is slightly more expensive, but worth it. And the Noctua NA-FC1's documentation impressed me on the level of detail, as custom with Noctua products.
And the system is running all Noctua stuff anyway: Noctua fans, Noctua CPU Cooler, Noctua fan extension cables, Noctua Y Splitter cables, might as well add one more quality Noctua in there, makes it complete.

Question 1:

If I my understanding is right, the fan hub / controller (NA-FC1) sends the PWM signal, provides power to the connected fans, reports the RPM back to the motherboard.
Considering this, in theory, I should be able to mix and match 120 iPPC 3000's / 140 iPPC 3000's in a single NA-FC1 without any problem (provided electrical load limits are not exceeded).
Ex: MoBo Fan Header (CHA_FAN1) + PSU (total 2 connections, one PWM signal input, one power source) to NA-FC1.
Then from NA-FC1 --> 3 way splitter and 2 x Noctua 140 iPPC 3000 RPM + 1 x Noctua 120 iPPC 3000 RPM.
And the 3 fans (two 140's + one 120) should share the same PWM signal and run at the same RPM / PWM Duty Cycle.

And in theory, should be able to mix and max Noctua 3000 RPM / Other brand PWM fans with lesser than 3000 Max RPM (provided it stays within electrical limit), the PWM fans should run at the % PWM Duty Cycle that's set. Ex: 50% of 3000 RPM = 1500 RPM & 50% of 2000 RPM = 1000 RPM.
And the fan hub / controller will allow each fan to draw however much power it wants to draw as the 3000 RPM's might need more vs 2000 Max RPM fans.
And I assume there isn't a requirement in the fan hub / controller (NA-FC1) that all the connected fans should be of the same make / model and should draw the same amount of power from the hub.

Question 2:
(It's answered, just would like your perspective on this)
https://forums.tomshardware.com/threads/fan-motor-and-psu-wire-proximity-electromagnetic-interference.3681199/#post-22169824

Thank for your time, I appreciate it a lot.
 
Last edited:

Paperdoc

Glorious
Ambassador
#1 You are entirely correct in all of this.

#2 I fully agree, the probability of trouble is near nil. Most of the electrical flow in the PSU output wires is DC, although that can have fluctuations as components' power draws change. It is the CHANGE of current flow that can induce noise signals in adjacent wires. But a fan is rarely affected by such small noise in its power supply.

From the other direction, can current flows inside the fan motor affect the DC power flow and induce noise in the PSU's output wires and feed that to mobo, graphics card, etc? Well to start, the current for a fan is small, so any noise it could generate is small. Beyond that, virtually all power consumer devices in your system have filters on their power input lines to ensure they do NOT "inherit" moise in that fashion. Noise pickup that affects computer components almost always is though low-voltage analog or digital signnal lines with poor shielding, and NOT through power supply lines.
 

ASK THE COMMUNITY

TRENDING THREADS