Demanding case/rad fans (wattage)

[VoidParticle]

The Noctua NF-A14 industrialPPC 3000rpm fans at maximum rpm use 6,6w. Not only can this be demanding for a single fan header if the motherboard uses 6w headers, but even at 12w headers you would never split 2 of these fans to the header.

My question is how do I successfully find a fan splitter device like this ¹ or this ² while knowing how many watts the total device or each header/cable can handle? Nobody advertises how many watts the fans either add up to or can sustain per header. Basically I'm saying my motherboard tells me how many watts per fan header but these devices do not which is a pain.

I do plan on allowing these fans to go full rpm if they need to based on heat. Yes these fans are overkill but I want all the headroom I can for heat dissipation off my rads.
 

[W-L]

4 minutes ago, VoidParticle said:

The Noctua NF-A14 industrialPPC 3000rpm fans at maximum rpm use 6,6w. Not only can this be demanding for a single fan header if the motherboard uses 6w headers, but even at 12w headers you would never split 2 of these fans to the header.

My question is how do I successfully find a fan splitter device like this ¹ or this ² while knowing how many watts the total device or each header/cable can handle? Nobody advertises how many watts the fans either add up to or can sustain per header. Basically I'm saying my motherboard tells me how many watts per fan header but these devices do not which is a pain.

I do plan on allowing these fans to go full rpm if they need to based on heat. Yes these fans are overkill but I want all the headroom I can for heat dissipation off my rads.
 

You can use power fan splitters such as those and use PWM control to vary the speed of the fans but note that having extreme high pressure and high speed fans can only do so much for cooling. In most cases especially for liquid cooling your limiting factor is the rate at which the heat can transfer into the waterblock and fluid. 

[VoidParticle]

1 minute ago, W-L said:

You can use power fan splitters such as those and use PWM control to vary the speed of the fans but note that having extreme high pressure and high speed fans can only do so much for cooling. In most cases especially for liquid cooling your limiting factor is the rate at which the heat can transfer into the waterblock and fluid. 

yeah that's fine. For me 3000rpm fans are there so that 2000rpm isn't my ceiling, but if they were to go to 3000rpm from a fan curve I'd like to know that 6.6w x # of fans on device will be supported.

[W-L]

2 minutes ago, VoidParticle said:

yeah that's fine. For me 3000rpm fans are there so that 2000rpm isn't my ceiling, but if they were to go to 3000rpm from a fan curve I'd like to know that 6.6w x # of fans on device will be supported.

As long as your alright with the noise, since those fans are not silent even dialed back. For the splitters as long as your have it plugged into a PWM capable header and aux (PSU) power it will work as a group. 

[Majestic]

You do realise that those fans kicking at 3k rpm is only really designed to be used in serverrooms? There is no way you're ever going to run those at 3k rpm. Or even 2k rpm on a consumer enviroment. Max rpm on my NF-F12's is already very loud (1200 or 1500, not sure).

[VoidParticle]

8 hours ago, Majestic said:

You do realise that those fans kicking at 3k rpm is only really designed to be used in serverrooms? There is no way you're ever going to run those at 3k rpm. Or even 2k rpm on a consumer enviroment. Max rpm on my NF-F12's is already very loud (1200 or 1500, not sure).

what are you talking about? Fans rated for 3k rpm can hit 3k rpm with their maximum voltage being delivered to them. They wouldn't need to be at 3k rpm all day like a server but if I set up a PWM fan curve that allows them to hit 3k rpm as the cpu or something reaches 85 degrees, they'll hit that rpm +-10%. I've also seen overclocking tests where they are using 50mm thick rads and you see a huge benefit running fans at 2.2k rpm. Also each Noctua fan is rated for it's size and max rpm along with the airflow and static pressure. I'm pretty sure 3k rpm fans are 3k rpm fans.

[VoidParticle]

8 hours ago, W-L said:

As long as your alright with the noise, since those fans are not silent even dialed back. For the splitters as long as your have it plugged into a PWM capable header and aux (PSU) power it will work as a group. 

Yeah I'm not one of those ultraquiet 400rpm guys. If things get unbearably loud I'll dial my overclock or fan curve back a bit. Thanks for answering though. I just wanted to make sure the fans were gonna work on one of these connectors/cables. My computer is also going to be 6 feet away from me and not at my feet. Seems to knock the decimal tests back a bit.

[Majestic]

7 hours ago, VoidParticle said:

what are you talking about? Fans rated for 3k rpm can hit 3k rpm with their maximum voltage being delivered to them. They wouldn't need to be at 3k rpm all day like a server but if I set up a PWM fan curve that allows them to hit 3k rpm as the cpu or something reaches 85 degrees, they'll hit that rpm +-10%. I've also seen overclocking tests where they are using 50mm thick rads and you see a huge benefit running fans at 2.2k rpm. Also each Noctua fan is rated for it's size and max rpm along with the airflow and static pressure. I'm pretty sure 3k rpm fans are 3k rpm fans.

I'm talking about the acoustics being unreasonably loud mate. There is no way in a non-server enviroment you would be sitting next to one of those fans running 3K. They go over 60dB(A). And they're controlled by PWM, not voltage. So they run 3K at maximum duty cycle, not voltage.

[Mick Naughty]

Prefer to use a him. That way I can actually lower the fan speed below 40% for 2k fans and 45% for 3k fans. 

But if you don't care about noise and for some reason need 3k fan at max all day for a server room then it wont really matter

 

[Tiberiusisgame]

I actually run 5 of these fans in my case. 2 140s and 3 120s. I love them and sometimes I do run them at 3,000 RPM. I agree there's a diminishing return on the heat dissipation and it sounds like you're on the deck of an aircraft carrier, but it's a satisfying feeling to have your entire desk vibrating from fan whirrr.

 

I can run all 5 of these fans at full tilt plus 2 80mm fans and 10 LED strips on a Corsair Commander Mini. Like @Mick Naughty said, it's better at fan control than the motherboard.

 

But as was suggested above, if you do fan control from the board, just use the PWM signal and run the rest directly from the power supply.

I did burn out a motherboard header after I drew roughly 2 Amps from the fan connector. It was a test on a 1st generation Intel board. No smoke but the LEDs on the board now misbehave and obviously that fan header is dead.

 

I also tried to run these fans at 15 volts (variable voltage power supply) as a test and am happy to note that they can handle the extra power. They start to oscillate and become unstable but they work. I have a feeling the motors in these fans are the same as Noctua's 24volt models, just didn't prove it.

[LogicalDrm]

Its not about watts. Its about amps. Single normal mobo header is 1A. IPPC fans are 0.3A per. Meaning that you can run 3 of single mobo header. If you are using 120mm ones. 140mm is 0.55A. So to be safe, one per header. Some mobos have high amp headers, 2-3A per. Meaning more fans per header. Or you can get hub using SATA power. At that point it depends on how much your PSU can deliver per SATA.

[VoidParticle]

On 12/20/2017 at 7:09 AM, Tiberiusisgame said:

I actually run 5 of these fans in my case. 2 140s and 3 120s. I love them and sometimes I do run them at 3,000 RPM. I agree there's a diminishing return on the heat dissipation and it sounds like you're on the deck of an aircraft carrier, but it's a satisfying feeling to have your entire desk vibrating from fan whirrr.

 

I can run all 5 of these fans at full tilt plus 2 80mm fans and 10 LED strips on a Corsair Commander Mini. Like @Mick Naughty said, it's better at fan control than the motherboard.

 

But as was suggested above, if you do fan control from the board, just use the PWM signal and run the rest directly from the power supply.

I did burn out a motherboard header after I drew roughly 2 Amps from the fan connector. It was a test on a 1st generation Intel board. No smoke but the LEDs on the board now misbehave and obviously that fan header is dead.

 

I also tried to run these fans at 15 volts (variable voltage power supply) as a test and am happy to note that they can handle the extra power. They start to oscillate and become unstable but they work. I have a feeling the motors in these fans are the same as Noctua's 24volt models, just didn't prove it.

If I were to get the commander mini or pro is there any way of finding out how many amps run per 4 pin fan header? I can't find a spec sheet and I'd like to not need to run 2 of these units for the fans in my case. If the 3000rpm fans use 0.55amps and I'm using a push/pull configuration for 2 rads then I'd be running 10 fans just for rads. I don't know the limit of how many fans can be attached to one fan header on a Corsair Commander mini or pro.

[Mick Naughty]

You really plan on running 10 3k fans at max?

[VoidParticle]

4 minutes ago, Mick Naughty said:

You really plan on running 10 3k fans at max?

No but I feel if I'm leaving anything automated they should be able to hit 3k. Which means I don't wanna kill a Corsair Pro or my fans just because I didn't know what the limits were. I could just buy 2 of these units and hook up 1 fan each header, or buy 1 and hook up 2 or 3 fans per header if I know what I'm doing.

[Mick Naughty]

So why not just regular 1200/1500 rpm noctua's and not worry about it? 

44 minutes ago, VoidParticle said:

No but I feel if I'm leaving anything automated they should be able to hit 3k. Which means I don't wanna kill a Corsair Pro or my fans just because I didn't know what the limits were. I could just buy 2 of these units and hook up 1 fan each header, or buy 1 and hook up 2 or 3 fans per header if I know what I'm doing.

 

[VoidParticle]

5 minutes ago, Mick Naughty said:

So why not just regular 1200/1500 rpm noctua's and not worry about it? 

 

Because 2 gtx rads can benefit from 2.2k+ rpm and I'm not worried about noise that much. Also I'll be using rads that use 140mm fans (which run higher amps/watts than 120mm) so that I cover more surface area while also gaining the benefit of the thickness of the rads. If I only use a push fan setup I would like 2.2k rpm. If I run push/pull fan setup I'd want 2k rpm with some headroom. My computer is also not going to be at my direct side and would be about 5-6 feet away from me and behind a bit of my desk so that 45 dB fans would sound quieter than if the computer was at my side. I'm also overclocking and experimenting with getting that delta T to ambient air temperature down as much as possible (for summer heat and my benchmark boner).

 

Also after researching pwm I'm more inclined to find an automated way to control fans by temp using only voltage. I didn't realize pwm cycles 12v in intervals trying to bring fans up to full speed and then cutting the power. I'd rather have the fans at a steady voltage without their "whirring up" every 1/2 second. I'm pretty sure this would make the 2k and 3k rpm fans sound the same at 2k rpm compared to pwm's way of making both fans 2k rpm.

[Mick Naughty]

So there isn't enough rad space I assume to warrant this. 

Also need to be concerned about powering the hum or what ever, the max power the psu is gonna give you at that point.

 

Might as well go all out and get noctuas 24v fans.

 

[VoidParticle]

4 minutes ago, Mick Naughty said:

So there isn't enough rad space I assume to warrant this. 

Also need to be concerned about powering the hum or what ever, the max power the psu is gonna give you at that point.

 

Might as well go all out and get noctuas 24v fans.

 

I'd still be at my initial question of this post asking what do I use for this? Now I'm certain I can't use my motherboard's fan headers, and these devices like Corsair Commander Pro don't advertise their max amps or you can only assume they are about the same as a motherboard fan header. I don't have any knowledge for a DIY approach nor would I find it safe without enough research. I'm also certain the Corsair Commander Pro can only supply 12v. Square 1 would be to find a fan controller that can supply 24v?

[Mick Naughty]

1 minute ago, VoidParticle said:

I'd still be at my initial question of this post asking what do I use for this? Now I'm certain I can't use my motherboard's fan headers, and these devices like Corsair Commander Pro don't advertise their max amps or you can only assume they are about the same as a motherboard fan header. I don't have any knowledge for a DIY approach nor would I find it safe without enough research. I'm also certain the Corsair Commander Pro can only supply 12v. Square 1 would be to find a fan controller that can supply 24v?

Controller wont convert 24v but another psu would. 

 

Could email them and ask or sacrifice rpm and just run 2k fan. If you don't have enough rads, maybe expanding the loop would be more beneficial.

[LogicalDrm]

5 hours ago, VoidParticle said:

No but I feel if I'm leaving anything automated they should be able to hit 3k. Which means I don't wanna kill a Corsair Pro or my fans just because I didn't know what the limits were. I could just buy 2 of these units and hook up 1 fan each header, or buy 1 and hook up 2 or 3 fans per header if I know what I'm doing.

SATA power connector can deliver 13A. Divide that between headers and so on. But I think its about 3A per header.