School me on radiators and fans - first real custom water loop.

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
I ordered 3 PWM fan hubs that use Molex for the power. At less than $4 a piece from China how could I not?

That way I can have up to 11A per hub instead of the SATA powered ones that max out at 4.5A per hub.

These would give me the option of running the high power Delta fans if I actually decided to try them out for giggles.
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
ummm... that would be why not. thats awfully cheap...

There really isn't much to a fan hub. It just connects the pwm signal together for all the fans and then to the motherboard fan header.

And then it pulls the power for the fans from the PSU.

Also has a single filtering capacitor.

No case of course like the name brand ones.. but those are all SATA powered anyway so that doesn't help me.

This same one comes in a SATA powered model as well.

s-l1600.jpg
 
Joined
Feb 6, 2013
Messages
724
Late to this party but wanted to say that as for fans get the best HSP PWM fans you can get at the best cost. Radiators will perform better or worst depending on what fan speed they are designed for. PWM takes care of that. Since you have a closed case just get Arctic fans, great value no compromise. I will always push HWLabs gts rads until someone can beat that value to performance proposition.
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
I sent bGears and email asking for more detailed info on the b-Blaster 140x38 fan. They said operating voltage is 9-15v with a startup voltage of 7v.

They said the dB(a) is 40-70 depending on the speed... which seems kinda silly to me. Minimum rpm is supposedly about 2600 which is also silly if that is true.

We shall see what it actually does when it arrives and I have a chance to test it out. I'll probably end up hooking it up to an adjustable bench test power supply since my PWM hubs will not be here for a while.
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
So.. as far as pumps go... D5 is higher flow but DDC is higher pressure..

Does it really matter what the brand is for either of those designs?

I do want a PWM controlled one.
 

VirtualMirage

Limp Gawd
Joined
Nov 29, 2011
Messages
467
So.. as far as pumps go... D5 is higher flow but DDC is higher pressure..

Does it really matter what the brand is for either of those designs?

I do want a PWM controlled one.
The D5 also relies on the coolant to keep it cool while a DDC expels heat into the case via a heatsink and requires some kind of active cooling (e.g., air movement). A D5 is usually quieter than a DDC too. D5s usually take up more room than a DDC. I believe the D5 has a little more flexibility when it comes to mounting options, especially if you plan to hang it from fans or a radiator that are at the top of a case. Each have their own mounting advantages and disadvantages.

From what I have read, unless you are tight on space and/or are running a lot of restriction over a short run, a D5 should serve most people perfectly fine in most modern case setups.

Even in my SFF build, I opted for the D5 over the DDC since I could fit it (barely).
 

cdabc123

2[H]4U
Joined
Jun 21, 2016
Messages
3,162
I would advise going with some reasonable fans and no high amp delta ones. They work phenomenally but it will sound like a server is running full tilt next to you. My watercooled rig runs 1 120mm delta fan (~5amp) through a heatercore from a car. It gets the job done when it ramps up but makes a whole bunch of noise in the process. On idle I do not run the delta fan and just let the rising heat cool the computer
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
I really don't see why the super high rpm fans would have to be noisy or run at a lame high minimum speed. Back in the day, Dell used super high speed fans on some of their towers and when the system was running normally the fans ran at some low rpm where they were silent.

The speed control on them was also very precise / quick throughout the full rpm range.

In cases where the CPU was overheating for whatever reason they would ramp up to full speed and you could hear them from the other side of the house... Louder than any server fan I have ever heard running full tilt.

I actually have a couple of those fans around but they are only 92mm if I remember correctly and I am pretty sure they use 5 wires.
 

pendragon1

Fully [H]
Joined
Oct 7, 2000
Messages
27,201
I really don't see why the super high rpm fans would have to be noisy or run at a lame high minimum speed. Back in the day, Dell used super high speed fans on some of their towers and when the system was running normally the fans ran at some low rpm where they were silent.

The speed control on them was also very precise / quick throughout the full rpm range.

In cases where the CPU was overheating for whatever reason they would ramp up to full speed and you could hear them from the other side of the house... Louder than any server fan I have ever heard running full tilt.

I actually have a couple of those fans around but they are only 92mm if I remember correctly and I am pretty sure they use 5 wires.
i have several of those dell fans. they still have a start point and they are not nearly as fast or push as much air as what you are comparing to. dell used a custom 4pin pwm plug/header.
 

cdabc123

2[H]4U
Joined
Jun 21, 2016
Messages
3,162
I really don't see why the super high rpm fans would have to be noisy or run at a lame high minimum speed. Back in the day, Dell used super high speed fans on some of their towers and when the system was running normally the fans ran at some low rpm where they were silent.

The speed control on them was also very precise / quick throughout the full rpm range.

In cases where the CPU was overheating for whatever reason they would ramp up to full speed and you could hear them from the other side of the house... Louder than any server fan I have ever heard running full tilt.

I actually have a couple of those fans around but they are only 92mm if I remember correctly and I am pretty sure they use 5 wires.

They are often designed for a different use case. The motors make more noise and the fan blades are designed to flow air at a certain rpm. If you have no need for a excessive amount of airflow or pressure then find a fan designed to run in the range you need it too.

Dell desktop fans did cover a fairly wide range however consumer fans are by far the best for noise/prefermence. A radiator does not need any stupid amount of airflow to work well
 

VirtualMirage

Limp Gawd
Joined
Nov 29, 2011
Messages
467
I have a D5 pump/reservoir on the way now.
EKWB EK-XRES 140 REVO D5 pwm
Did you order a used or new old stock?

I thought EKWB discontinued those and replaced them with the Quantum Kinetic line. I don’t think there is much difference other than the newer one has improved port layouts at the pump (3 ports vs 2 ports, too, IIRC) and a reservoir cap with 3 ports versus 1 port. I can’t recall if the older model also has the anti-vortex swap out tube that I find is more useful when the pump is horizontal.
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
I ordered used for over $100 less than a new one goes for. Supposedly it was "lightly used" so if it seems like the pump is making noise upon testing I will be asking for a refund.
 

thesecond

[H]ard|Gawd
Joined
Jan 15, 2010
Messages
1,074
How's this for an idea... 38mm for push and 25mm for pull.

Reason I am thinking this is because most of the 38mm have the secondary stationary blade thing that is supposed to help with static pressure.. and it is also giving the blade more space between it and the radiator which is a good thing.

I don't see that being helpful at all when used as puller fans.

In fact, it would be helpful if the puller fans were moved away from the radiator a bit as well. Will reduce the dead spot where the motor is and reduce turbulence and thus overall noise.

I might also experiment with doing the same thing on the push side since I will have a bit of room inside the case to mount the fans lower.
Its a cool idea, but it wont work like you think it will. Turbine effect will mean the stronger fan will do all the work, and the weaker fan will do nothing. You have to match your fans to see performance gains.

Just mount your fans in push or push/pull, and then run sufficient exhaust to get air out of the case.
 

cdabc123

2[H]4U
Joined
Jun 21, 2016
Messages
3,162
Its a cool idea, but it wont work like you think it will. Turbine effect will mean the stronger fan will do all the work, and the weaker fan will do nothing. You have to match your fans to see performance gains.

Just mount your fans in push or push/pull, and then run sufficient exhaust to get air out of the case.

I disagree. Air speed through the radiator will be reduced significantly so just about any fan on the pull side will help. I also agree with a small gap if possible on the pull side (1/4 inch or so)

I also believe the second set of stationary blade would help static pressure and flow through the radiator however the types of fans that have that are normally excessively loud do not throttle down well and are generally made for high airflow with moderate pressure. Ie. They will be loud.
 

thesecond

[H]ard|Gawd
Joined
Jan 15, 2010
Messages
1,074
I disagree. Air speed through the radiator will be reduced significantly so just about any fan on the pull side will help. I also agree with a small gap if possible on the pull side (1/4 inch or so)

I also believe the second set of stationary blade would help static pressure and flow through the radiator however the types of fans that have that are normally excessively loud do not throttle down well and are generally made for high airflow with moderate pressure. Ie. They will be loud.
Lots of people have tested this before you. You should do more reading before you commit.

You can start here. https://hardforum.com/threads/diffe... is being forced,burning out the motor sooner.
 

pendragon1

Fully [H]
Joined
Oct 7, 2000
Messages
27,201
and ive never heard this "gap" thing either. you dont want any gap between the fans and rad, so much so that you can buy gaskets for it...
 

vegeta535

Supreme [H]ardness
Joined
Jul 19, 2013
Messages
6,302
I ordered used for over $100 less than a new one goes for. Supposedly it was "lightly used" so if it seems like the pump is making noise upon testing I will be asking for a refund.
They are loud at full speed. I keep mine at 40% and it is silent.
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
and ive never heard this "gap" thing either. you dont want any gap between the fans and rad, so much so that you can buy gaskets for it...
Not a gap really... I was talking about having a shroud that extends the fans away from the radiator but the setup would be sealed against air leaking. That way you get better airflow from the spots that are not directly below the fan blades and it should reduce noise a bit as well.

It is really easy to prototype these types of setups with just some cardboard and tape.
 
Last edited:

pendragon1

Fully [H]
Joined
Oct 7, 2000
Messages
27,201
Not a gap really... I was talking about having a shroud that extends the fans away from the radiator but the setup would be sealed against air leaking. That way you get better airflow from the spots that are not directly below the fan blades and it should reduce noise a bit as well.
ah, ok, that probably would work.
 

VirtualMirage

Limp Gawd
Joined
Nov 29, 2011
Messages
467
They are loud at full speed. I keep mine at 40% and it is silent.
Mine runs between 2,430 and 2,500 RPMs and is silent as well. It's like the sweet spot, go slower and it gets noisier, go higher and gets noisier. Running it faster doesn't make much of a difference for temps either, not enough to warrant the noise and increased power consumption at least. I do have it set if the temps get too out of control that as a fail safe it ramps up the pump speed (something I have not experienced). But otherwise, it remains pretty constant within 100 RPMs.
 

cdabc123

2[H]4U
Joined
Jun 21, 2016
Messages
3,162
Lots of people have tested this before you. You should do more reading before you commit.

You can start here. https://hardforum.com/threads/diffe... is being forced,burning out the motor sooner.

That link does nothing but point out that with a stupid extreme you may have problems. The point where you would begin to wear them faster is the point beyond the designed rpm range of the fan anything within the range and it wont do a thing to harm them. If you would like to disprove my claim please give me an experiment that gives the cfm coming out of the radiator and show me an example of a fan that cannot accelerate air at that cfm. Anything short is just an opinion on what people think will happen. (Links to a forum post going to a forum post is not evidence of anything its just abunch of unbacked opinions)

If your cfm was on the high end it would be best to go with a high airflow fan on the pull side, however I highly disbelieve the cfm is ever high enough to justify that.

As far as gap goes a shroud will be needed.
 
Last edited:

vegeta535

Supreme [H]ardness
Joined
Jul 19, 2013
Messages
6,302
Mine runs between 2,430 and 2,500 RPMs and is silent as well. It's like the sweet spot, go slower and it gets noisier, go higher and gets noisier. Running it faster doesn't make much of a difference for temps either, not enough to warrant the noise and increased power consumption at least. I do have it set if the temps get too out of control that as a fail safe it ramps up the pump speed (something I have not experienced). But otherwise, it remains pretty constant within 100 RPMs.
Mine stays around 2800rpm. Actually surprised it has enough power to push a 360, 2x 280. gpu ands cpu block. There is a 2c difference running at 40% vs 100%. That is not worth it when everything runs cool already.
 

Tsumi

[H]F Junkie
Joined
Mar 18, 2010
Messages
13,465

If you followed the xtreme systems link, you will find the following testing:

UK3-l = 97.4CFM
UK3-h = 109.5CFM

UK1 tests
UK3-l + UK1-h = 97.1CFM (-.3%)
UK1-l + UK3-h = 110.3CFM (.7%)

Terminology:
UK3 = UltraKaze at 3000 RPM
UK1 = UltraKaze at 1000 RPM
-l = pull configuration
-h = push configuration

You can see that even with one fan at 3000 RPM and the other at 1000 RPM, with the slower fan being in pull, it still increased airflow across the radiator. Interestingly, but not surprisingly, putting the slower fan in push actually decreased performance from just the high speed fan in pull.

Even with a 2000 RPM difference, putting a fan on the pull side will increase performance. How much it increased it by was very small, but was an increase nonetheless.

and ive never heard this "gap" thing either. you dont want any gap between the fans and rad, so much so that you can buy gaskets for it...

A gap created by a shroud improves performance by allowing airflow to even out. This means that instead of some air trying to move through at high speed and others at lower speed, it moves through more uniformly and has a higher average flow. It also improves heat transfer where there would normally be a dead spot at the motor hub.
 

pendragon1

Fully [H]
Joined
Oct 7, 2000
Messages
27,201
A gap created by a shroud improves performance by allowing airflow to even out. This means that instead of some air trying to move through at high speed and others at lower speed, it moves through more uniformly and has a higher average flow. It also improves heat transfer where there would normally be a dead spot at the motor hub.
there was no shroud mentioned and it was already explained by who i quoted.
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
Ok, so I am planning on going with 10/16 soft tubing.

Looking at the case, it looks like I might need a couple 90 degree fittings... but I'll have to verify the pump/reservoir will fit where I want to mount it once it arrives.

The radiator and the waterblock arrived today. Did some seeing how the radiator would fit and it looks like I will most likely be mounting it almost all the way towards the back of the case with the ports at the back of the case. This will give me a super short run from the waterblock to the radiator and not block any drive bays.

Did some looking at fittings and trying to figure out what ones I want. May end up buying used and just replacing the o-rings which I already have plenty of.

As far as angled fittings that swivel, what is the best brand to get? Or should I just look at reviews?
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
Aggghhh... Nobody has the EK Duraclear 10/16 tubing in stock for a good price. EK wants $18 for 3 meters which is plenty but they are out of stock. Everywhere else that has it in stock wants about double that.

I can get black tubing but what fun is that?

Edit.. Ended up ordering a combination of EK and XSPC fittings from Performance PCs and they also had clear tubing in stock.

I think I have everything ordered except fans now.
 
Last edited:

VirtualMirage

Limp Gawd
Joined
Nov 29, 2011
Messages
467
Aggghhh... Nobody has the EK Duraclear 10/16 tubing in stock for a good price. EK wants $18 for 3 meters which is plenty but they are out of stock. Everywhere else that has it in stock wants about double that.

I can get black tubing but what fun is that?

Edit.. Ended up ordering a combination of EK and XSPC fittings from Performance PCs and they also had clear tubing in stock.

I think I have everything ordered except fans now.
I'd recommend some PrimoChill PrimoFlex Advanced LRT tubing. They sell it in clear and Amazon has it in stock.

I use the same brand of tubing for mine, but I have white.
 
Joined
Apr 16, 2018
Messages
964
I agree primochill lrt is top notch stuff but if your looking for eks black rubberish looking stuff they dont make anything like it that i know of.
For 90 degree rotaries i usually go with barrow, ek, bitspower, alphacool or byski. I dont use xspc rotaries because of some leaking issues ive had with them. With rotaries look out for any kind of sloppy or loose turning ones. Those are the ones that will leak. Only use the ones that turn hard. They are just pressed together so if they get loose it generally means the metal is coming apart.

You can but primochill lrt by the foot at performance-pcs.com
 

cyclone3d

[H]F Junkie
Joined
Aug 16, 2004
Messages
14,269
Here is what I ordered from performance-pcs.com
8x XSPC straight
5x EK-Torque 90 degree rotaries
9 foot XSPC clear tubing

I'm hoping this is all I will need. Was going to order directly from EK but they didn't have the clear tubing in stock and they wanted a bit more than $20 for shipping.
 
Top