Full speed vs temperature controlled fan placement

[Tripredacus]

I just did an upgrade for my gaming computer, and while I thought I had everything figured out ahead of time, one thing I didn't think of was fan headers. My chassis being the Chenbro SR105 has a front fan. The old board had a "front fan" header in addition to case and CPU fan headers. The new board only has a CPU and case fan header. In order to keep the airflow channel intact, I used a molex adapter to connect to the front fan.

The problem is that the rear fan is temperature controlled, and is running about 2700 rpm at idle. The front fan is running at (what I imagine) is full speed, or whatever speed is dictated by the molex connector.

Would it be better if the rear fan ran at this "full speed" and have the front fan be controlled by the board? My only concern there is that the board is expecting the that header to be connected as an exhaust fan rather than an intake and thus the temperature detection would be incorrect.

[Mcinwwl]

I'm loosing details here.

So you have 3 fans: CPU, rear and front. Right? Or Wrong?

I've seen pic in another topic, CPU one exhausts to the opposite direction to the mobo. obvious.

Front one ( I assume) blows the air in? at full speed.

Rear is mobo controlled and is an intake one. And that I do not grasp.

1 hour ago, Tripredacus said:

My only concern there is that the board is expecting the that header to be connected as an exhaust fan rather than an intake

Why do not do it "the godly way"? AFAIR it just requires to flip the fan 180 degrees to make it blow the other way around. Correct me if I'm wrong, but I see no reason to go against what MOBO expects here.

If I'm wrong or that's impossible in your case you can try to modify the fan speed with Asus software and check the temperature after some stress test and compare results.

[jaclaz]

12 hours ago, Tripredacus said:

My only concern there is that the board is expecting the that header to be connected as an exhaust fan rather than an intake and thus the temperature detection would be incorrect.

I am not following you. 

The temperature detection is temperature detection, and it works even if you have no fans connected.

The  basic logic behind the temperature detection driven fan controller (what we highly specialized technicians call a thermostat) is (or should be):
I am hot, have the fan blow (or suck) more air until I am cool enough. (i.e. switch to ON)
I am cool, have the fan blow (or suck) less air (until I become hot again, i.e. switch to OFF).

A more sophisticated approach implies to ramp up and down in steps the fan speed.

Now, the point might be if to more fan speed corresponds a reduction in (local) heat or not, i.e. if the fan (no matter if intake or exhaust) is capable (at max speed) to effectively reduce the heat.

@Mcinwwl

Well, no.

The idea (in PC desktops) is to have a "flow" of fresh air, you shouldn't have on opposite sizes two fans blowing (or sucking) air, air should flow from front (and bottom) to back (and top).

Two fans blowing in would create a (strong) positive pressure inside the case, i.e. (hot) air will need to exit from wherever it can.

Two fans blowing out would create a (strong) negative pressure inside the case i.e. (fresh) air will need to enter from wherever it can.

What is usually wanted is neutral pressure and a definite air flow:

https://www.neweggbusiness.com/smartbuyer/over-easy/pc-cooling-how-to-set-up-computer-case-fans/

https://www.howtogeek.com/303078/how-to-manage-your-pcs-fans-for-optimal-airflow-and-cooling/

and that depends from a number of factors, including the capability of the fans (and the amount of heat, and the response of the fan controller(s), etc.)

In theory a fixed speed intake fan and a variable speed exhaust one tends to create a (slightly) positive pressure, whilst a fixed speed exhaust fan and a variable speed intake one will tend to create a negative one, but it all depends on the capacity and actual speed of the fans.

 jaclaz

 

 

[Tripredacus]

Mcinwll, from the picture I posted, there is a front fan that brings air in, this air goes over the two spindle disks (the fan is actually part of the drive enclosure), then goes over the motherboard, then the rear fan sends it out the back. The CPU is in direct line with this design, obviously it is pushing air away from the CPU. The PSU's fan actually faces the CPU as well. Basically you could draw a straight line from the front to the back of the chassis for airflow, with some turbulence happening inside due to CPU, PSU and video card fans.

I know the rear (output) fan's speed because it is reported in the BIOS. The front fan (intake) is much faster. I can tell it because it is louder, but also because it is connected directly to the PSU with no controller in-between. You can see a bare view of this chassis type here: https://ru.gecid.com/cases/chenbro_sr20969/
I say "type" because the link is for the SR20969, which has some slight differences, but the fan placement and the drive bay is the same as the SR105.

As far as thermostats go with fans, I was under the presumption that the placement of the thermostat in relation to the fan make a difference, it is the reason I am posting the thread. My concern was that if I connected the front fan to the fan header (a header designed for a rear fan) would it be an issue.

It sounds to me that a slight negative pressure would be better than a slight positive one.

If there is any sort of fan controller I can connect to the PSU and not to the board (or perhaps to the board using an un-used header like the USB3.0 connector) that has its own thermostat, I can connect the front fan to that, in addition add the optional front fan to the bottom section to help out the video card. Note that the video card I have is one of the newer types that does not turn the fans on until it his a low temperature threshhold. It is kind of weird to be looking inside of the machine when it is on and the fans aren't moving.

[jaclaz]

Curously we are having a strictly related thread (cross posting):

 

What you may want is a thermostatically controlled variable speed fan (example) :

https://www.arctic.ac/eu_en/arctic-f12-tc.html

Such devices are independent from the PC board and from manual controls, they simply have a temperature sensor and run according to temperature.

The TC in the model name should mean Temperature Control:

https://www.arctic.ac/eu_en/products/cooling/case-fan.html

Are your fans PWM (i.e. 4 pins)?
https://www.ekwb.com/blog/what-is-pwm-and-how-does-it-work/

If yes, you may look for a PWM fans hub, again example:

https://www.enostech.com/silverstonetek-1-to-8-pwm-fan-hub-sst-cpf04-review/

jaclaz

Edited March 18, 2020 by jaclaz

[Mcinwwl]

9 hours ago, jaclaz said:

The idea (in PC desktops) is to have a "flow" of fresh air, you shouldn't have on opposite sizes two fans blowing (or sucking) air, air should flow from front (and bottom) to back (and top).

I get that, and I knew that before, but I was also

9 hours ago, jaclaz said:

not following you. 

'you' being the topic author in this case. Now it gets clearer.

What makes me wonder, is why the CPU cooling exhausts 'to the side'. What I usually see in gaming rigs is some big-a** radiator and a fan that blows front-to-back, to play in the same team as rear and front one. Ant those cooing systems are to expensive, when compared to other parts you bought.

Edit: I even checked the internet if my memory is not playing tricks on me and it is not, such setups exist: https://images.morele.net/listy_zakupowe/4048673/10272_full.jpeg

With such setup, the only problem remains:

9 hours ago, jaclaz said:

In theory a fixed speed intake fan and a variable speed exhaust one tends to create a (slightly) positive pressure, whilst a fixed speed exhaust fan and a variable speed intake one will tend to create a negative one, but it all depends on the capacity and actual speed of the fans.

and "slightly" positive or negative pressure sounds a minor troublemaker, when compared to changing the flow direction by 90 degrees in the mid-way.

 

If, CPU cooling remains like it is, I need to quit now, as I lack knowledge how to fine tune the rest. However, I remain interested in the outcome, as this is simply interesting.

Edited March 18, 2020 by Mcinwwl
added image.

[jaclaz]

The reality is that noone spends any time to actually do some sane calculations on airflow and air dynamics inside a case as each case and each CPU fan/heatsink are different and if you attach to them any number of case fans, different GPU's (and their fans), besides the PSU one, with different airflows you will have an almost infinite number of possibilities, then you add to it the external variable of the room temperature (that can be anything between - say - 16 and 32 C ) and the (large) variations in heat produced by the CPU and by the other components and in the end you find out that the only possible solution is actually  the current "solution" (if the "normal" CPU and PSU fans are not enough) add more fans and have some thermostatic control of them to slow them down when there is no need for them.

Even with the (in theory) more advanced solution (the PWM fans controlled by a single sensor via a PWM hub) there will anyway be some differences unless all case fans are exactly the same make and model. (in the sense that even if the PWM cotrol sets fans at - say - 60% - one will likely be at 50% ad one at 70%, and the air flow produced by them is not linear) so ultimately the best approach is to experiment.

A nice set of experiments:

https://www.bit-tech.net/reviews/tech/the-big-cooling-investigation/1/

jaclaz

 

 

[Tripredacus]

17 hours ago, Mcinwwl said:

What makes me wonder, is why the CPU cooling exhausts 'to the side'. What I usually see in gaming rigs is some big-a** radiator and a fan that blows front-to-back, to play in the same team as rear and front one.

It depends on the CPU and how you are using it. I have seen the Ryzen fans that look just like that, but a desktop Intel CPU isn't going to come with such a fan. I'm not overclocking nor trying to push the limits on anything. Most of the high-end video settings are not even noticeable to me and if turning things down increases FPS with no discernable change in graphics quality, I'm going to do that. And another thing is cost. The PCs that I build are done as cheap as possible. I mean, the case I'm using I found in the trash, that is just one example.

I'll take a look at those fans with the built-in temperature sensor. The fans I have now are the "2 pin" variety, despite the board having 4 pin sockets and 3 or 4 pin connectors, there are only 2 pins on either the front or rear case fan. Only the CPU fan is 3 pin.

There was never any real reason for me to even have a front fan, other than the case had one and it seems like a good idea.

[jaclaz]

Yep, but what is the motherboard?

Has it a 3 pin or a 4 pin header?

And if it is a 4 pin, is it PWM?

Some details:

 https://noctua.at/en/how-can-i-check-if-my-4-pin-fan-header-supports-pwm

Basically, you can regulate both a 2 Pins and a 3 Pins fan but ONLY by lowering the (theoretical) 12 V on pin 2 (i.e. using a voltage regulator).

The issue is that this kind of voltage regulating is not "ideal" because most fans won't simply rotate below a given voltage threshold, let's say below 5 Volts, so you do not have the possibility to regulate from 0 RPM to max (let's say 1000 RPM) but only (still say) between 400 and 1000 RPM.

Here is am example of a "classic" variable speed temperature driven controller:

http://www.heatsink-guide.com/tempcontrol.htm

but - really - nowadays it makes no sense to build this kind of stuff (if not for the fun of it) when a TC fan can bought for anything between 5 and 12 Euro. 

The advantage with PWM (four pins) fans is that the voltage on pin 2 remains constant at 12 V whilst the PWN acts as a high speed cut-off, allowing for a wider range (and also more accurate) speeds.

jaclaz

 

[jumper]

For 2 and 3 pin fans, use a chopper / pulse-width controller circuit.

 

[Tripredacus]

22 hours ago, jaclaz said:

Yep, but what is the motherboard?

It is the Asus H110M-C/CSM: https://www.asus.com/us/Motherboards/H110M-C-CSM/specifications/

According to the manual, both fan headers support PWM.

[jaclaz]

3 hours ago, Tripredacus said:

It is the Asus H110M-C/CSM: https://www.asus.com/us/Motherboards/H110M-C-CSM/specifications/

According to the manual, both fan headers support PWM.

Good , then, if I were you (and you do know how cheap  I am) I would get a PWM Y cable, example:

https://www.quietpc.com/gel-pwm-cable

and a couple PWM (4 pin) fans, again example, of suitable airflow/speed:

https://www.quietpc.com/120mmfans

and call it a day.

Check also this:

https://www.overclockers.com/pwm-fan-roundup-twenty-four-120-mm-case-fans-tested/

Since you shouldn't be in a situation of extreme need for cooling/airflow I would go for more silent fans.

If you can get a couple of Arctic ones, you won't need the Y cable, if you get one with the "pass through" connector (that they call "PST"). i.e.:

PST:

https://www.arctic.ac/eu_en/f12-pwm-pst.html

vs. "normal":

https://www.arctic.ac/eu_en/arctic-f12-pwm.html

and actually, since the cost is not that much different, I would personally go for a P12 (as opposed to a F12) as it should be more silent (and also efficient) in "normal" operation:

https://www.arctic.ac/eu_en/p12-pwm-pst.html

jaclaz