Water Cooling: It's better down where it's wetter

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > Water Cooling: It's better down where it's wetter

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

PerrineClostermann posted:

Any goons thinking of using water? Any else already cooling with water?

I am, and honestly it was a mistake. Air cooling will always provide better cooling with lower noise levels because heat pipes move heat more effectively than pumped water, without the noise of a pump. Water cooling only makes sense in situations you need to pump the heat to a remote radiator due to space constraints, like in SFF PCs. On the plus side, putting a water cooler on your GPU lets you exhaust heat from the case in a manner similar to a blower, but without the noise.

# ¿ Aug 11, 2016 21:11

Adbot: ADBOT LOVES YOU

# ¿ Apr 29, 2024 13:29

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

Well yeah if you are willing to use an external radiator about the size of your case you can get pretty impressive results with low noise, but I don't think that's what most people are considering when they think about water cooling.

# ¿ Aug 11, 2016 22:54

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

Shrimp or Shrimps posted:

Don't see why not. I live in SEA and summers here hit 34~5c at 90% humidity and I watercooled just fine without an aircon. In the end, watercooling just keeps the delta between ambient and max temps lower.

Unless you meant trying to actually cool your bedroom with water?

Linus was using a bedroom as an office that was full of a bunch of high-end computers that heated the room up very badly. The idea was to make a massive shared watercooling system that would pump the heat from all of the systems into a radiator outside the house, keeping the room and thus the people and computers inside cooler. It turns out that it basically didn't work for poo poo. It was way more work than expected, there were leaks, and most of the heat was lost in the pipes before it got outside anyway so the room was still hot and cooling still sucked. And then after it was all done they got evicted and had to tear it down to move! Here's the start of the saga:
https://www.youtube.com/watch?v=b8bLtg9J1Oc

# ¿ Sep 8, 2016 18:09

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

Is anyone aware of any large (200mm+) static-pressure fans better than the BitFenix Spectre Pro 230mm? It's a 200x30mm fan with 156CFM and 1.81mmH2O at 900rpm, which seems pretty decent. I just wish there was an option for a higher max RPM to achieve higher pressure, but I guess I can overvolt it to 14V or so.

I figured this was a good thread to ask since people mostly need static-pressure fans for radiators. My actual goal is experimenting with filtered positive pressure ventilation to make a dust-free computer, versus the normal case filters that just catch some dust but clog quickly. Since filters restrict airflow I need as much pressure as I can get.

# ¿ Sep 28, 2016 22:25

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

rage-saq posted:

Reminder: don't use prime95 as it's all SSE accelerated and causes a small part of the COU to get to like 90c.
Intel Xtreme Tuning Utility has a good load test

Prime95 is cool and good, the point is that you don't need to freak out if your CPU gets too hot during the AVX2 test, because the CPU will crank up the core voltage when AVX2 is used (just like it does when you run a real app). You shouldn't just avoid testing your CPU stability because you're worried a particular test won't pass.

# ¿ Oct 25, 2016 00:54

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

Just a quick clarification, there's two kinds of PWM: Using the 4th pin of the header to send a PWM control signal to the fan's built-in speed controller, and actually applying PWM to the 12V power for the fan. I think the former should be available on any 4-pin header. The latter is rare and used to control fans that don't support speed control and don't react well to voltage control. I'm not sure what Don Lapre is saying these Corsair fans need but since they have 4-pin plugs I would expect them to work from any 4-pin header on your board, but I haven't looked at your specific board.

Alereon fucked around with this message at 07:15 on Nov 27, 2016

# ¿ Nov 27, 2016 07:12

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

Don Lapre posted:

A lot of motherboards don't actually send a pwm signal on their 4 pin headers but a constant 5v. Many boards only actually send a pwm signal on the CPU header.

Thanks for the info, that's ridiculous as there's no reason to even include a 4-pin connector if you're not going to send a PWM control signal, 4-pin fans plug into 3-pin headers just fine. I guess I'm spoiled by good motherboards with speed control on every header.

# ¿ Nov 28, 2016 19:22

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

PerrineClostermann posted:

And to complicate PWM, I've found there are headers/controllers that don't actually send a PWM control signal on a 4th pin, but still claim PWM. Turns out they apply PWM to the 12v power supply, turning fans on and off rapidly. Sigh.

To be fair that is actually a totally legitimate way to control speed and is usually better than voltage-based speed control if the fan doesn't have a built-in controller. It's how you control larger motors as well.

# ¿ Nov 28, 2016 20:02

Adbot: ADBOT LOVES YOU

# ¿ Apr 29, 2024 13:29

Alereon: Feb 6, 2004; Dehumanize yourself and face to Trumpshed; College Slice

Yeah I agree, there's no reason for it to be such a clusterfuck at all.

# ¿ Nov 28, 2016 20:05

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > Water Cooling: It's better down where it's wetter