2016/02/27 03:18:08
Bristol_Jonesey
Yes, I see exactly where you're coming from.
 
For now I'm going to run it without the cover on but monitor the temperatures critically.
 
Being mounted in a rack means dust build up is a lesser concern than temperature
2016/02/27 12:24:35
tlw
mettelus
+1 to this. Bear in mind that heat transfer is proportional to flow rate of the coolant over the surface. The "cooling fin" design of fanless gear is pretty reliant on external airflow being provided (or open to the general atmosphere like air-cooled motorcycles, or when you pulled the cover off).


Though air-cooled engines can get very hot indeed if run without a moving airflow passing over them.

mettelus
 A case fan will not run at high speed if the system is running cool. Other options are to leave the cover off, put holes into the upper portion of the lid (to enhance natural convection), or mount the system so it has holes as the bottom and top (and will function naturally like a chimney). I honestly would go the case fan route first and foremost though.


Personally I wouldn't run with an open case unless the cpu fan and gpu/psu fans (if fitted) were running at a high speed to be certain no overheating happens. Cases can be used to direct airflow to where it is needed though it can take a bit of experimenting to find the optimum fan positions. Cases should do this by design, but many seem not to.

Around 15 years ago a friend's son, then in his late teens, bought himself one of those "ready for extreme overclocking" gaming computers that were fashionable at the time. Perspex case sides, lots of grilles, LED strings lighting up the inside and 12 90mm fans. Cpu, psu, two on the gpu, two each case front and rear and a block of four in a grille on the side blowing straight at the stock Intel cpu cooler and gpu card. The noise level was impressive in all the wrong ways and running speedfan and gpuz showed the cpu, gpu and drives were running hot as Hades, right up near their maximum temp limits.

A little cigar smoke showed that the fans were fighting each other, with areas of dead air inside the case including around the cpu cooler and drive bays. Disconnecting the four fans in the case side brought the temperatures down to the point the system reduced the speed of the remaining fans. Taping over the side fan vents got the temperatures down some more.

Fluid mechanics are strange.

Holes in the top of the case are certainly worth a try. My Fractal Design case has a sliding panel that lets you open a grille in the case top. In a vertical tower case leaving the top 5.25" drive bay empty and open to let air out of the top of the case can work well.
2016/02/27 15:42:15
mettelus
+1 to all of the above. The below is more clarification/edification on the reasoning for case fans being the number one choice if anyone wants to read or this thread pops up in years to come. Regardless, temperature monitoring should always be used to guide performance of any modifications chosen.
 
tlw

Though air-cooled engines can get very hot indeed if run without a moving airflow passing over them.


 
Anyone who has ridden an air-cooled bike hard and then stopped at an "epic" red light can attest to this. I should have clarified that a bit (the point I really wanted to make is what tlw said above), since when in motion it is fine and dandy, but when stopped the rider is sitting on a ""slow cooker" which is only getting heat transfer from natural convention.
 
[more of an aside here] Just to clarify (and throw a little engineering in for fun), heat transfer is driven by delta T (difference from heat source to heat sink), with the mass flow rate and specific heat capacity of the fluid proportional to the heat transferred (typically written as q=mCΔT, or variations thereof). For fluid-flow heat transfer, air is considered a "fluid." Practical application everyone can attest to is "wind chill" - going into a cold environment is not so bad with no air flow, but in a windy environment heat transfer is noticeable immediately, driven also because the "laminar layer" ("dead air") near the skin's surface gets "thinner" without warm clothes to trap it there. An important point with this is "wind chill" cannot make anything colder than the delta T; it just jacks up the heat transfer rate. [/aside]
 
Both of the above fit into the recommendation for case fan(s). The case is not only a protective cover, but also forces air flow in a specific direction, pulling air from the box near the CPU/PSU and forcing it into the room (hopefully) on the other side of the case. Without the "constraint" of maintaining cold flow over the heat source (maintaining delta T at a good value), you run into "natural convection" which is more vertical like how a chimney works (or the example of sitting at a red light on a motorbike). The case fan also helps to stir the heat coming out so that it dissipates quicker when it enters the room - depending on flow restraints outside the box it is also possible to pull the heat right back into the front (eventually).
 
Temperature monitoring is highly recommended if you alter "intended design," which may be fine for your situation, but the forced air flow from a case fan does ensure a better heat transfer inside the box (and lower fan speeds all around) - essentially "wind chill" for your CPU, which in this case is a good thing.
2016/02/28 22:23:32
TerraSin
Couple things to consider from my personal builds and experience:
Case: I use cases specifically designed to diminish sound. Both of my computers use Corsair Obsidian 550D cases which have sound dampening material all throughout the case as well as rubber grommets. They really thought of everything when it came to removing sound and I can't recommend this case enough. They are getting harder to find though.
 
Fans: Not all fans are created equal. The bearing type the fans use have a lot to do with reliability and sound output. Sleeve bearings are the cheapest and noisiest of the bearings with the least reliability. Many cases come with these preinstalled. I use all Noctua SSO2-Bearing fans now and they are truly the best and quietest fans I've used which they should be considering the $20 a pop price tag they come with.
 
CPU Heatsink: Much like fans, the fan used on the CPU matters a lot when it comes to noise output. The stock coolers your CPU comes with are junk. They serve their purpose when it comes to keeping a CPU cool for normal use but that's about it. The more the CPU heats up when under stress, the more noise the fan produces as it tries to keep up with the tasks to keep the processor cool. The only real option is to install a good after market cooler. Again, on my newest build I opted for a Noctua NH-D14. This thing is giant but it's extremely quiet. Hands down the best cooler I've used.
 
This is what the final build looked like:

2016/02/29 02:59:35
Sycraft
Noctua fans are what I'd look at for low noise. Of all the ones I've tried, they are kings for low noise.
2016/02/29 12:14:16
patm300e
Sycraft
Noctua fans are what I'd look at for low noise. Of all the ones I've tried, they are kings for low noise.


+1 on Noctua Fans.  For info on a bunch of them:
http://acousticpc.com/noctua_store.html
 
© 2024 APG vNext Commercial Version 5.1

Use My Existing Forum Account

Use My Social Media Account