How much ventilation should my custom case have?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Devonavar
How much ventilation should my custom case have?
So first my question without much context - I'm planning a build in a custom case, cooled by a single 120mm exhaust fan. How much area should I have for intake? An 120mm round exhaust is around 17.53 square inches. Should I have the same area for intake? If I only have, say, half that, or a third of that, will the air moving faster going in be a significant noise source? Would there be other issues?
More context: I'm having a custom wooden case made for planned new build. The idea is to have a solid front, top, and sides. I expect that to be good for keeping noise from reaching my ears (and to look good), but it means ventilation holes go in the bottom, and the larger they are the larger the case is, roughly. It also affects how tall the feet for the case need to be. The planned components (discussed some in this thread) are an i3-530 with a Thermalright-HR01 Plus connected with a duct to the exhaust fan, a Radeon HD 5570 graphics card, a PicoPSU, and an SSD. I expect it to use around 90W under load.
More context: I'm having a custom wooden case made for planned new build. The idea is to have a solid front, top, and sides. I expect that to be good for keeping noise from reaching my ears (and to look good), but it means ventilation holes go in the bottom, and the larger they are the larger the case is, roughly. It also affects how tall the feet for the case need to be. The planned components (discussed some in this thread) are an i3-530 with a Thermalright-HR01 Plus connected with a duct to the exhaust fan, a Radeon HD 5570 graphics card, a PicoPSU, and an SSD. I expect it to use around 90W under load.
-
- Friend of SPCR
- Posts: 2887
- Joined: Mon Feb 28, 2005 4:21 pm
- Location: New York City zzzz
- Contact:
If I was you, i you want a cold case....
install two 120mm fan holes or at least one in the bottom pointing upwards. One exhaust is never as cool as in and out setup. I agree with solid front, sides and top. the PSU should be on bottom too if it is a fanned unit. I would make it isolated then. I use fanless so I must put my psu near the exhaust fan where it can act as a window to draw outside air through it. This normally is the top.
The more I think about it, the more a wooden case is superior and much nicer looking.
install two 120mm fan holes or at least one in the bottom pointing upwards. One exhaust is never as cool as in and out setup. I agree with solid front, sides and top. the PSU should be on bottom too if it is a fanned unit. I would make it isolated then. I use fanless so I must put my psu near the exhaust fan where it can act as a window to draw outside air through it. This normally is the top.
The more I think about it, the more a wooden case is superior and much nicer looking.
Re: How much ventilation should my custom case have?
You want positive pressure to avoid dust buildbup.Ondo wrote:So first my question without much context - I'm planning a build in a custom case, cooled by a single 120mm exhaust fan. How much area should I have for intake? An 120mm round exhaust is around 17.53 square inches. Should I have the same area for intake? If I only have, say, half that, or a third of that, will the air moving faster going in be a significant noise source? Would there be other issues?
More context: I'm having a custom wooden case made for planned new build. The idea is to have a solid front, top, and sides. I expect that to be good for keeping noise from reaching my ears (and to look good), but it means ventilation holes go in the bottom, and the larger they are the larger the case is, roughly. It also affects how tall the feet for the case need to be. The planned components (discussed some in this thread) are an i3-530 with a Thermalright-HR01 Plus connected with a duct to the exhaust fan, a Radeon HD 5570 graphics card, a PicoPSU, and an SSD. I expect it to use around 90W under load.
Use one 120mm fan in and one out. The one going in should push more air than the one pulling it out.
Use the fan in to push fresh cool air into the HR-01 fin array.
-
- Friend of SPCR
- Posts: 2887
- Joined: Mon Feb 28, 2005 4:21 pm
- Location: New York City zzzz
- Contact:
Re: How much ventilation should my custom case have?
Yeah, I agree here with more pressure from bottom fan to be more than the pressure from the exhaust fan, in other words, one at like 10 volts the other at like 7 volts. I would even opt for 2 on the bottom if it was a deep case. You wont be hearing them if they are moving very slowly. thats a lot of surface area with minimal noise.ces wrote:You want positive pressure to avoid dust buildbup.Ondo wrote:So first my question without much context - I'm planning a build in a custom case, cooled by a single 120mm exhaust fan. How much area should I have for intake? An 120mm round exhaust is around 17.53 square inches. Should I have the same area for intake? If I only have, say, half that, or a third of that, will the air moving faster going in be a significant noise source? Would there be other issues?
More context: I'm having a custom wooden case made for planned new build. The idea is to have a solid front, top, and sides. I expect that to be good for keeping noise from reaching my ears (and to look good), but it means ventilation holes go in the bottom, and the larger they are the larger the case is, roughly. It also affects how tall the feet for the case need to be. The planned components (discussed some in this thread) are an i3-530 with a Thermalright-HR01 Plus connected with a duct to the exhaust fan, a Radeon HD 5570 graphics card, a PicoPSU, and an SSD. I expect it to use around 90W under load.
Use one 120mm fan in and one out. The one going in should push more air than the one pulling it out.
Use the fan in to push fresh cool air into the HR-01 fin array.
-
- Posts: 266
- Joined: Wed Mar 31, 2010 8:19 am
- Location: OV, The Netherlands
It'll use around 90 watts under load, and the PicoPSU I'm getting supplies 150. It's not even close to a problem.BlackWhizz wrote:You wont run that CPU and Video card of a PicoPSU. Thats just impossible.
I'm much more worried about noise that about it being cool. I'm hoping to be able to get enough cooling with a single fan at low speed.~El~Jefe~ wrote:One exhaust is never as cool as in and out setup.
From what I've read, if I go positive pressure and add dust filters, it'll make a bit more noise than if I go without. I'm leaning towards maximizing silence, and just using compressed air to clean it out periodically.ces wrote:You want positive pressure to avoid dust buildbup.
Negative pressure is no more or less likely to generate noise than an equal amount of positive pressure. It will depend to a large extent on the fan.Ondo wrote:BlackWhizz wrote:From what I've read, if I go positive pressure and add dust filters, it'll make a bit more noise than if I go without. I'm leaning towards maximizing silence, and just using compressed air to clean it out periodically.
Try this:
Use a Seasonic x650 PSU. If you keep the power usage of your system below 300watts (depending on ambient temp) the fan will never kick in.
Use an HR-01 plus with the optional thermalright duct (red is the short one, blue is the long one). Use that to direct cool air in - into the HR-01 plus fan array. Cooler outside air means you need less air current. Use a slow 500 rpm fan.
You really need to get some kind of slight air movement around your video card. It doesn't take much, just a whiff of current to change the stale hot air around its fins will make a world of difference. If you do a tower format, at least put a 120mm blow hole above the card.
One you get all that working, reconsider useing a FilterRight filter over the intake. Theoretically that might generate noise. But maybe nothing you will hear.
I bet a 500rpm Gentle Typhoon will not generate any more noise with the filterright than without. It will not even notice the additional impedance of the filter.
-
- Posts: 15
- Joined: Thu Sep 21, 2006 6:44 pm
- Location: here
-
- Friend of SPCR
- Posts: 2887
- Joined: Mon Feb 28, 2005 4:21 pm
- Location: New York City zzzz
- Contact:
positive pressure cools very well and if they are mounted underneath a case at like 7 volts, you can't hear them. You need to rubber mount them too, they make stand off plugs for that, something could be fashioned like that for a wood case.
If you only had one fan and wanted to game, rear top exhaust is best.
However, you can more silently cool with slower moving fans and more of them. One medium speed fan can be quite noticable. THere is no need for pico, i hate that thing. Complicates builds. x650 or a Nightjar wont run fans yet produce a lot of power. You would have to have a really slow system to run pico. On board video and single core type of thing. I guess an amd dual core super efficient one could work on a pico setup for a media box.
Positive pressure causes turbulence in a case cooling down lots of parts. making a distinction of cool vs quiet wont make sense as cool is quiet. hot and silent .... means you failed!
a 5570 draws a lot of current. 150 watts total might be fine. hardrives draw 6-9 watts each. motherboards about 40. you are very very close to popping the max wattage on a 150 if you throw in a game. Fans use a 1 or so a piece. Cold components take away a few % of wattage draw....
If you only had one fan and wanted to game, rear top exhaust is best.
However, you can more silently cool with slower moving fans and more of them. One medium speed fan can be quite noticable. THere is no need for pico, i hate that thing. Complicates builds. x650 or a Nightjar wont run fans yet produce a lot of power. You would have to have a really slow system to run pico. On board video and single core type of thing. I guess an amd dual core super efficient one could work on a pico setup for a media box.
Positive pressure causes turbulence in a case cooling down lots of parts. making a distinction of cool vs quiet wont make sense as cool is quiet. hot and silent .... means you failed!
a 5570 draws a lot of current. 150 watts total might be fine. hardrives draw 6-9 watts each. motherboards about 40. you are very very close to popping the max wattage on a 150 if you throw in a game. Fans use a 1 or so a piece. Cold components take away a few % of wattage draw....
for every 90 degree turn in airflow you will reduce the amount of flow by, IIRC, 12%. If you make your 17.5 sq-in hole be sure it's a direct shot into the case.
If it's redirected air, like in a 3480, P18x, Solo, etc you should account for this 12% loss in your intake size.
This, of course, assumes you have maximized the flow through your 120mm exhaust.
If it's redirected air, like in a 3480, P18x, Solo, etc you should account for this 12% loss in your intake size.
This, of course, assumes you have maximized the flow through your 120mm exhaust.
1. Every cubic inch of air the goes in will force a cubic inch of air out. If the exit is smaller than the intake, it will just cause the air to flow faster through the exist to make up for its smaller size. The faster airflow, at leasst in theory, has the potential to generate noise.psyopper wrote:for every 90 degree turn in airflow you will reduce the amount of flow by, IIRC, 12%. If you make your 17.5 sq-in hole be sure it's a direct shot into the case. If it's redirected air, like in a 3480, P18x, Solo, etc you should account for this 12% loss in your intake size.
This, of course, assumes you have maximized the flow through your 120mm exhaust.
2. Where does that 12% figure come from?
3. I believe that air currents inside a case are too complex to use formula intended for simple ducts.
Sure - it's the filter I'm worried about.ces wrote:Negative pressure is no more or less likely to generate noise than an equal amount of positive pressure. It will depend to a large extent on the fan.
Whether I go positive pressure or negative, there just aren't going to be many holes in this case. Either way, checking to see if I can put a filter over any intakes without increasing noise is a good idea, and I'll try for that.
You're using a power calculator. It's wrong. Motherboards do not draw about 40 - not even close. SPCR's review of the i5-661 shows an entire system drawing 69W under load (here).~El~Jefe~ wrote:a 5570 draws a lot of current. 150 watts total might be fine. hardrives draw 6-9 watts each. motherboards about 40.
Cool and audible also means I failed. A blanket statement that 2 fans are cooler than 1 is not useful - they're also louder. I'd love to know whether my system can be cooled quietly with just one, or if two would be better, but there are a lot of factors that go into that. My best guess is that one will work, because the system uses so little power.~El~Jefe~ wrote:making a distinction of cool vs quiet wont make sense as cool is quiet. hot and silent .... means you failed!
Right. But if you can't get it in, you can't replace it.ces wrote: 1. Every cubic inch of air the goes in will force a cubic inch of air out. If the exit is smaller than the intake, it will just cause the air to flow faster through the exist to make up for its smaller size. The faster airflow, at leasst in theory, has the potential to generate noise.
I heard it from an old aerospace engineer who used to to turbine and forced induction work. You should have heard him talk about efficiencies across an intercooler. This seems to speak to what I am referring to:ces wrote: 2. Where does that 12% figure come from?
http://www.engineeringtoolbox.com/minor ... d_208.html
Again, I'm not referring to flow within the case. I am talking about restricted intakes.ces wrote: 3. I believe that air currents inside a case are too complex to use formula intended for simple ducts.
So a sharp bend increases velocity but a vaned bend decreases velocity?psyopper wrote:I heard it from an old aerospace engineer who used to to turbine and forced induction work. You should have heard him talk about efficiencies across an intercooler. This seems to speak to what I am referring to:
http://www.engineeringtoolbox.com/minor ... d_208.html
"90o bend, sharp 1.3
90o bend, with vanes 0.7"
On this point I believe you are in error.psyopper wrote:Right. But if you can't get it in, you can't replace it.ces wrote: 1. Every cubic inch of air the goes in will force a cubic inch of air out. If the exit is smaller than the intake, it will just cause the air to flow faster through the exist to make up for its smaller size. The faster airflow, at leasst in theory, has the potential to generate noise.
As you decrease the circumference of the exit, you will increase the speed of the air. I think there is even a name for this "the venturi effect".
To the extent the restriction increases impedance, it will reduce the total amount of air coming in and going out, but whatever is leaving through the smaller hole, will be moving faster than what is coming in the larger hole.
OMFG. Are you serious? Read the page again, it's titled "Minor Loss Coefficients". You have a 1.3 Coefficient LOSS in flow around a 90 degree bend; you have a 0.7 Coefficient LOSS in flow around a vaned bend, I presume because the flow of air is assisted by the vanes. Neither of those increases the flow of air, the vanes just decrease the flow less.ces wrote:So a sharp bend increases velocity but a vaned bend decreases velocity?psyopper wrote:http://www.engineeringtoolbox.com/minor ... d_208.html
"90o bend, sharp 1.3
90o bend, with vanes 0.7"
Sure.... but you completely misunderstand my point. If the intake is so restricted that it reduces the amount of air coming in, then it also decreases the amount of air avialable to exit your venturi at a faster speed.ces wrote:On this point I believe you are in error.psyopper wrote:Right. But if you can't get it in, you can't replace it.ces wrote: 1. Every cubic inch of air the goes in will force a cubic inch of air out. If the exit is smaller than the intake, it will just cause the air to flow faster through the exist to make up for its smaller size. The faster airflow, at leasst in theory, has the potential to generate noise.
As you decrease the circumference of the exit, you will increase the speed of the air. I think there is even a name for this "the venturi effect".
To the extent the restriction increases impedance, it will reduce the total amount of air coming in and going out, but whatever is leaving through the smaller hole, will be moving faster than what is coming in the larger hole.
And as usual you seem to have usurped a thread for your own meandering purpose. Good on ya!
Just prior to the equation, the title reads"Minor Loss - Head or Pressure Loss in Air Duct Components - can be expressed as". On a linked page it describes "minor losses" of pressure as follows: "Losses occur in straight pipes and ducts as major loss and in system components as minor loss. Components as valves, bends, tees add head loss common termed as minor loss to the fluid flow system. The minor loss can be significant compared to the major loss - in fact when a valve is closed or nearly closed the minor loss is infinite. For an open valve the minor loss may often be neglected (a full bore ball valve)."psyopper wrote:OMFG. Are you serious? Read the page again, it's titled "Minor Loss Coefficients". You have a 1.3 Coefficient LOSS in flow around a 90 degree bend; you have a 0.7 Coefficient LOSS in flow around a vaned bend, I presume because the flow of air is assisted by the vanes. Neither of those increases the flow of air, the vanes just decrease the flow less.ces wrote:So a sharp bend increases velocity but a vaned bend decreases velocity?psyopper wrote:http://www.engineeringtoolbox.com/minor ... d_208.html
"90o bend, sharp 1.3
90o bend, with vanes 0.7"
So I thought that was an equation describing how to predict the loss of pressure as a result of various components added to a duct. Those two numbers just didn't seem consistent to me.
I guess I just don't understand those coefficients as well as you do.
Oh I thought you were talking about a 120mm intake hole and a 92mm exit hole. Though in either case the volume of air going in one hole will equal the volume of air going out the other. The air compensates for the restriction of the smaller hole by moving faster through the smaller hole. (At really high pressures the exiting air might compress a bit, but not at the pressures we're talking about.)psyopper wrote:Sure.... but you completely misunderstand my point. If the intake is so restricted that it reduces the amount of air coming in, then it also decreases the amount of air avialable to exit your venturi at a faster speed.
-
- Friend of SPCR
- Posts: 2887
- Joined: Mon Feb 28, 2005 4:21 pm
- Location: New York City zzzz
- Contact:
You might want to consider that you are really annoying. Go make you your own case then and don't post.Ondo wrote:Sure - it's the filter I'm worried about.ces wrote:Negative pressure is no more or less likely to generate noise than an equal amount of positive pressure. It will depend to a large extent on the fan.
Whether I go positive pressure or negative, there just aren't going to be many holes in this case. Either way, checking to see if I can put a filter over any intakes without increasing noise is a good idea, and I'll try for that.
You're using a power calculator. It's wrong. Motherboards do not draw about 40 - not even close. SPCR's review of the i5-661 shows an entire system drawing 69W under load (here).~El~Jefe~ wrote:a 5570 draws a lot of current. 150 watts total might be fine. hardrives draw 6-9 watts each. motherboards about 40.
Cool and audible also means I failed. A blanket statement that 2 fans are cooler than 1 is not useful - they're also louder. I'd love to know whether my system can be cooled quietly with just one, or if two would be better, but there are a lot of factors that go into that. My best guess is that one will work, because the system uses so little power.~El~Jefe~ wrote:making a distinction of cool vs quiet wont make sense as cool is quiet. hot and silent .... means you failed!
how irritating.
All Ondo is doing is questioning what you are saying and proving the basis for his questioning. In my opinion you shouldn't make him feel unwelcome here because of that.~El~Jefe~ wrote:You might want to consider that you are really annoying. Go make you your own case then and don't post.
how irritating.