Overclock.net banner

1 - 19 of 19 Posts

·
EhmayWuntee
Joined
·
1,415 Posts
Discussion Starter #1
I am working on a custom end table build and need some help making sure I have enough airflow for the system.

My original question was just for the PSU which I tried in my build log, but wasn't getting any bites to try and help so I figured the best chance for help is to start a thread here and get the air cooling experts.
I have also decided to ask on the overall PC cooling while I am here and try to be efficient with my thread and especially, your time.

So first up is the PSU, mine will be in its own "closed system" so to speak because it will be secluded on its own.
How much space would you recommend I have available for fresh air to flow into a medium sized closed drawer and keep the PSU from overheating?

Here is a quick photo to help you see what I am building:

The lower drawer will house the PSU (hanging from above so there is a ~3" gap below it) and it will be the only component needing cooling there.
My plan was to have a few holes drilled or something along those lines in the rear of the cabinet and hope the intake fan of the PSU will suck air into the drawer and circulate back out the rear.
Perhaps air will passively get into the drawer from small gaps around the drawer and no holes are needed, or can be kept minimal?
I technically have holes in the bottom of the top drawer to run cables through and that drawer will have intake fans, so maybe that will supply enough passive air for the PSU?
If needed, how many holes or what size opening do you recommend on the rear for the PSU?
To clarify, I don't have any intentions of putting a dedicated intake fan somewhere to pull air into the bottom drawer which can then be picked up by the PSU intake. Hopefully you agree this isn't needed.
I don't imagine a PSU (or at least low wattage ones) to need much cooling. I know a lot of cases even have it where the PSU only drawers in air from the bottom gap between the table/floor and the bottom of the case, so maybe 1/2" vertical space.

The more extreme solution would be to cut a hole in the floor of the lower drawer directly below where the PSU intake fan will sit and it can draw air up from the bottom large opening. This is closer to the idea of some cases.
Even if I add a filter or something so it isn't fully open, I would like to avoid this solution if possible.

Secondly I would like to get clarification on the remaining PC components cooling.
My plan, which can also be viewed from the previous picture, was to have two 140mm fans as intake, one on each side of the table, then there would be a single exhaust out the rear middle of the drawer (120 or 140mm I don't recall).
The CPU cooler fan may also need to be factored in for the airflow, it will just be directing air into the rear exhaust fan. I didn't think this would matter a ton, but was worth detailing at least.
Is this enough airflow to keep the cpu/mobo/gpu cool?
This is not a full on gaming PC or anything high end or crazy overclocked right now (G3258), but I suppose it could be upgraded in the future to an i5/i7 or something if I get bored and find a good deal so I need to slightly thing ahead.
Again though, I plan to keep its use focused on media and maybe minor gaming on the side as a secondary PC. I will always have a different primary rig for my big usage and better components.
This PC will likely also be turned off most of the time and only on when in use.

Thank you for any help/input you can provide and if I need to add clarification somewhere, please ask.
 

·
RGBKB.net
Joined
·
2,460 Posts
I don't have enough time right now to read all, but in regards to the PSU, you will need roughly 10% the airflow required for a normal computer.
If your PSU is Bronze rated or higher, it will be operating somewhere around 90% efficiency. That means for 1000w pulled from the wall, 900w is going to the PC and 100w is being dissipated by the PSU. Hope this helps
wink.gif
 

·
Overclock Failed...
Joined
·
13,565 Posts
Quote:
Originally Posted by mrtbahgs View Post

My plan was to have a few holes drilled or something along those lines in the rear of the cabinet and hope the intake fan of the PSU will suck air into the drawer and circulate back out the rear.

That won't be working. You're planning on air wafting into and out of the drawer. You need the same airflow into and out of the drawer as the PSU fan itself moves.

Perhaps air will passively get into the drawer from small gaps around the drawer and no holes are needed, or can be kept minimal?
I technically have holes in the bottom of the top drawer to run cables through and that drawer will have intake fans, so maybe that will supply enough passive air for the PSU?
If needed, how many holes or what size opening do you recommend on the rear for the PSU?

The above (holes) is foolhardy. Put a PSU in a box with a few holes and in a half hour you'll smell burnt plastic..

Stop this nonsense and mount the PSU with it's fan intake in an opening in the back of the drawer and provide an opening for tha air it intakes to exhaust.

Secondly I would like to get clarification on the remaining PC components cooling.


The fan(s) for cooling the rest of the components should move the same amount of air into and out of the drawer as the fan(s) on the CPU cooler move through it's fins.That can be done with a single fan (intake or exhaust) just so there is the appropriate (un-fanned -Is that a real word? ) opening too
 

·
Premium Member
Joined
·
27,134 Posts
Mount and vent the PSU the same way they are mounted and vented in most cases. Bottom in and back out.

Main computer systems needs a minimum of at least as much air flow through case as components use, but that is not the most important part of what is needed to cool components. The air needs to flow so it supplies air at or near room temperature to components. To do this all of the components heated exhaust air must leave the compartment / case without contaminating / mixing with the cool air going to them .. because if it does mix with the cool air it will raise it's temperature. For this reason the compartment / case airflow volume usually needs to be 1.3 - 1.5 times as much as components actual flow .. and also needs to change in flow rate in unison with component. In other words, case fans need to be speed controlled the same as component fans. 5th post in "Ways to Better Cooling" link in my sig gives a basic guide to case airflow. 1st post is index, click on topics of interest to see them.
 

·
EhmayWuntee
Joined
·
1,415 Posts
Discussion Starter #6
Thanks for the replies thus far.

I will post a few other pics when I get home to better clarify what it going on, the way i interpret your replies leads me to believe we aren't quite on the same page for this setup.

Bill, I do have a dedicated opening in the rear of the drawer for 100% of the exhaust to completely leave the enclosure very easily, kind of like in your photo.
As far as the intake side, I assume your photo has the PSU fan facing upwards and drawing air down into the psu?
Mine would be kind of like this in the sense of the gap with free air moving in, just that mine is upside down with the gap between the fan and bottom of the drawer.
If your model should work though where the single large opening for intake to flow and feed the PSU is fine, then I might be close with the same idea of a few smaller holes to do the same.
The difference now would be if they are very near or under the intake fan of the PSU or if they are just in the drawer and I guess creating a longer path for the air to flow before it reaches the fan.

Doyll, I think I am on target for the components side of things judging by the way I understand your post. I have looked at your links before, but will revisit them tonight.
In regards to the standard bottom in, back out, I was hoping to get by with a slight alteration where its more of a 12" away side in to flow towards the bottom and then back out...
I was hoping the gap would be sufficient to still draw air in and I guess follow a least resistance type of path to make it into the intake.

Again I think a photo or 2 will help me get my design across better and then if you still think its asking for trouble or expecting the smell of burnt plastic, then I will obviously have to modify it a bit to be practical.
 

·
Premium Member
Joined
·
27,134 Posts
Just one more thing, the more isolated the cool intake air is from the heated exhaust air the better. If any of the heated exhaust mixes with the cool intake air, the harder the system works and the more heat it makes the more and hotter the exhaust air becomes resulting in the more heat it adds to the cool intake air makeing it even hotter resullting in even higher CPU, GPU PSU temps making the air they exhaust even hotter meaning this hotter air mixing with the cool intake air gets hotter stil and around and around it goes it an every increasingly higher heated air spiral.
tongue.gif
 

·
Registered
Joined
·
86 Posts
Cool project
thumb.gif


A 53w G3258 and a low-end GPU aren't going to generate enough heat that you really need to worry about it as long as you have any reasonable amount of air flowing into the enclosure. If you want to build this thing to work for a future high-end overclocked build however, it starts to get interesting. The intake fan on the GPU side is aimed right at the GPU intake, so that is awesome. You could rotate the CPU heatsink 90° to get the same effect there. If you really wanted to go nuts with it, you could then build ducts from each side intake straight to the GPU and CPU coolers. If you don't want to rotate the CPU heatsink, I would consider moving the CPU-side intake fan as far towards the front of the drawer as possible so that side intake fan will blow cool air in front of the CPU heatsink.
 

·
Registered
Joined
·
2,024 Posts
yes, the main issue is not "how much of a gap" .... because that would (in theory) be like 1 inch allround for a "box" with a 2 foot square base.
The main issue is to prevent air re-circulation... if your PSU spits it out the back, it needs to go somewhere - up/left/right/down... but it should not be allowed to be sucked back in

Assuming the PSU sits fan up in a "box" , exhausing out to the back, but taking in air from inside the box, then the openings , the "air-intakes" of the box should/ be at the front or bottom front, maybe even the sides, providing the intakes extend no further as 1/5th to 1/4 of the side length from the front.

If the PSU sits fan-down and thus taking air from outside the box and exhausting to the rear, then a barrier of sorts must be placed to avoid the PSU taking in the exhaust air.

It sounds more complex as it
smile.gif
Some smart ducting solves the issues.
 

·
EhmayWuntee
Joined
·
1,415 Posts
Discussion Starter #10
Man Doyll that was tough to read lol, I know what you were trying to say the whole time, but I still had to attempt to read it through.
Yes it is a very valid point and if i stay true to only running the system for 2 hours or less during the day and otherwise it is off, I think it wont be much of an issue with a recycled increase in temps (assuming I have much).
If it were on 24/7 or worse, running hard like folding 24/7, then I could see this being a major concern.

Castle, I agree that with the current components I think it is really safe, but yes I was trying to consider future changes as well such as picking up a used i5/i7 to swap in a few years from now, but that isn't a guarantee so I don't want to plan on too much because by then I'd possibly have a different cooler or a radiator from an AIO to cool things and have to cut more of the drawer. So while I want to try and think of the future, I also don't want to do it so much that I am not building for the present
tongue.gif
.

Moving the one side fan forward is a decent idea to help airflow reach the CPU a bit, but I also was tied to the symmetry of the build. So now i guess it falls into form over function to some degree.
The heatsink and side intake fan are like 10" apart or so, wouldn't that be a fair amount of space to where the intake air will disperse wide enough over that distance to feed the heatsink fan if left in its current middle position?
I don't know how noticable it would be if that side intake fan was slid forward 3-4". I guess I am assuming air will move in a cone shape and spread out as opposed to staying perfectly straight as if it's ducted.
Rotating the cooler might help feed the cool air in, but it would then push all the hot air onto the GPU PCB and not direct it out as easily so I don't think that is a better scenario for this build.

RnRollie, I understand what you mean with trying to keep the exhausted air from wanting to go back down and recirculate.
With my initial plans it is probably letting some of that exhaust get sucked back in.
Even though the holes would be under the PSU exhaust port and in theory all the warm air would rise up as opposed to fall down and be pulled in, that likely isnt 100% happening depending if suction is created below and forcing the warm air to flow down instead.

My PSU will be oriented with the fan down, but since the PSU is suspended from the ceiling, the gap for intake air is still the enclosure that the PSU is inside.

I forgot I can just relink some photos I already uploaded in my build log and not have to wait until I get home so here are a few more to try and show the design I have.
PSU hangs off the top drawer and sits in the space of the bottom drawer:



Base cabinet has an opening to let the exhaust freely flow out the back, what is not displayed was my planned addition of a few holes below this to draw in cool air:


This is the bottom drawer, perhaps my only good solution is to cut a hole into the bottom middle directly below the PSU intake and allow air to more easily make it into the fan, plus it is definitely separate fresh air from the exhaust:


I suppose instead of a solid opening which i was hoping to avoid, I could just drill like 10-20 holes in a pattern and get the same effect while avoiding that large hole... This may be the compromise solution.
 

·
Premium Member
Joined
·
27,134 Posts
Sorry, I went a little overboard. The point is it just keeps getting hotter and hotter. But don't think you will be safe for a couple hours. If your design does not have proper airflow the heat buildup when both CPU and GPU are working (like when gaming) can easily cause system to start throttling is 20-40 minutes.

There is no excuse for not doing it right the first time. My guess if if you don't, you will be wishing you had and kicking yourself for not.

Being a retired custom furniture and cabinet builder I'm going to back off. But I have to say your cabinet design is all wrong from beginning to end. Me thinks you need to study up on how to build a proper cabinet first, then incorporate a proper computer case design into it.
 

·
EhmayWuntee
Joined
·
1,415 Posts
Discussion Starter #12
Quote:
Originally Posted by doyll View Post

Being a retired custom furniture and cabinet builder I'm going to back off. But I have to say your cabinet design is all wrong from beginning to end. Me thinks you need to study up on how to build a proper cabinet first, then incorporate a proper computer case design into it.
Interesting, I'd be willing to hear a bit more on this if you have time, or I can certainly research it a bit.
I assume you are referring to the main exterior cabinet and nothing to do with the drawers, correct?
My only thought of what could be very wrong would be if you are saying it isn't very structurally sound, or is it just a different approach that isn't normally seen?
Having a background, experience, etc. to justify your logic and suggestions is welcome and paid much closer attention to than just someone giving an opinion that may have no true weight to it.
 

·
Premium Member
Joined
·
27,134 Posts
Some of both. Basically you don't have a chest of drawers design, nor it seems anything else
I would suggest you first decide what you want the unit to be when done. Simply design it with 2 drawers and top. I would only have a few braces on the back and bottom to give it structural integrity. PSU in bottom drawing air from below and exhausting out back. Computer drawer exhausting out back and drawing air in through side vents, both in drawer and sides of cabinet with them aligning when drawer is closed. If possible have intake vents on left side aligned to flow cool air directly to GPU.
 

·
EhmayWuntee
Joined
·
1,415 Posts
Discussion Starter #14
I follow most of what you are saying and a lot of it is how it is designed, i am a bit unclear on one part though, let me go over all of it quickly and see if you can clarify my confused area.
If this is too much text, the confusion will be at the very end.

Here is a rendering of the final product with everything closed:


The overall unit in its entirety is 24" wide x 18" deep x 24" tall.

The top drawer contains the majority of the PC (mobo, cpu, gpu, ram, and hdd) along with 2 intake fans (one on each side of the drawer and centered) and a single exhaust fan out the rear.
These intake fans will line up to the exterior cabinet openings so air freely moves into the enclosed top drawer and exits the rear.
The left side should line up the air to direct it right at the GPU and the right side will hopefully still provide enough air to flow towards the front of the drawer to enter the heatsink's fan which is blowing front to back and into the rear exhaust.

The bottom drawer contains just the PSU and an ODD which are both suspended off the top drawer so there is a 3"+ gap below both the PSU and ODD (PSU is intake fan side down).
The bottom drawer can be completely removed and it will not require me to unhook anything or remove any cables, its more of a surrounding enclosure.
The airflow for the PSU is still being decided a bit now (main reason for this thread) but worst case there will be holes or an opening directly lined up to the intake fan in the middle of the second drawer to allow air to go up into the PSU and then out the back.

The very bottom opening in the base cabinet is for additional storage and will likely be where I relocate my wireless AIO printer.

The base cabinet is a top and bottom with 2 sides and a back, but no front and a few spots will have holes/openings for air/cables as seen in the photos.

What I am confused on with your reply is I can't tell if you think I don't have enough structural bracing, or you think I have too much for the base cabinet.
I personally don't see how it could be lacking in structural integrity and I am not having like 100 lbs resting on the top or anything.
This is actually fairly similar to some small entertainment cabinets I have seen the only difference being that the entertainment cabinets are usually open shelving throughout and no drawers.
Your mentioning of a chest of drawers design actually sums it up pretty well for what I would consider this to be, a chest of drawers with an open third drawer space at the bottom.

I appreciate the help you (and everyone else) have provided thus far.
 

·
Premium Member
Joined
·
27,134 Posts
If total unit is 24" wide, the inside of drawer is approx. 22" wide. ATX motherboard is 12" x 9" PSU is 5.9" wide x 7.1" x 3.4". Set PSU on edge on right side of 22" wide drawer, this leaves 18.5" of width for motherboard ..that is 2" on each side of motherboard, which is plenty of room to clear anything on it.
  • 2x 140mm vents on left side of drawer aligned with GPU cooler will supply plenty of air for even the biggest GPUs.
  • 1x 140mm vent on right side will supply air for CPU cooler. This vent needs to be close enough to the front so it's air flows smoothly to CPU cooler
  • 1x 140mm vent on right side aligned with PSU intake.
  • Gaming GPUs create 150-300w of heat stock. Not sure of overclocked wattage. Biggest GPU coolers use 3x 92mm fans. 69sq cm x 3 = 207sq cm total area. 2x 140mm fans have 280sq cm airflow area. cfm rating of 92mm fans is 30-35cfm .. 140mm is 65-75cfm .. at full speed. Slower speeds are quieter.
  • Powerful CPUs create 94-130w of heat stock. Overclocked this can be 180-300w. Biggest CPU coolers use 140mm fan airflow.
I would leave back of drawer and cabinet open with only enough framework to support drawer and supply PCIe mounting points. The intake airflow will push air right out the back no problem. No exhaust fans are needed. Just use good fans on intakes

There is still plenty of room in front part of drawer for optical drive, HHD/SSD, etc. Honestly, I do not install an optical drive anymore. I have a slim USB optictal drive in a drawer in case it is ever needed, I use USB sticks to install OS and software.

Cable can go through grommets in bottom of drawer to hide excess length. A double bottom drawer that bottom can be removed to access when running cables would simply installation.

Cabinet only need about 1/3 of back and bottom to be solid for lateral structural stability. Feet on bottom make cabinet stable. A smooth bottom wiggles and rocks on carpet.
wink.gif
 

·
Registered
Joined
·
2,024 Posts
Basically... what it is, is an improved ThermalTake X9
smile.gif

... married to an IKEA Entertainment Center furniture OR to a DOMO Nightstand
smile.gif


Following doylls' suggestions is your best bet to get it to do what it should be doing
 

·
EhmayWuntee
Joined
·
1,415 Posts
Discussion Starter #17
Thanks for the detailed insight, I will take some of those into consideration, but I am sold on setting up my PSU to be hidden and below the mobo so I will have to plan its personal airflow properly.
I just really like the uniqueness in this design and it isn't like every other desk build out there.
Same for the ODD, no i likely don't need it (unless maybe for DVD movies), but I have an extra from an old build and it has a simple spot that is again hidden, so why not.

Not that it really changes anything, but I am using an mATX form factor which if I recall correctly, my board is 9.6"x9.6".
I do like the gpu air calcs, I never thought to look at it that way.
Currently its just a mid range GPU I had from an old RMA, but at some point I will likely switch it to my GTX670 when i upgrade my main rig so I should probably plan for that size and power.

The feet are a good suggestion, having 4 solid points to stand on versus the entire bottom surface, I can see rocking being a potential.
Worst case, I can fall back on this idea after trying out the flat surface since I don't want to rush into buying or building any random feet.

Now i have a bit more information to help plan this out and not build it to overheat on the first power up.
I think the PSU, even though its low wattage, was being underestimated on the airflow needed and it would have caused issues if I kept my original design.
I was likely expecting the PSU fan to work hard to suck air in from where ever it can as opposed to giving it a bit of a break and allowing intake holes to be close by.

As much as I'd like to keep it as minimal as possible, I also don't want to regret my finished product and have to make after thought cuts or even a partial rebuild.
At the same time, this is my first try at an end table like this and I knew it'd be a learning experience to help improve future hobby projects so its all part of the fun and experience.
 

·
Premium Member
Joined
·
27,134 Posts
Keep in mind PSU cables and connectors to MB, GPU, etc. are not designed to be flexed or support movement after being connected. Even repeated plugging / unplugging can lead to connection issues. They are nothing like a wall socket and lamp cord or a headphone jack and plug. They are more the 'plug and forget' kind of connector.
biggrin.gif
 

·
EhmayWuntee
Joined
·
1,415 Posts
Discussion Starter #19
I wouldn't have considered them that sensitive, but certainly didn't expect them to be like a plug in a wall socket, so its good to know they are more sensitive than my first inclination and I can be careful.

The design is made so that it can still be a one time plug and forget because I can completely remove the bottom drawer or top drawer independently and not have to disconnect anything other than the PSU cable from the wall and peripherals.
So I should be all good after initial setup and securing any excess cable slack.
 
1 - 19 of 19 Posts
Top