DIY case cooling

[godinman]

Hi All,

I've built my home office and to keep things tidy I'm going to build my pc into a drawer.

Now I'm going to DIY a sort of test bench build within the drawer using some old cases as donor parts but cooling is where I'm stuck pondering.

I've got room for 4 x 120mm fans front and back, or 1 x 230mm front and back, or pretty much any combination of the above. Side fans aren't an option.

What would people recommend? I could go for 8 x low rpm 120mm which could ramp up if needed I suppose?

Ideally I'd go for the 230mm option but do you think this would provide enough airflow?

I'm going to put filters in front of the fans and then stretchy speaker type material over the front of the whole drawer which will hopefully not restrict airflow too much. This will also keep fan noise down (I'll put filters outside the rear fan(s) for the same reason).

Edited to change fan dimensions.

Thanks.

 

[lude1962]

4x 140 = 280x280, drawer is 260x260?
also wheres the psu going?

 

[godinman]

I've got enough height that I can stack the system and have PSU and HDD's below and motherboard above.

Edit - just checked and available dimensions are -

W - 330
H - 240
D - 340

So it looks like it's between up to 8 x 120mm fans or 2 x 230mm fans.

(Have updated OP).

 

[LuckyBenski]

I would say big fans aren't particularly good at moving lots of air compared to an array of several smaller fans (within reason). I would treat the drawer as a normal PC case albeit one with thick insulating panels and not much venting. As such you might want more airflow than the average PC case. Probably 3 120s in front and same in back.

Because it's not a normal PC case with little vents and cracks everywhere, I doubt positive pressure is of any benefit to keep dust out. So, I'd aim for healthy cross flow. Line the fans up such that main components are in line with the cross flow. Yes, keeping speeds low and ramping up based on temperature should work fine. I'd avoid too low a minimum speed or maybe keep two opposite corner fans on a higher speed all the time. This should avoid the enclosure heating up and killing passively cooled components...

Sounds like an interesting project!

 

[doyll]

Hi All,

I've built my home office and to keep things tidy I'm going to build my pc into a drawer.

Now I'm going to DIY a sort of test bench build within the drawer using some old cases as donor parts but cooling is where I'm stuck pondering.

I've got room for 4 x 120mm fans front and back, or 1 x 230mm front and back, or pretty much any combination of the above. Side fans aren't an option.

What would people recommend? I could go for 8 x low rpm 120mm which could ramp up if needed I suppose?

Ideally I'd go for the 230mm option but do you think this would provide enough airflow?

I'm going to put filters in front of the fans and then stretchy speaker type material over the front of the whole drawer which will hopefully not restrict airflow too much. This will also keep fan noise down (I'll put filters outside the rear fan(s) for the same reason).

Edited to change fan dimensions.

Thanks.

I'm going to assume your fan layout is 2x high and 2x wide 120mm fans? Too bad you don't have enough room for 3x 140 side by side or 2x high and 2x wide.
2x 140mm usually do about the same job as 3x 120mm fans .. usually at lower noise levels. Probably the best way would be to have fans at back of drawer and only filters on front. Much of the depends on your position relative to drawer position. If drawer face is at an angle that sound coming out of it is moving away from you it will be 'quieter' than if it is facing you. The stretchy speaker type material should work for looks and filters.

 

[godinman]

It's something along these lines, I've measured a bit better and can fit 3 x 140mm if I squeeze the bottom one up a bit, so the top fans would be opposite the GPU and CPU while the bottom would be in-between the HDD's and PSU.

I could then either repeat the same config at the front or do mesh or something? (The pic shows a 230mm fan as I was experimenting).

 

[doyll]

Could you lower the motherboard tray so all 3x 140mm fans would be pulling air across it instead of one pulling from below it? 3x 140mm fans will give enough airflow for a big 140mm fanned CPU cooler and 2x-3x fanned GPU cooler .. but not enough for them and PSU. PSU needs another or at least a 120mm one.

Would 6x 120mm fans fit?

Also, removing PCIe slot covers removes lots of airflow restriction. I usually don't use the motherboard I/O cover because it lowers airflow over components between CPU and back of case.

 

[EsaT]

I'm going to put filters in front of the fans and then stretchy speaker type material over the front of the whole drawer which will hopefully not restrict airflow too much. This will also keep fan noise down

Unless having solid object blocking direct noise escape path and forcing sound waves to bounce around noise leaks out from holes completely undamped.

Also big fans at RPMs where their blades move slow enough to not create much noise produce barely any pressure.
Pressure is heavily reliant on RPM, so multiple smaller fans are better able to handle airflow impedances.
Though having smaller fans in rear could help by creating "negative" pressure inside case, if its otherwise solid.

 

[Armageus]

It's something along these lines, I've measured a bit better and can fit 3 x 140mm if I squeeze the bottom one up a bit

Surely the fan frames will clash / won't be able to secure them with screws?

 

[godinman]

Could you lower the motherboard tray so all 3x 140mm fans would be pulling air across it instead of one pulling from below it? 3x 140mm fans will give enough airflow for a big 140mm fanned CPU cooler and 2x-3x fanned GPU cooler .. but not enough for them and PSU. PSU needs another or at least a 120mm one.

Would 6x 120mm fans fit?

I can lower the tray about 30mm before I start getting too close the the PSU. I might be able to get 6 x 120mm to fit, it's tight but it's something I'll check.

Unless having solid object blocking direct noise escape path and forcing sound waves to bounce around noise leaks out from holes completely undamped.

Also big fans at RPMs where their blades move slow enough to not create much noise produce barely any pressure.
Pressure is heavily reliant on RPM, so multiple smaller fans are better able to handle airflow impedances.
Though having smaller fans in rear could help by creating "negative" pressure inside case, if its otherwise solid.

I see, air dynamics are more complex that I thought!

If I stuck to the 140mm fans but had them front&back of the case (push pull) do you think that would allow the air to full through some thin filter material on the front? Or as above would 6 x 120mm be a better bet?

Surely the fan frames will clash / won't be able to secure them with screws?

I've measured up using some of the thin framed fans like Gelid Silent Pro and I should be okay. Worst case I'll be building them on to a frame and then fixing that separately (still internally) but I don't think it'll come to that.

------------

I could increase the width a few mm if I needed to go for the 120mm fans, I'd replace the panel with some honeycomb mesh to maintain strength and give me something to fix to. (So the image below is just for a sizing example rather than a design).

I'm still not set on the front cover, I need to maximise airflow but I don't want visible fan grills and I want it to blend in fairly well with the other drawers in the office. I think I might end up going for honeycomb mesh with white anti-dust stretchy speaker fabric pulled right over it. I just don't know if it'll look naff so am open to suggestions. (currently the front it just a big 230mm hole in the design, not final)

 

[doyll]

120mm fans tend to have slightly higher pressure ratings. New Noctua NF-A12x25 PWM are very good and quiet too .. but they are £29.99 each. Can be found for a little less if you shop around.

The reason I was asking what orientation the front is to your seating arrangement is to determine fan noise direct line to your ears. Having fans at back helps limit noise pattern. There are some fans like NF-A12x25 that will remain almost silent at 1200rpm from a meter or 2 away. PH-F120MP are good .. not as good as NF-A12x25 but are only £10.99 each and if OcUK restocks the 3-packs at £26.00 makes them only £8.67 each.

There are some nice looking filters similar to what speakers use. Maybe one of them will work.

 

[EsaT]

I can lower the tray about 30mm before I start getting too close the the PSU. I might be able to get 6 x 120mm to fit, it's tight but it's something I'll check.


I see, air dynamics are more complex that I thought!

If I stuck to the 140mm fans but had them front&back of the case (push pull) do you think that would allow the air to full through some thin filter material on the front? Or as above would 6 x 120mm be a better bet?

PSU "compartment" really doesn't need that much of airflow:
High end PSUs produce bare 5% of PC's heat output.
So would need to have awfully big pile of HDDs for them to matter in higher end gaming PC.


Having fans in series (both intake and exhaust fan) can be related to adding batteries in series:
Max airflow (current) stays same, but there's more pressure (voltage) capability to overcome resistance.

As for specs of fans (assuming they're honest) airflow is stated for situation in which there's zero back pressure/pressure difference between sides of fans.
That means essentially fan blowing to down wind and not really doing any work itself.
Pressure is again measured in situation when there's zero airflow, with fan basically blowing into/sucking air from closed box.
In actual use their operating point is somewhere between these two extremes.

 

[LuckyBenski]

High end PSUs produce bare 5% of PC's heat output.
So would need to have awfully big pile of HDDs for them to matter in higher end gaming PC.

Surely based on efficiency ratings it's more like 10%+?

 

[EsaT]

Surely based on efficiency ratings it's more like 10%+?

80+ Titanium PSUs reach 96% efficiency with European mains voltage.
80+ Golds are ~92%.
Low US mains voltage and primary side losses from it drops efficiency 2%.
Higher mains voltage has its advantages, because its current which causes voltage/power losses.
Though proper sizing can make similar difference to efficiency.