5
\$\begingroup\$

I have a medium sized electrical cabinet (1100mm wide x 400mm deep x 1100 mm tall) and I am installing 2 cooling vents / fans in the top and bottom opposite corners.

I am only using 1 filtered fan and one filtered vent to keep costs down, should I place the fan at the top or the bottom of the cabinet?

Are there any relevant industrial standards or conventions on this? Or is it a matter of preference and not really significant.

\$\endgroup\$
5
  • \$\begingroup\$ Not sure of standards, but it would be sensible to keep the moving, noisy fan as far as feasible from a user. \$\endgroup\$ Oct 12, 2022 at 23:24
  • 5
    \$\begingroup\$ Keep in mind that some dust inside your cabinet is less of a problem than a clogged filter. So, if the user doesn't have the discipline to keep the filter clean, you don't want a filter. \$\endgroup\$
    – Mattman944
    Oct 13, 2022 at 8:18
  • \$\begingroup\$ @Mattman944 it is an important system design issue. Some equipment may shutdown gracefully in the event of overvoltage. Some cabinets may have sensors and alarms. Etc. But filters always need maintenance and that is something to keep in mind. Also, some dusty environments are worse than others. Sometimes dust inside the cabinet might actually be a big problem. \$\endgroup\$
    – user57037
    Oct 13, 2022 at 17:14
  • 1
    \$\begingroup\$ Never seen a PSU with positive pressure; only extractors. Extracting is better and more even. AC condensers too. You don't blow at 'em; you suck it up and spit it out. Same with plumbing: asking isn't going to work; you make it go away. \$\endgroup\$
    – Mazura
    Oct 14, 2022 at 4:50
  • \$\begingroup\$ Please describe the location of the cabinet - inside or outside? Chance of getting rained on, or a water pipe leaking above ? \$\endgroup\$
    – Criggie
    Oct 15, 2022 at 1:27

3 Answers 3

19
\$\begingroup\$

It is worth considering what an axial fan actually does to the air.

If you've ever put your hand in front of a fan (the side that blows air towards you/your hand), you feel a very noticeable and strong flow of air.

On the other hand, place your hand behind the fan - the side that is sucking air in - and you will feel much less air flow.

Of course, the fan is sucking in as much are as it is blowing out, so why is there such a noticeable difference?

Fans work by taking slow moving air and accelerating it to high speed air. This results in a relatively tight column of faster moving air being blown in a direction (along the axis of the fan blade's rotation), while slow moving air is sucked in from pretty much all directions in a just under 180° arc, horizontal and vertical from behind the fan.

If your cabinet is poorly sealed and dust is a primary concern, then you probably want to have the fan at the bottom pushing air into the cabinet.

However, if your cabinet is fairly airtight except for the intake and exhaust vents, then you will get superior thermals for anything passively cooled inside by putting an exhaust fan at the top instead.

The hot air will naturally move to the top of the cabinet and without much preference for where, and an exhaust fan at the top will suck in all the hottest air from all directions at the top, exhausting it out in a column of air up and away. This results in much more even removal of hot air, keeping the ambient temperature throughout the entire cabinet lower.

If you have the fan blow cool air in from the bottom instead, you will have much greater turbulent effects from the higher and more directional air and depending on the location and shape of whatever is in the cabinet, this can have a quite measurable detrimental effect on thermals over all.

If you are blowing air out, you can blow it out into a clear open space, while the air being sucked in by that fan is all going to come in from the easiest (lowest impedance) paths, whatever they may be.

If you are blowing that directional column of fast moving air into the cabinet from the bottom, some of it is going to impinge on surfaces inside the cabinet (or the cabinet walls themselves) and generally not flow as well out of the case. Additionally, hot air will not be removed evenly, and the part of the cabinet the fan is blowing directly at might be cooler than the average temperature you'd get with an exhaust fan at the top, you will have the potential for various pockets of hot air that are not really replaced with new, cool air as efficiently.

But the overall average temperature will be lower with an exhaust fan at the top. Simply because you get more air flow that way. Static pressure is static pressure, the fan only cares about the pressure difference across it. Exhausting means lower (negative) pressure on the intake/inside the cabinet, and higher pressure at the exhaust. Intaking air means .. lower (negative, relative to the cabinet insides) pressure at the intake and higher pressure on the fan's exhaust side, only now it is exhausting into the cabinet. There is no difference from the fan's perspective. Either way, there is a static pressure gradient in the same direction. It makes no difference.

BUT...consider why hot air rises. It is because it is less dense. What does that mean for a gas? Higher pressure. Just look at hot air balloons - the whole balloon is inflated and given structure simply by heating the air inside, producing positive pressure.

So intaking cold air from the bottom is going to create positive pressure inside the cabinet... but there is already some positive pressure in the cabinet simply from the air being hotter than outside the cabinet. This increases the static pressure on the fan must overcome, which reduces the air flow a given fan can achieve.

If, however, you have a fan at the top exhausting air, you have slightly less static pressure across it. You have the positive pressure from the hot air inside the cabinet trying to push out, and a fan that is trying to exhaust it. Where the intake fan is acting against this pressure, the exhaust fan is aided by it. The result is a measurable improvement in air flow vs. the intake fan. Just think of it like this: is it easier to inflate or deflate a balloon?

The improvement can be significant or negligible (more often the latter, or at least a fairly modest improvement) depending on all sorts of factors. But it will be there and be improvement all the same. If you are confident that little to no unfiltered air will get sucked into the cabinet, you should go with an exhaust fan blowing out the top. If you have lots of cracks or seems air can get into though and are worried about dust, then go for an intake fan at the bottom for the positive pressure. Just be wary of hot spots etc.

\$\endgroup\$
3
  • 5
    \$\begingroup\$ Sometimes forced turbulent airflow inside the cabinet can be better than natural convection. This is true especially if power levels are small and the goal is to keep the temperature differences very small, as natural convection needs a large enough temperature difference to start flowing. \$\endgroup\$
    – jpa
    Oct 13, 2022 at 11:01
  • \$\begingroup\$ Very thorough answer, thanks for that \$\endgroup\$
    – asdfsdgf
    Oct 14, 2022 at 0:40
  • \$\begingroup\$ Lots of great points, but a minor one can still be added (to muddy up the waters a bit): any sort of a grating/grille you might use to keep people from sticking fingers into the fan is going to present more of an obstruction (and create more noise) on the high-velocity, exhaust side of any fan. So if noise or efficiency is an important concern, it's somewhat better to have a fan suck air into the cabinet instead of blowing it out. \$\endgroup\$
    – TooTea
    Oct 14, 2022 at 12:56
20
\$\begingroup\$

You want air to flow up and out. This facilitates natural convection rather than trying to make it flow in reverse. So if the fan is blowing IN to the enclosure put it at the bottom. If the fan is an exhaust fan put it at the top.

If you blow filtered air into the enclosure the whole enclosure will have a very slightly elevated pressure which might help keep dust out. If you blow air out of the enclosure, the whole enclosure will have a slightly lower pressure and it may tend to suck unfiltered air in through every crack and seam and gap.

I am not an expert on cabinet ventilation. But when I have done them I have tried to run them at positive pressure with a filter on the inlet fan.

\$\endgroup\$
7
  • \$\begingroup\$ Worth noting that the vast majority of such fans have double-sided enclosures, precisely so that they can be changed from 'suck' to 'blow' merely by flipping them over. Having a fan at the bottom very, very slightly increases the air pressure inside, which is a good thing, whereas a fan at the top very, very slightly decreases air pressure, which is counter-productive. \$\endgroup\$
    – MikeB
    Oct 13, 2022 at 12:23
  • \$\begingroup\$ Why is positive air pressure desirable? Just to keep out dust? From a cooling perspective, isn't slightly negative better than slightly positive pressure? \$\endgroup\$ Oct 13, 2022 at 17:22
  • \$\begingroup\$ @TylerShellberg Is there a reason for negative pressure having more heat transfer? None come to mind, but I could be missing something. I might even assume the opposite, where higher pressure is more density, and therefore might be able to transfer more heat due to increased mass available to take heat away. \$\endgroup\$
    – JMac
    Oct 13, 2022 at 17:55
  • \$\begingroup\$ I don't know about cooling. If you look at metacollin's answer it says that exhaust on top will work slightly better in terms of airflow than blowing air in at the bottom and venting on top. All else being equal, I would thing that more airflow will lead to better cooling. I assume the density difference is really minimal. My only rationale for positive pressure is dust control. @TylerShellberg \$\endgroup\$
    – user57037
    Oct 13, 2022 at 18:18
  • 2
    \$\begingroup\$ There are very slight differences in cooling efficiency. But the big difference is that positive pressure puts more of the dust on your filter while negative pressure puts more dust on the things you're trying to cool. \$\endgroup\$ Oct 13, 2022 at 21:19
8
\$\begingroup\$

It's application specific. Normally a single fan would be an extract, but you need to consider and ideally test the actual airflow. If air is blowing in from an intake, it will take the easiest route to get out of the case, and if the fan is an extract it will suck air from the lowest available static pressure source(s). If you imagine a fan next to a hole, the air will mostly suck or blow air mostly from the hole, not ventilating the rest of the enclosure. Different fans have different capabilities regarding static pressure, which is their ability to get a long column of air moving, and you need to take that into account. Are there other holes in the case? Can air enter around a door or between panels? Will the air flow around particularly warm components?

Also, you don't want to filter an outlet. You only need filters on inlets.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ If my computer's case fans were extractors instead, like how they used to make computers, then I could put the cover back on it.... Blowing air at something and then hoping it finds its way out of the case is silly. Extracting make it go brrrrr, which is what all the other fans do while the cover's off. - Put the inlet as far off the floor as possible (dust). Don't make any holes in the top (spills). \$\endgroup\$
    – Mazura
    Oct 14, 2022 at 4:42
  • \$\begingroup\$ @Mazura You can just unscrew the fans and turn them around backwards, if that's something you want. The screw holes are symmetrical. \$\endgroup\$
    – user253751
    Oct 14, 2022 at 13:12

Not the answer you're looking for? Browse other questions tagged or ask your own question.