Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Closet cooling and fan static pressure

kaspor

Electrical
Aug 12, 2021
52
Hello

I have a closet which is approx 2.7m high by 1m deep by 1.2m wide. I have l equipment in there that needs to sit at max 30degC. There are two double doors on the closet.

In order to keep it at 30degC I’ve installed 4 extraction fans on a thermostat. The fans draw air in through gaps in the doors and from underneath and exhaust it at the top. The fans are similar to these


The problem is the noise. They’re generation 30dba each. I’ve insulated them as much as I can but they’re still loud.

So I then started looking at quieter fans such as these


They have a much lower db but looking at the performance curve it looks to have a much lower static pressure rating


Does this mean they will inherently take longer to get the temp in the closet down as they are not able to draw the same amount of air in?

I’m worried that if I change the fans over that the newer fans will be useless as they won’t be able to draw air in like my current ones.

Also cutting holes or vents for intake isn’t an option unfortunately.
 
Replies continue below

Recommended for you

This is all very odd.

You have something generating heat inside your wardrobe?

Which is in your bedroom??

Four fans that size for something that small is absolutely nuts.

You need to provide a very low resistance to air flow coming in. At the moment I doubt those fans are drawing in more than 10% of their maximum flow and possibly less.

Or find some other way of cooking whatever it is you have hiding in your closet.

What heat flow are we talking here or ach in the closer??
 
Thanks for the reply. It’s all of my server equipment. It’s a closet in the home, not in my bedroom.

Because of the enclosed space heat eventually dissipates within the closet and it gets to 6 degrees above ambient. Ambient temp can get up to 28deg.

I would love to have a bigger area for airflow coming in but that would mean cutting a hole in the door or a hole in the wall which I want to try and avoid.

The fans work well but they’re noisy.

I think the Noctua fans which are renowned to be very quiet would not work well because they don’t have a high enough static pressure rating?
 
The simple answer is to go buy a replacement door and cut whatever hole you want in that and save the original door to put back when needed.

If one is particularly cost sensitive, a couple of sheets of plywood might do it, but I haven't priced things like hollow core doors vs. plywood sheet and that might vary with location.

This would also allow whatever fancy upgrades, such as access windows or blinking lights to be added to the door that may or may not be driven by the server.
 
Not understanding the flow path for air into the closet through the fan and back out of the closet. If you just have some fans blowing in a closet, there is a good likelihood that you are heating the air in the closet with the dissipated fan energy and air movement.
 
The fans are possibly noisy because they are running at a low flow. One of those fans should be overkill for such a ventilation duty, but my guess is a lot of the air just leaks in high up and doesn't flow the air past your server.

Can you not remove the plinth or cut agroove on it and drill some large holes in the floor of the closet?

You need to direct whee the air inlet is going and allow for a much lower pressure drop. These fans have very low differential pressure. The quiet fans are much lower power and more like 12V computer cooling fans.

If the ambient air is 28 C you won't get enough air in there to keep it below 30C.

But why is it in a closet to start with?
 
If you have a fairly restricted flow path for the air you need to use a blower with higher static pressure capability than an axial fan. That would be a squirrel cage type of blower. These can be very quiet, or not.
How are your fans exhausting air at the top without cutting holes in the doors. Are they louvered doors? If they are louvered you may not need fans. The fans may be fighting natural convection though the lovers.
 
Doors are typically undercut 1" to 2" to allow flow from rooms with door closed back to the return air louver in HVAC systems or to allow air to flow into restrooms with exhaust fans. Your case with flow into a closet is similar.

Undercutting consists of cutting the bottom of the door across the entire bottom surface by 1 to 2 inches. Likely you already have doors such in your bathroom and other locations with undercut but you never notice it, that you can use as examples of how much to cut off. MY house has many doors that are undercut.
 
Thanks for the replies! I have attached a sketch which hopefully paints a better picture of the install.

1736716412336.png
The simple answer is to go buy a replacement door and cut whatever hole you want in that and save the original door to put back when needed.

If one is particularly cost sensitive, a couple of sheets of plywood might do it, but I haven't priced things like hollow core doors vs. plywood sheet and that might vary with location.

This would also allow whatever fancy upgrades, such as access windows or blinking lights to be added to the door that may or may not be driven by the server.
Unfortunately this is not an option. My wife will shoot me :(

Not understanding the flow path for air into the closet through the fan and back out of the closet. If you just have some fans blowing in a closet, there is a good likelihood that you are heating the air in the closet with the dissipated fan energy and air movement.
I have attached a diagram to this post which hopefully clears this up. Cooler air is being drawn through the gaps in the doors. Warm air is drawn out the top of the closet.

The fans are possibly noisy because they are running at a low flow. One of those fans should be overkill for such a ventilation duty, but my guess is a lot of the air just leaks in high up and doesn't flow the air past your server.

Can you not remove the plinth or cut agroove on it and drill some large holes in the floor of the closet?

You need to direct whee the air inlet is going and allow for a much lower pressure drop. These fans have very low differential pressure. The quiet fans are much lower power and more like 12V computer cooling fans.

If the ambient air is 28 C you won't get enough air in there to keep it below 30C.

But why is it in a closet to start with?

I don't think they're noisy because of low flow. Even if I open the doors, the noise is the same. With the doors open there would be very little flow restriction. The noise is from the motors. If the ambient is 28 and I run these fans, the temp sits at 30.

If you have a fairly restricted flow path for the air you need to use a blower with higher static pressure capability than an axial fan. That would be a squirrel cage type of blower. These can be very quiet, or not.
How are your fans exhausting air at the top without cutting holes in the doors. Are they louvered doors? If they are louvered you may not need fans. The fans may be fighting natural convection though the lovers.
See attached sketch

Doors are typically undercut 1" to 2" to allow flow from rooms with door closed back to the return air louver in HVAC systems or to allow air to flow into restrooms with exhaust fans. Your case with flow into a closet is similar.

Undercutting consists of cutting the bottom of the door across the entire bottom surface by 1 to 2 inches. Likely you already have doors such in your bathroom and other locations with undercut but you never notice it, that you can use as examples of how much to cut off. MY house has many doors that are undercut.

Yes they are undercut, see attached sketch.

When I look at all the fan specs, it seems that the quieter fans are quieter mainly because they have a lower static pressure at the same flow rate. Granted there's some noise reduction due to the bearings they're using or maybe a better designed motor, but what I can see is that a lower DB always results in lower static pressure on the datasheets. In this situation, lower static pressure won't generate the air changes?

Is it worth trying to reduce the speed of the motors? These are induction motors so they can safely run at 50%. I can put a fan controller on them to see if that helps?
 
You might consider using PC cooling fans, which are typically very quiet; I saw a YouTube video where someone built an indoor air filter using an array of PC cooling fans. and MERV13 filters. Having some filters on the air inlet side might not be a bad idea, since the cooling capability of most electronic equipment degrades with the added load of dust sucked into the filters of the equipment fans.

Nevertheless, I would think you might want to plumb the exhaust air to an outlet on the roof, unless you really want the extra heat in the house. On hot days, that's just an extra burden in the house A/C.
 
You might consider using PC cooling fans, which are typically very quiet; I saw a YouTube video where someone built an indoor air filter using an array of PC cooling fans. and MERV13 filters. Having some filters on the air inlet side might not be a bad idea, since the cooling capability of most electronic equipment degrades with the added load of dust sucked into the filters of the equipment fans.

Nevertheless, I would think you might want to plumb the exhaust air to an outlet on the roof, unless you really want the extra heat in the house. On hot days, that's just an extra burden in the house A/C.

Thanks for the reply. I don't think PC cooling fans will have enough power to draw air in. Flow vs Static pressure curves for PC cooling fans is pretty poor from what I have read? Exhausting to the roof and hence atmosphere isn't a good idea where I live. The cabinet exhausts to a room that is 40m2, so heat build up isn't an issue. Also dust isn't an issue, the area is very clean.
 
Last edited:
What about dust in the intake air ? Fans in a forced draft setup, with intake air filters, would be better to deal with both dust and heat.
 
What about dust in the intake air ? Fans in a forced draft setup, with intake air filters, would be better to deal with both dust and heat.

I inspect the fans once every 6 months for dust. I can tell you there's no dust. Filters would just add unnecessary airflow resistance in this situation I feel.
 
1) I am confused about the noise issue - combining (4) 30 dB fans doesn't really add up to a very loud emitter.
2) How are you insulating for sound? What material are you using?
3) Do you have HVAC ducts in the attic space? Can you tap off a small supply line into the closet ceiling?
4) Without knowledge of the actual arrangement it is hard to guess what the problem is.


db.PNG
 
1) I am confused about the noise issue - combining (4) 30 dB fans doesn't really add up to a very loud emitter.
2) How are you insulating for sound? What material are you using?
3) Do you have HVAC ducts in the attic space? Can you tap off a small supply line into the closet ceiling?
4) Without knowledge of the actual arrangement it is hard to guess what the problem is.


View attachment 3441

Thanks for the reply - I understand it's 36dB when added up, but looking to get it lower if possible.

I have used soundscreen insulation batts (or insulwool) to insulate around the fans.

No there is no attic space, it is a flat roof home. No access to roof space otherwise I would've done this.

The problem is the noise and if there's anything I can do to reduce the noise (even if its a few dB). The fans mentioned in the OP were one option but not sure if they'll do the trick given low static pressure ratings and my installation arrangement.

The only other thing I can think of is to reduce the speed of the fans with a capacitive fan controller.
 
What happens when you run only one fan? You would get a 6dB sound reduction. How did you arrive at four fans? If they are on a thermostat, how much time are they on? Also, what is the wattage of your power sources in the server(s)? Do you have any idea of how much replacement air you need to bring the temperature down? If you could snap a pic of the fan mounting and arrangement it would probably be helpful.
 
What happens when you run only one fan? You would get a 6dB sound reduction. How did you arrive at four fans? If they are on a thermostat, how much time are they on? Also, what is the wattage of your power sources in the server(s)? Do you have any idea of how much replacement air you need to bring the temperature down? If you could snap a pic of the fan mounting and arrangement it would probably be helpful.

I don't have a dB meter, but yes you're probably right. They turn on for 10-15 mins then turn off, once every couple of hours. It depends on ambient temperature etc.

I found that with one fan, it would be on for an hour to achieve the same internal ambient.

I haven't done a heat load calculation for what is inside the cabinet, that would be quite the exercise as I have about 30 pieces of equipment in there. I can you however that temperature increases inside the cabinet at 1deg per hour up to 6-7deg above ambient.

The fans are just mounted to the dry wall blowing straight into the next room. I'll try and get a picture later.
 
The fans don't get the dust because the equipment on the shelves are filling with it.

The detail about the wife was critical information, left out. I thought you might be a renter with a landlord as I don't see any other reason not to make the change. Why does she care?
 
The present fans noise level is 34 dB per data sheet. For 4 fans the combined noise level will be 40 db which is the point where the charts available on the internet show the noise level at the beginning of when ones sleep will be disturbed.

With the 1" undercut of the doors (and not considering the added flow due to the door vertical crack), I calculate that the point of operation of the existing fans based on the intersection of the fan curve with the system head loss curve (only loss is assumed through the 1" undercut) per fan is about 80 m3/hr at 8 Pa, or 320 m3/hr total (188 CFM).

The best choice out of the new fans indicated above that gives the highest static pressure output at given flow for lowest dB noise is the NF-A12x25. I calculate that the operation point based on the intersection of the fan curve with the system head loss curve (only loss is assumed through the 1" undercut) per fan is about 70 m3/hr at about 0.6 mmH2O, or 280 m3/hr total (165 CFM). So there will only be 23 CFM difference from the present fans. The sound level per fan is 22.6 dB or 28.6 dB for four fans operating together. Charts for dB on the internet shows that this is approximately the sound in a bedroom at night. Note a 10 dB reduction in noise level gives a noise intensity 10x less (Watt/m2).

Note also that the new fans you selected are basically PC fans and are meant to be mounted to PC's or similar devices. I assume they can be rigged to mount on your cabinet/gyp board wall.
 
Last edited:
The present fans noise level is 34 dB per data sheet. For 4 fans the combined noise level will be 40 db which is the point where the charts available on the internet show the noise level at the beginning of when ones sleep will be disturbed.

With the 1" undercut of the doors (and not considering the added flow due to the door vertical crack), I calculate that the point of operation of the existing fans based on the intersection of the fan curve with the system head loss curve (only loss is assumed through the 1" undercut) per fan is about 80 m3/hr at 8 Pa, or 320 m3/hr total (188 CFM).

The best choice out of the new fans indicated above that gives the highest static pressure output at given flow for lowest dB noise is the NF-A12x25. I calculate that the operation point based on the intersection of the fan curve with the system head loss curve (only loss is assumed through the 1" undercut) per fan is about 70 m3/hr at about 0.6 mmH2O, or 280 m3/hr total (165 CFM). So there will only be 23 CFM difference from the present fans. The sound level per fan is 22.6 dB or 28.6 dB for four fans operating together. Charts for dB on the internet shows that this is approximately the sound in a bedroom at night. Note a 10 dB reduction in noise level gives a noise intensity 10x less (Watt/m2).

Note also that the new fans you selected are basically PC fans and are meant to be mounted to PC's or similar devices. I assume they can be rigged to mount on your cabinet/gyp board wall.

Thanks for the reply, this is great information. What if instead of going the nuclear option of replacing the fans (quite the job!), I slowed the existing ones down by say 30 or 50% at night? I know this will make a difference to the motor noise, but is there a correlation between m3/h and fan speed?
 

Part and Inventory Search

Sponsor