Small, quiet fan to replace a loud one?

20 Nov 16 12:21, you wrote to Rodney Pont:

RW> Yup, by far the easiest way to reduce fan noise is to increase RW> diameter.

another way is to find fans with large blades that move more air than many small thin blades... house fans are an example... quiet ones generally have only four wide blades whereas others with five, six or more blades that are all thin generally whine like jet engines...

then there's the bearings... many cheap fans use simple nylon sleeve bearings... better fans use ball bearings... the best ones use good sealed roller bearings... which bearings one wants depends on the orientation of the fan... roller bearings won't do much for a fan that is oriented to blow up or down but mount it on its side and those bearings do their job very well...

)\/(ark

Always Mount a Scratch Monkey Do you manage your own servers? If you are not running an IDS/IPS yer doin' it wrong... ... That's the way the veal cutlets.

Reply to
mark lewis
Loading thread data ...

So I recently bought

- a case that is adequate, but the included fan has turned out to be irritatingly loud.

Anyone know of a decent 3cm fan I can replace this one with, that is either much quieter, or it would also be fine if it ran slower (to be quieter). BOINC (which I run when not using the Pi as a RetroPie box) heats my Pi 3 up to the point where I need some cooling to prevent the CPU from slowing. But with the current fan the CPU temperature only

TIA, Jamie Kahn Genet

--
If you're not part of the solution, you're part of the precipitate.
Reply to
Jamie Kahn Genet

Two suggestions:

1) Use a 5w wire wound resistor in series with the fan to reduce its speed. Assuming the fan's quoted current draw is correct, a 4.7 ohm resistor should reduce its supply voltage to a bit over 4v, giving a useful speed and noise reduction. Resistors are cheap, so you can afford buy a few to experiment with or your local Maplins should have a 50 or 100 ohm variable pot that should last long enough to let you experiment with it to determine what size fixed resistor is best for the job.

2) Use the chimney effect to cool the Pi.

Try removing the fan and replacing it with a vertical cardboard or plastic tube of at least the diameter of the fan exhaust hole and 30-50cm in length, mounted by either fitting it in the hole or using tape to seal it to the top plate.

This is guaranteed to be a silent cooling system regardless of whether it provides enough cooling. If it works and you care about aesthetics, you can buy clear acryllic tube on eBay.

--
martin@   | Martin Gregorie 
gregorie. | Essex, UK 
org       |
Reply to
Martin Gregorie

You might have better luck looking for a 40mm fan, look for a lower dB value, and mount it on the outside of the case.

Is there a good exit for the airflow? if that's restricted the fan will be a lot louder than it should be and the cpu temperature will be higher. Can you leave an end panel off and see what that's like?

--
Faster, cheaper, quieter than HS2 
and built in 5 years; 
UKUltraspeed
Reply to
Rodney Pont

Yup, by far the easiest way to reduce fan noise is to increase diameter.

(Maybe combine with a chimney, as suggested above.)

Reply to
Roger Bell_West

Now there's a smart idea :-) I've not thought about the stack effect since high school physics (where I mainly thought about the girls in my class, instead of the work)!

*stumbles about mentally with the flow rate calculations for a couple minutes* Hmmm... now I need a point of comparison. What kind of flow rate does a typical small computer fan aim for? Or what would normally be sufficient to keep a Pi 3 from lowering it's CPU speed even in summer? I'll have to find out...
--
If you're not part of the solution, you're part of the precipitate.
Reply to
Jamie Kahn Genet

That's certainly doable. Any quality fan suggestions? I'm not bothered where it's shipped from (given it'll be a cheap purchase). I can even use a remailing service or a US or UK friend to get it to me in NZ, if need be.

Yeah, plenty of space, and the fan is mounted in a flip top for the case. With the lid open the fan still makes as much noise. It must've been very cheaply put together.

--
If you're not part of the solution, you're part of the precipitate.
Reply to
Jamie Kahn Genet

I have one of these cases and got round the problem by switching the fan on and off via a GPIO signal and a transistor. OK it's still noisy when running but mostly doesn't run at all.

--
nev 
getting the wrong stick end since 1953
Reply to
nev young

I like Noctua fans and since you need a 5v one the one below seems suitable and it comes with a low noise kit (a wire with a resistor in it) but you will have to do something to wire it up I think.

formatting link

--
Faster, cheaper, quieter than HS2 
and built in 5 years; 
UKUltraspeed
Reply to
Rodney Pont

Don't screw the body of the fan direct to the body of the case. Use something like these vibration mounts.

New 20pcs Black Soft Rubber Cooling PC Fan Screw Mount Pin Rivet Anti

formatting link

They will however add about a couple of mm fan stand-off from the case body.

--
Adrian C
Reply to
Adrian Caspersz

Why not use PWM to control the fan speed as temperature changes?

Reply to
Rob Morley

I'm not sure I would believe anything in the spec for that fan...

"Sound Level - 18dBA"

18 dBA is nearly silent. I would have to put my ear next to the fan to hear it. If they are claiming this and the fan is irritatingly loud, this is either not the same fan or they just made up the specs.

Fan noise generally related to fan speed. "Because there is a 5th power relation between noise level and fan speed, a small change in fan speed will cause a large change in fan noise." So slowing it by not too much will reduce the noise... assuming the noise is due to the fan itself and not a case issue.

If there are components close to the fan blades, it will create a tone at a frequency of the rotational speed times the number of blades. Is the noise a tonal sound or mostly just white noise? You might try mounting the fan on the outside of the case.

If you don't want to reduce the airflow much, you can try adding a baffle to the fan exhaust. I'm not sure this would help tremendously as some sound will also pass through the inside space and out through the case and it will be at least as large as the case itself, but can help without changing the airflow.

Just one question... do you know you *need* a fan?

--

Rick C
Reply to
rickman

If a chimney will cool this unit, I would bet nothing at all would also cool it adequately. I used a 100 watt lightbulb in a 4 inch x 3 foot dryer tube to see how much chimney effect I would get and was unable to measure any air movement at all. How much air flow would you expect from a 2 watt CPU?

--

Rick C
Reply to
rickman

I've been using vent holes in the top, over the CPU and holes in the side. My RPi2B used to shutdown from over-heating, but not now.

Reply to
4ctestsystems

Il giorno domenica 20 novembre 2016 11:11:38 UTC+1, Jamie Kahn Genet ha scritto:

use a bigger heatsink, and leave the top of the case open.

Bye Jack

Reply to
jack4747

Because putting a resistor in the circuit is easier and is probably adequate to the task. Besides, I don't much like hearing fans whining up and down with load.

--
martin@   | Martin Gregorie 
gregorie. | Essex, UK 
org       |
Reply to
Martin Gregorie

Also use a slightly bigger case, a little more internal volume makes a lot of difference to the cooling.

Plus seemingly ignoring the laws a physics; a white case seems to be cooler than a black case.

---druck

Reply to
druck

It's even better if it's matt black on the inside. In fact the optimum is silver outside, matt black inside, and made of metal.

--
W J G
Reply to
Folderol

I feel there's something fundamentally wrong (some may call it irony!) about using a device designed to generate heat in a system where you're trying to dissipate heat, but maybe it's just me ;-)

-Gordon

Reply to
Gordon Henderson

0.2 amps through 4.7 ohms dissipates 19 mW (IIRC P = I^2*R), so is quite unlikely to cause heat dissipation problems.
--
martin@   | Martin Gregorie 
gregorie. | Essex, UK 
org       |
Reply to
Martin Gregorie

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.