Author Topic: Noisy fans  (Read 498 times)

0 Members and 1 Guest are viewing this topic.

Offline peter-h

  • Regular Contributor
  • *
  • Posts: 158
  • Country: gb
  • Doing electronics since the 1960s...
Noisy fans
« on: September 29, 2019, 02:32:41 pm »
Who do equipment makers put in such noisy cooling fans?

There is no need for it. In the PC world, low noise fans have been around for years. And you can have a temperature controlled fan if you really want the flow rate at high ambient temps.

The end result is that anyone working in electronics is surrounded by noise from every direction. Mostly white noise which really messes up one's hearing.
 
The following users thanked this post: tooki, Electro Detective

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 2594
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Noisy fans
« Reply #1 on: September 29, 2019, 02:37:42 pm »
a) because computers these days do not need much cooling and the fans serve mostly estetical value*
b) because it is cheaper to put in smaller heatsink and put a cheap vacuum-cleaner rated fan near it (y u no they need to shave those 3 cents off from a $3000 instrument)
c) because most test equipment, unlike modern PCs, produce real heat, not just "heat".


*not counting high performance gaming setups, that double as frying pans
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 1010
  • Country: fi
    • My home page and email address
Re: Noisy fans
« Reply #2 on: September 29, 2019, 08:36:42 pm »
My personal observation is that desktop computers have shifted to larger fans, most commonly 120×120×25mm ones.
Good ones are basically silent at 800 RPM, but still move quite a lot of air.  I personally use voltage-controlled fans that run at 1000 - 1200 RPM at 12V, and a controller that uses the built-in tachometer, downvolting the fans to 800-900 RPM; depending a bit on use case (pressure differential needed) and blade geometry.

Equipment tends to use smaller fans, and thus need to do seriously higher RPMs to move enough air.  It is much worse than linear dependence, because the hub size is basically constant; the effective fan disk surface area is much smaller.  At 60×60mm (2.36"×2.36"), you often need 2000-4000 RPM to get significant airflow, and there just isn't any known tech that can do that with relatively little noise: they all whine.  The fans in equipment seem to be chosen based on power rating rather than audible noise generated, since it is not a feature users seem to make a buying decision on; unlike computers.

Laptops are designed to spread the heat out, and only use a tiny fan or two when a relatively high temperature is reached; and then sound like a small turbine engine.  They are even worse than any equipment -- but only when the machine is stressed enough to need that additional cooling.

It sounds counterintuitive, but a bigger fan that consumes more energy than a smaller fan, installed in less optimal place (much worse flow patterns), can be much quieter than a smaller fan that seems optimal for a given enclosure, because of the different RPMs needed to move enough air for the cooling required.  This is just not something many enclosure designers seem to realize, so we get noisy but intuitively/technically straightforward enclosures.
 
The following users thanked this post: tooki

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 17642
  • Country: nl
    • NCT Developments
Re: Noisy fans
« Reply #3 on: September 29, 2019, 08:58:06 pm »
Who do equipment makers put in such noisy cooling fans?

There is no need for it. In the PC world, low noise fans have been around for years. And you can have a temperature controlled fan if you really want the flow rate at high ambient temps.
But a PC doesn't have to work at full power at 40 to 50 degrees C for 20 years 24/7. Professional test equipment is designed to do just that. However some equipment really needs the cooling and some doesn't. I tried to put lower noise fans in my newly acquired Lecroy 7200A scope and this didn't work out very well (some parts got nearly 10 deg. C hotter). OTOH I have put temperature controlled fans in my HP6012A power supplies and I never managed to get them spin faster than the standard RPM.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 9217
  • Country: us
Re: Noisy fans
« Reply #4 on: September 29, 2019, 09:56:57 pm »
I think it mostly comes down to the fact that the equipment is not intended to be used in homes or quiet offices, so quiet operation is not typically something that is marketed or one of the design criteria that gets a lot of emphasis. What *is* emphasized is performance and reliability, so if a more powerful fan results in potentially longer lived equipment or improved thermal stability with corresponding improved electrical stability then it makes sense to go with the more powerful fan. With a given piece of equipment, ask yourself if they would have sold many more of them had they put more effort into making it quieter. If the answer is no, then there's your reason.
 

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2096
  • Country: au
Re: Noisy fans
« Reply #5 on: September 29, 2019, 11:07:57 pm »
I've got some gear with 'Fan On Demand' feature and whilst they make a racket when they kick in under actual 'DEMAND' conditions,
they shut up totally when the demand is gone and or the internals have cooled down.  :clap:

I'm with OP peter-h on this, it's not rocket science or 'profit devouring dollars' to install 'fan on demand' features,
with multiple quiet fans kicking in as singles or sets if and when 'demand' requires it,

not needlessly running 24/7 pulling in dust, wasting power, praying all that noisy air may be sort of finding it's way to the actual hot spots that need it (usually NOT  :palm:)
and most important > driving the user nuts!   |O
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1830
  • Country: us
Re: Noisy fans
« Reply #6 on: September 30, 2019, 04:02:30 pm »
I remember the old DEC PDP-11 computers. They had three fans stuck in there because the board spacing was so tight. That was noisy.  Even then the room had to be air conditioned or they would create errors.  If one fan went out or it got dusty they would fail. 50% more cabinet space would made them much cheaper to run.
 

Offline peter-h

  • Regular Contributor
  • *
  • Posts: 158
  • Country: gb
  • Doing electronics since the 1960s...
Re: Noisy fans
« Reply #7 on: September 30, 2019, 07:48:58 pm »
Totally fanless is quite hard to do. I have some fanless machines. Lots of heat pipes and it works with a low power CPU.

I have had some fanless power supplies and all blew up really quickly despite being hugely over-specced for the job. Too many hot components... whereas a motherboard is easier, with heat pipes needed for the CPU and the main chip.

Howeve, the point I was trying to make is that in most applications you are not extracting kilowatts by forced airflow. You just need moving air... a low noise fan is just fine.
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 1010
  • Country: fi
    • My home page and email address
Re: Noisy fans
« Reply #8 on: September 30, 2019, 09:32:10 pm »
To clarify my own position, in practical terms:

If you have a 140mm × 140mm × 45mm enclosure, instead of putting one, two, or three 40mm×40mm fans at the smallest face, use a slightly thicker case (45mm to 70mm), and a low-RPM 120mm×120mm×25mm quiet fan.  If there is no clear path, and the innards are full of wires and heatsinks, pick a quiet fan that has slightly lower airflow, but higher static pressure.  Such a fan might even consume more power at 800 RPM than the small fans at 3000 RPM, but it will be much quieter when providing the same cooling/total airflow.

Too often I see the fan being placed on the smallest face, even when there is room for a much bigger fan on some other face.

FWIW, using rubber grommets or even silicone gasket (from a tube) works well to avoid chassis resonation.  Soft silicone grommets work best IMO.
 

Offline Red Squirrel

  • Super Contributor
  • ***
  • Posts: 2254
  • Country: ca
Re: Noisy fans
« Reply #9 on: September 30, 2019, 11:32:54 pm »
I think one issue too is lot of products will just put the fan in as an after thought and it's the cheapest one they can find.  Also it seems 99% of the time it runs for nothing. I have some 1U switches with super loud fans that probably don't even need to run that much.  Why can't they thermal throttle it at least.  The inverter in my shed also has a fan and the fan runs based on load instead of temp which is a stupid design. There is no reason that fan even needs to run in winter but it does and it ends up draining the battery.  Will probably modify it some time and same with the switches.  It's not a big enough issue, more of an OCD thing. :P
« Last Edit: September 30, 2019, 11:34:30 pm by Red Squirrel »
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 9217
  • Country: us
Re: Noisy fans
« Reply #10 on: October 02, 2019, 05:12:21 am »
I have a cheap NVR for my security cameras and it came with a noisy little fan in it. The way the cover is designed it didn't seem to move much air so I unplugged the fan completely, I let it run for a while and nothing seemed to be getting worryingly loud. It's been running for 2 years now with no fan and no troubles.

I had a very similar experience with some web controlled power strips, the old models had fans that would go noisy, I noticed the newer more compact models had almost identical guts but no fan at all. I removed the noisy fans from a bunch of them and they didn't even get warm to the touch. Some stuff has a fan just because.
 

Offline VK3DRB

  • Super Contributor
  • ***
  • Posts: 1487
  • Country: au
Re: Noisy fans
« Reply #11 on: October 02, 2019, 08:33:15 am »
What is really annoying is the idiots at companies like NVIDIA (graphics cards) and Pioneer (audio amps) designing in crappy phosphor-bronze bearing fans. When the lubricant dries out, they rattle making a buzzing sound, run slow and inefficient and sometimes they cannot be replaced easily. For the sake of a couple of dollars, they could have put in far better fans. Instead, they lose customers who look elsewhere for better quality.

Quality is remembered long after the price is forgotten.
 

Offline SteveyG

  • Supporter
  • ****
  • Posts: 730
  • Country: gb
Re: Noisy fans
« Reply #12 on: October 02, 2019, 08:39:08 am »
To clarify my own position, in practical terms:

If you have a 140mm × 140mm × 45mm enclosure, instead of putting one, two, or three 40mm×40mm fans at the smallest face, use a slightly thicker case (45mm to 70mm), and a low-RPM 120mm×120mm×25mm quiet fan.

Generally, equipment is designed to be stacked, so it is rare the fan could go on the largest face. There are also the IP rating issues associated with vents which constrains which faces you can place the fan on.

a) because computers these days do not need much cooling and the fans serve mostly estetical value*

The majority of PC processors in use today will still have a TDP in excess of 50W, which needs a decent amount of airflow. I wouldn't say the fans are for aesthetics.
 
The following users thanked this post: tooki

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 1010
  • Country: fi
    • My home page and email address
Re: Noisy fans
« Reply #13 on: October 02, 2019, 11:34:33 pm »
Generally, equipment is designed to be stacked, so it is rare the fan could go on the largest face.
Even if there is just a few millimeter gap between devices, it is still better airflow path than those screaming fans.

I've worked a lot with 1U rack enclosures (high performance computing cluster compute nodes), and man, do I hate those mini fans they use in them.  They scream.

There are also the IP rating issues associated with vents which constrains which faces you can place the fan on.
I admit, I am completely unaware of those.  (I know basic IPxy ratings, but that's it.)

However, even in a 1U rack mount chassis there is usually enough room for larger centrifugal fans, if the exhaust has to be on a small face of the enclosure.  They are by no means "silent", and even "quiet" ones are hard to come by, but they are not as nearly noisy as the small turbine-like fans (say, 50×50mm or smaller; the 25×25mm are utterly crazy IMO).  (Rack mount servers don't need to be silent, because there is no demand; if you need to go into a server hall, you better have hearing protection anyway.)
 

Offline MrMobodies

  • Frequent Contributor
  • **
  • Posts: 612
  • Country: gb
Re: Noisy fans
« Reply #14 on: October 04, 2019, 12:34:23 am »
I have a collection of old Thermaltake Smart fans 1 and 2 some of which can spin up to 4000 - 6000rpm and they can be very noisy without the variable control of thermal sensor connected but they have been reliable over the years and still work. I have noticed the larger ones are the quieter than the smaller ones (with the same RPM) at certain airflow and the noise from them don't bother me so far. They're not led lit ones but the plastic is orange and some are black.

About ten years ago had some large CPU blowers that were pulled out of some Dell servers. I connected them up to Thermaltake speed controllers to slow them down and I got a decent enough airflow where I want just to cool down a shelf where I put drives for testing and anything that got hot and that seemed much more quieter but it was only a bit of airflow I needed in a certain place and it didn't need screwing down or it didn't move about on the surface.
« Last Edit: October 04, 2019, 12:45:29 am by MrMobodies »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf