NY Times article about fan noise cancellation

[jib]

The NY Times (noless) posted an article today about a professor and his students working on a project trying to reduce fan noise:

Dr. Sommerfeldt's system has four miniature speakers and four even tinier microphones set in a ring around the computer fan. The microphones and other sensors detect the noise of the fan blades and, with the help of digital signal processing and algorithms, radiate opposing tones from the speakers. The whole system can be tucked into the same space that a conventional computer cooling fan would occupy.

Sounds interesting .. if someone searched the forum I bet we could find someone suggesting this a long time back. Who knows maybe he or some of his students are one of us

[Keyoh]

Dang, beat me to it.

[Sledge]

Who knows maybe he or some of his students are one of us

If so, then I'd doubt they'd admit it. The patent for an effective noise cancellation system for a fan would be worth... zillions

[Nebor]

We can meet him to it. We just need to gut a pair of Bose Noise cancelling headphones, use the mic from it, amplify the output from the noise cancelling circuit to slightly larger speakers, and presto!

[Rusty075]

Sounds interesting .. if someone searched the forum I bet we could find someone suggesting this a long time back. Who knows maybe he or some of his students are one of us

In theory it wouldn't be that complicated: Since fan noise is of a constant frequency all you'd need is a DSP plugin to convert the noise 180 degrees out of phase. Your computer already has the hardware it needs; a soundcard with a line in for the mic's and a line out for the playback speakers.


But in reality its alot more complicated. The fan noise comes from too big a source to be damped by 1 speaker. You'd need an array of mic's and speakers working together. The system would require individual tuning based on its particular environment, and that tuning would be much more sensitive than what a hobbiest could do. Even the slightest imperfection in the sound wave interferences and instead of reducing the noise you make it twice as loud....

One way this could be made to work would be to put the mic, speaker, and processing chip all inside the hub of the fan. By being within the source of the noise 1 speaker could cancel it out by itself. Of course this would require some seriously tiny electronics, and would probably result in a fan that costs $500 each.

Anybody know a good patent attorney?

[yermolovd]

Wasn't this discussed here before? Anyway, there have been many discussions of that on other forums, and everyone comes to a conclusion that this is hard to do (too much work I guess). Well, if they get the thing to work, Ill be only happy .

[MikeC]

Someone else is writing an article on the same topic for New Scientist Magazine and did a quick email interview with me on the topic, which might be germaine to post here...

Mike -

Thanks for getting back quickly. The article is slated for the print edition dated 29 May. I'm not sure if we'll put it on-line; only some of the articles go up. By the way -- you have a nice site that helped me understand what's going on. Good work.

-- Best, Jeff

---------------------------------------------------
At 8:54 AM -0700 5/21/04, Silent Mike wrote:

Hi Jeff,

My understanding about noise cancellation is that effectiveness is limited to specific fields, like the perfect "illusion of image" that audio aficionados seek. Outside of the target zone, the image collapses, or the noise cancellation no longer works -- or no longer works as well. This is one reason that noise cancellation headphones make a lot of sense. (They also work very well.)

The elimination of higher frequency tones from fan noise that's demonstrated by the WAV file in http://www.acoustics.org/press/147th/Sommerfeldt.htm is impressive. If this result is not zonally restricted and generally achievable at low enough cost with a drop-in substitute for standard fans, then yes, it is certainly worthwhile. I think there would be demand for both aftermarket quieting mod kits of existing office gear as well as for OEMs. Audible tones are generally much more annoying than the broadband white noise of fan-caused air turbulence to most people. What is the anticipated cost?

Having said that, most makers of office machinery seem to put almost no emphasis on low noise. If it is a design goal, it is certainly secondary to others, such as low cost, convenience, size, etc. Making low noise a primary design target may be all that's needed to bring the noise of office gear down.

Both commercial and amateur practitioners of the art of PC silencing at www.silentpcreview.com have achieved great noise reduction in PCs with careful implementation of existing technologies, often bringing the overall PC noise down to ~20 dBA/1m. This is far below the ambient level of any office, and effectively silent except in an incredibly quiet room. All with nothing more than careful design and modification of existing technologies and the judicious use of reduced speed higher quality fans, directed airflow, unobstructed airflow paths, and more efficient heatsink designs.

An old adage about noise is that it is much more effective to eliminate the source rather than try to block or damp it after it is produced. Which category does noise cancellation fit into?

Please contact me if you have more questions/comments, and do send me a link to your article when it is published.

Mike Chin
editor/publisher
www.silentpcreview.com

Something I meant to bring up and forgot about is the issue of whether noise cancellation ADDS to the total acoustic energy in a space, and even if we hear nothing because of the cancellation, what effects might all that unheard energy be having on us? It seems relevant because with noise cancellation, you are actually adding more energy to the environment, and we know that all forms of radiated energy have their effects on living beings. If we consider the number of IT cooling fans in a large office... now double that!

[greeef]

i can answer that succintly - it doesnt add noise.

Noise cancellatin is very easy to explain. Take a sine wave, you have a noise. Now take an identical sine wave, flip it 180 degrees, and run both of them through the same speaker. They cancel each other out, the total sound output is nil.

Where this destructive reinforcement occurs acoustic energy is reduced. However, what you say about target zones is very relevant - with higher frequency sounds, the distance between the two sources (noisy device and noise canceling speaker) must be minimal, otherwise approaching the device from different angle would delay the waveforms enough that positive reinforcement could occur.

My other worry would be that these devices aren't 100% accurate, and they could end up simply masking some of the unwanted noise with louder but less objectionable tones.

griff

[MikeC]

greeef --

I understand how noise cancellation works...

So you have two identical signals out of phase.... does that mean you aren't doubling the amount of energy fed into this acoustic space? I don't believe this is a question or issue I've seen serious studies of anywhere. It is especially relevant when you have say 200 fans and 200 noise-cancellation fans -- surely there will be a GREAT deal of acoustic energy outside the null zones. And even in the null zones.... I wonder if perception or non-perception is all there is to it. IE, there's no free lunch... above and beyond the energy it takes to drive the noise-cancellation device.

[AZBrandon]

If so, then I'd doubt they'd admit it. The patent for an effective noise cancellation system for a fan would be worth... zillions

Well as others have mentioned, the technology has been around for a while. Honda even used it in the 1997-2001 Honda Prelude stereo system to try to reduce the apparent sound of road noise in the car. Did it work? Well.. you don't hear about it being used on any other cars, that's for sure. The only other place I know of it being particularly common is private aircraft headsets. A Cessna 172/182 is a rather unpleasant place to be, acoustically speaking, and noise-canceling headsets go a long way to reducing the apparent noise.

Ultimately, I suspect it will simply be too expensive to be practical for home PC's. It might be useful in industrial environments or something where they have massive fans cooling warehouses and stuff, but I think for PC's it will likely never catch on, especially as manufacturers get better at reducing the total heat output of PC's.

[trodas]

Well, not only expensive, but also far from being efective - especially when multiple fans are considred
Look, some of the fan noise "wave" get spread out of the fan even with all this "oposite waves" (it's for example just matter of "not enought speakers and mics" to cover all the noise from one fan as well as subject of "not enought fast system reaction time" - as the system catching the wave that must exist in the first place and then the second one it MIGHT catch with "antiwave") and therefore there will be a LOT's of intereferences all around...

So my suggestion is - let's stick to realy BIG PASSIVE heatsinks

[fancontrol]

Assuming engineers continue to surprise us and find a way to make the cost of a noise cancellation system resonable, the big hurdle to overcome is interference patterns. Unless you can actively cancel noise right at the source (fan, aircraft engine, tyre) or the sensor (ears) you have the above mentioned problem of keeping the signal and it's opposite in sync.

I read an Intel white paper where they did their best to use active noise cancellation to keep a PC quiet. I took it as an honsest effort to ensure that working on making components quieter was a better idea than adding a component to cancel noise. Hunt around; I'm sure you can find it. The bottom line was that the best they could do was... not very good. And the chassis grew significantly to make room for all the noise cancelling gear.

My vote: put the moving parts in the closet. Networks are fast enough to make this real. It just isn't main stream. Yet. I don't think it'll be long before my DVR, cable box, PC, answering machine, DVD changer, etc. are all forgotten like the water heater. I don't see it, I don't hear it, I don't think about it, but I can get what I want from it anywhere in the house.

[Inexplicable]

So you have two identical signals out of phase.... does that mean you aren't doubling the amount of energy fed into this acoustic space?

If I understand correctly, noise cancellation is basically about breaking up pressure waves (synchronized gas molecule movements). I would suggest that a lot of the energy in fact turns into heat (the background noise of molecules randomly bumping into each other). Noise gets absorbed this way in any case, noise cancellation just helps it along. It can't be perfect, of course, but I wouldn't think a great deal of energy magically gets transferred outside the null zone, although you could get some pretty strange sum waves there.

[mynci]

hope this doesnt annoy the guys at new scientist. ill take it down after a few days.

i am a subscriber and due to a heavy work load hadnt had chance to even open this weeks edition else i would have scaned it when i got it (thursday).

still here it is:



no mention of spcr though

mynci