Active Noise Cancellation of Blade-Pass Frequency using Microphones and LoudSpeakers

Active Noise Cancellation of Blade-Pass Frequency using Microphones and LoudSpeakers

These Active Noise Control Papers where Loudspeakers have been used to cancel out blade passing frequency tones are all over the place.

do a google with these terms:

Active Noise Control Blade Frequency

Here's a few links:

http://www.mecheng.adelaide.edu.au/anvc/publications/papers/1996/cement-paper.html

http://www.mecheng.adelaide.edu.au/anvc/publications/papers/2002/AC2_194.html

http://www.aercoustics.com/awards/cce1995/cce95.html

~BdK

ok not trying to flame just understand. i have a case with 4 120mm fans on it so if i tried active noise cancellation i would need 4 different frequences exactly out of phase with each fan to make it silent.

Click to expand...

There are more than one frequency tone being emitted by your fans. We can calculate/estimate the primary Blade Pass Frequency tone being produced using

RPM*Blades/60.

Using microphones, we should be able to reproduce all of the sounds in our PC in loudspeakers in real time.

To make these sound reproductions 180 degrees out of phase only entails swapping the positive and negative wires around on the speakers being used. The sound level and location of these speakers and microphones in each environment would require some experimentation to get it right.

This technique is something that's been applied in real world situations I referenced above.

can anyone calculate the odds of me starting my comp and 2 fans being out of phase and the other two also being out of phase? and would this make my comp quiet? (discounting lots of other things in there making noise.)

Click to expand...

That's not what I'm advocating here. I'm not entirely sure as to what you mean when you say

2 fans being out of phase and the other two also being out of phase

Click to expand...

hth

~BdK

This is rather interesting. So we're going to use destructive inteference properties of a soundwave on fans So we're trying to construct nodes and antinodes at the same point on the wave to cancel it out eh....very interesting....

This has definitely been done in the real world as you said, eg noise cancellation headphones for jack hammer operaters.

I'm gonna have to get out my microphone and try this. I think it would be far more accurate to use the microphone to record then flip the wave in an audio editor than to use RPM*Blades/60.

This could be applied to quite a lot of things in the computer environment eg hdd's and cd drives. A script could easily be written to play the respective cancellation wave whenever your hdd or cd is being used. In theory, you could produce a totally silent computer.

LOL maybe you can even make a silent keyboard

I'll go find my microphone now.......

hmm.. i didnt read much of this, but are u saying using a microphone to record to computer and then ur speakers will output frequency to silence? well wouldnt that have delay? i mean fans are constant, kinda, but keyboard?

It might be possible to record a significant enough chunk of constant noise from your computer, then extract one phase, reverse it, then play it constantly, in sync with the noise from your computer. Granted, this probably won't be perfect and you'll get areas where you will experience the phenomenon of beat, that is, the sound might fade in and fade out as the rpms change.

It shouldn't be that hard to come up with a similar way of doing it on the computer as they do in noise cancellation headphones.

In headphones, there is an active microphone that constantly records the sound and then the onboard chip produces the reverse phase wave in realtime and plays it into the headphones so in essence, you hear nothing from the outside. These headphones are quite common among pilots and I've once tried a pair on onboard a plane and they work freakishly well. I heard nothing from the plane's engines, just the music that was playing from my cd player which I had plugged the headphones into.

So theoretically, there must be some sort of software alternative that can do this in realtime on our pc's, which are so much more powerful than a chip inside the headphone. If such software doesnt exist, someone should make it....

I just realised something. This probably won't work well with speakers as the placement of the speakers will only allow noise cancellation to occur at ONE position in that room as the wave is only going to cancel at that spot, so if you move yourself, you'll hear the noise from your computer in all its glory. You're better off to use headphones me thinks...

wow, cool! I had already read about this in Tom Clancy's 'Cardinal of Kremlin' though

But one thing: is this actually possible with a normal microphone and a normal normal set of speakers? And would the speakers have to be positioned with care or something?

I realize it could be in one of the links above, but I just had time to skim them. I'll read them again later.

what if that noise cancelling speaker is placed directly opposite the primary source of noise???

Can a speaker or speakers be strategically placed surrounding (as in a surround sound setup) and cancelling in real time the noise???

And "recording" per se probably isn't necessary. Just using microphones, appropriate amplifier(s), and out of phase speakers should be enough.

And as far as software to do this, I was thinking of the exact same thing last night.

Using the microphone input on a sound card, monitor (versus record) the noise output in real time of your PC.

Using software that's always on to actively control the noise, replay that noise 180 degrees out of phase through a PC's speakers to cancel out that noise.

Surround Sound speaker setups may or may not be a better choise than traditional stereo setups.

As far as using the formula:

RPM*Blades/60.

I shared that formula with the intent being that it's practical purpose is to be used in fan selection, to identify fans with a more desireable lower frequency noise.

In many of the cases I quoted up top, however, in active noise cancellation, the majority of the noise they were trying to cancel happened to be at the Blade Pass Frequency.

But yes, in order to effectively cancel out noise, it has to be the actual noise reproduced out of phase in real time.

~BdK

This came from "the other thread" regarding fan selection based on Blade Pass Frequency:

http://forum.oc-forums.com/vb/showthread.php?s=&postid=1697860#post1697860

That thread took us on a tangent into discussing Active Noise Cancellation of Blade Pass Frequency noise (as well as other noise).

I felt these topics deserved separation, so I started the thread you're reading right now.

Originally posted by KCN ok not trying to flame just understand. i have a case with 4 120mm fans on it so if i tried active noise cancellation i would need 4 different frequences exactly out of phase with each fan to make it silent.

Click to expand...

quote posted in response by CrystalMethod: "Theoretically" , yes. The odds of you achieving such a feat would be astronomical though. Close in this senario won't get you anything. You need to be dead on.

Click to expand...

This isn't the type of scenario that I've been advocating, and "no" I don't even think it's theoretically possible because each fan is a little different, not exactly identical, in placement, construction, etc.

Soundwaves can be exactly (exactly enough for these purposes) reproduced 180 degrees out of phase however.

hth
~BdK

lazerin said:

I just realised something. This probably won't work well with speakers as the placement of the speakers will only allow noise cancellation to occur at ONE position in that room as the wave is only going to cancel at that spot, so if you move yourself, you'll hear the noise from your computer in all its glory. You're better off to use headphones me thinks...

Click to expand...

bingo.

In a few more words, the interference pattern of a sound source and a similar one (out of phase) is very complicated. There would be several regions where they are exactly inverted and thus cancelled out, but most other regions would experience "beats" or even constructive interference (increase in noise). The cancellation is very localized. This is why it works well for headphones -- the mics sample the sound right at your ears and cancel the sound at those locations.

Also, notice that while your fan is rated for X rpm's at 12VDC, your computer generally will deliver something off from that, such as 12.4VDC. Combine this with the fact that most parts will also have some error / discrepancy between rated and actual operation (even at precisely 12VDC), and you won't be able to use the fans rated RPMs to calculate the blade pass frequency. (But patching into MBM's shared memory structure might work, were it not for the above comment, though.)

Still, this is always a fun idea to beat around.

-- Paul

you don't need to know BPF to cancel the noise

you don't need to know bpf to cancel the noise. so all the voltage/rpm/etc. fan variance discussion is a moot point when it comes to active noise cancellation using microphones.

the only really useful application of bpf as far as I'm concerned thus far, is in FAN SELECTION based on the frequency of one of the components of that fans noise which can be roughly predicted using the bpf formula B*R/60.

Knowing BPF can help in fan selection.

Using active noise cancellation techniques may be able to help in System Noise Control.

~BdK

You might be able to use speakers......
If you go to the old subwoofer placement technique. That is, instead of placing the subwoofer, place the speakers where your head is usually at. Playback the cancellation wave, then move around the room until you find the spot where you experience cancellation. Then place your speakers there. Bare in mind, you CANNOT move your head an inch or else it'll be out of phase

unless your speakers are REALLY BIG!

-BdK

You'd also want to be very careful with this, cause if the waves become in phase, eg the nodes line up with the nodes and the anti nodes line up with the anti nodes....you'll experience constructive inteference of waves and the sound would be amplified 2x!!!!!!! This type of constructive inteference is strong enough to cause a lot of damage in the real world....eg shattering a cyrstal glass by playing its' natural frequency back to itself.

What about something attached to the PC that vibrates the PC in response to and out of phase from noise created therein???

sort of like an "aura shaker" used in car stereos in lieu of speakers to create the EFFECT of sounds you can feel but not hear.

I know this aura shaker is used in low frequency applications, but something vibrating at a given frequency will make noise, and hence should be able to cancel noise too.

I guess this would be the opposite of a transducer.

-BdK

OOOOHHHH CHECK THIS OUT!!!!!!!!!!

http://www.val.me.vt.edu/Experiment/experiments.html

See!! I told you software existed.... (even tho the software is only a demo....hehe)

Rather interesting eh?

Noise cancellation has already been tested in cars. The cars were fitted with hardware that were able to record, create cancellation waves, and play them back in real time, all with solid state devices eg chips.

Maybe I could just go buy a pair of ridiculously expensive noise cancellation headphones and rip the chip out of it

so, could putting a speaker directly in front of a PC cancel out that PC's noise with effectiveness caparative to the 2 speaker example shown at that software link???

When you have 2 speakers face to face and out of phase, can you walk all around those speakers and without hearing them, or at least hearing them as much quieter???

~BdK

anc in cars

ANC was abandoned in cars due to safety concerns. i.e. not hearing a car's horn because of noise cancellation.

BdK