Active Fan Audible Noise Canceling Example Project Needed

[danadakk]

I have fans mounted on a wood stove that create a fair amount of audible whine noise.

Seeking a Arduino or Raspberry Pi based project that processes a mic and drive speakers
to do active cancellation of the fan noise.

Do not want to replace fans at this point.


Regards, Dana.

 

[danadakk]

Made some progress, some stuff on youtube.

 

[BradtheRad]

Our wood stove had a fan which made a loud whoosh. I built a speed control using a triac. By reducing speed a bit it made the room noticeably quieter.

Fortunately the fan had the type of motor whose speed could be reduced by shortening the duty cycle as a triac does. Not all motors are compatible with a triac.

 

[KlausST]

Hi

Active noise cancelling works good on "known conditions" and only to an upper frequency limit.
How good it can work in free air (in opposite to headphones) depends on where and how the noise is generated, how the noise is reflected by the room, and where in the room you want it to be cancelled. You can't achieve almost equal cancellation at every position in the room.

It is way more difficult than noise cancelling on headphones.

I don't want to discourage you ...
If it was my task ... I'd treat the problem rather than the symptom.

Klaus

 

[danadakk]

Our wood stove had a fan which made a loud whoosh. I built a speed control using a triac. By reducing speed a bit it made the room noticeably quieter.

Fortunately the fan had the type of motor whose speed could be reduced by shortening the duty cycle as a triac does. Not all motors are compatible with a triac.

They are DC fans on a bench power supply, noise it seems linear with Vapplied.
And more airlfow - more heat, so I do not want to sacrifice performance if at all
possible. My noise is a fairly high pitch whine. I think the fan noise is being amped
by the sheet metal it is attached to, so thinking I will take a speaker and just set it
magnet side on sheet metal, detect freq and invert phase back to sheet metal.

Its a crap shoot but I am going to try it, make some FFT measurements to see how
complex the harmonic content is/isn't of the noise. I will use a part that can easily gen up
4 harmonics, sines, using wavetables driven by 4 DDS, all onchip, and sum them, see what
happens. Maybe use an algorithm to effect some minimalist kind of AI to autonomously
find correct amplitudes and freqs to get min noise.

Adding this to the other 5,000 projects I am working on.....


Regards, Dana.