[Knack]

"just wondering" wrote in message
...
"Knack" wrote in
:

We currently have one of the quietest ceiling fan models in our
bedroom, but I'm wondering if I can suppress that fan noise even
further. However, we don't want to wear earplugs or noise-cancelling
headphones while we sleep.

Just an idea-- Was wondering if it could be possible to somehow make
the sensor and signal processor of one of those noise-canceling
headphones (For example, Panasonic RP-HC500) drive a larger amplifier,
which would in turn power a pair of small cheap PC type speakers. The
little speakers could then be spaced apart above the headboard of our
bed.

Was hoping that I could obtain all components that I need (except the
RP-HC500 and the PC speakers) at Radio Shack or some other local
electronics components retailer. However, I'm not an electronics
tinkerer so I have no clue as to what the specs would be for those
components.

From school (long ago) I have a vague and very limited knowledge of
electronics (Ohm's law, impedence matching), but know nothing about
amplifiers and next to nothing about filters/tuners.

Don't want to spend more than a total of US$400 on this. Is the idea
feasible? If so, then if someone could give me some tips/steering
about the specifics of what I would need, then I'd be greatly
appreciative.



Just to bring some sanity to the discussion, let me point out that what
can vs can not be done will depend on the characterstic of the fan noise
that you want to cancel. Since you don't give any detials, there really
is no difinitive answer. Perhaps the fan motor is coupling 120Hz to the
ceiling structure and the ceiling is acting as a sound board. If this is
the case, mechanically isolating the fan would help. An alternative to
active noise cancellation would be to reduce the speed of the fan.

The ceiling fan is well designed. It has an elegant flexible elastomer mount
within its flush mounting base, which in turn makes contact with the room's
soft, low density ceiling tiles that have excellent vibration absorbing
characteristics of their own. So I'm quite certain that practically none of
what I hear is sounding off the ceiling.

However, I can definitely hear two distinctly different noises coming out of
this ceiling fan: a low rumbling sound and a higher pitched sound.
Presumably, the former is due to air turbulence and the latter is from the
electric motor.

Unfortunately the higher pitch sound component is the more objectionable of
the two that I wish to attenuate. Surely, someone with a flush mounting 52"
(132 cm) ceiling fan of their own, and a microphone, and an oscilloscope
could easily tell me what those two frequencies are. Hopefully the two noise
frequencies are both within the attenuable range for ANC. Since posting this
thread I've learned that active noise control (ANC) produces *good* results
only for frequencies below 300 hz.

A typical bedroom is not a huge space. And remember that the design need
only satisfy a maximum of 2 sleepers who's ears will always be located at
approximately the same places whenever noise control is needed. Although the
location of a double-bed cannot be expected to match one of the ceiling's
centerlines exactly, it will be reasonably close, because there will be
space on each side of the bed for separate access (It's not logical to put
the *length* of a double-bed right up against a wall without providing any
access space there.).

For example, our bedroom measures 10'6" x 12' (3.2 m x 3.7 m) and has a
queen size bed located with its centerline only 14" (36 cm) off the
centerline of the room's longer wall. Consequently in 3D space the distance
to the noise source (ceiling fan) from the center of each sleeper's head is
a difference of only 9%. Each sleeper's head would be equally distant at
about 25" (64 cm) from the proposed noise sensor.

Thus, based on the layout of my own bedroom I'm guessing that the size of a
typical bedroom nor the the slightly asymmetrical locations of 2 sleepers in
it shouldn't present much of a problem.

A major problem with implementing the idea is that there will be no
available electronic schematic diagram for any ANC headphone with which to
build the system around. I just realized that the device would have to be
hacked by someone with considerable electronic experience having data about
various ICs in its circuitry. If ICs for the device's signal processor
output and its amplifier input are both proprietary, then I guess I'm stuck.