Noise Blocking Servo Headphones & Servo Mics

I'm thinking it might be feasible using a negative feedback arrangement with panels perhaps driven capacitively. I seem to recall seeing a circuit that worked something like that - though I don't know whether it was for a headphone or a larger space. I don't know how practical it would be to have entire walls that functioned as speakers - and I doubt it would be a big hit with people who care about decorating... but the idea is intriguing.

on.

ugh

o

At first glance it seems like that would much simpler, infinitely more uniform, much cheaper and more energy efficient than lots of coils and magnets.

A "capacitor panel" would need to be able to be charged / discharged in less than a millisecond and to balance the acoustical "pressure" the electrostatic "pressure" may have to be large.

Anyone have any ball park numbers for caps that would summarily rule this out? How many microns can a cap squeeze inward? If it requires

92 kV maybe it's not ready for Home Depot.

With 3 layers the panel facing outside would accelerate to and from the middle panel squeezing and expanding the dielectric material. That voltage could then be used to squeeze or expand the dielectric between the panel facing inside and the middle panel.

A plus pressure causes the overall panel to get thinner and a "minus" pressure thicker so that the inside panel never moves.

It would be interesting to see if it was possible to get something to work with the "servo" inherent in the design, i.e., with nothing more complicated than one or 2 DC voltage sources.

If it was easy to roll up like wall paper. Someone mentioned a clear conductor so windows could be made to work as well.

Building codes now require builders use [passive] sound proofing materials. The problem with this is you can't hear what's outside when the jets are gone.

Silence should be possible with the flip of a switch.

Bret Cahill

Look into "electrostatic speaker drivers". They are fairly simple arrays that you can build yourself. They typically have a thin aluminized mylar film suspended between metal grids. You apply a high bias voltage (5000 V or so) between the metal grids, and you run the output of your stereo through a big step-up transformer.

They have the best high frequency response of any speaker type, very smooth and extended. They are not very sensitive unless the panels are large. The problem is that the high frequencies are coming from a large flat panel, so they "beam" straight out... the opposite of the ideal point source that radiates equally in all directions. So you have to align them to point at your favorite listening chair.

Obviously, that would be no good for noise cancellation. As noted previously, the drivers need to be small to cancel high frequencies. But then I think you would run into problems because with less suspended area, the edge stiffness of the suspension (where the film is clamped) would dominate and the sensitivity would drop even further than the simple reduction in area would imply. But you can't raise the grid or driving voltage too high, or you get arcing.

In short, bad idea... and that doesn't even consider the issues with using them as sensors.

Best regards,

Bob Masta DAQARTA v6.02 Data AcQuisition And Real-Time Analysis

Scope, Spectrum, Spectrogram, Sound Level Meter Frequency Counter, FREE Signal Generator Pitch Track, Pitch-to-MIDI Science with your sound card!

ems:

rs and

could

ar to

ms

2. >

ositi=3D

thro=3D

o a

to=3D

l

ams

t s

If "high frequency" means anything over 2 kHz then maybe that can best be done with conventional / passive materials.

Large displacement caused by the sound =3D> high voltage voltage in caps.

Using the panel as a sensor might not be all that much of an advantage.

OK, maybe I am using the wrong term here. I wasn't suggesting to use capacitors to drive the panel; but that the whole panel itself could act as a large capacitor with one of the plates free to move.

The free-moving "plate" (probably a thin plastic film with a metallic conductive surface) would be driven by a voltage in response to the noise (but of opposite sign, so as to cancel it). The other "plate" of the capacitor could be wire mesh, the holes allowing air to move quickly so as to permit adequate motion of the film.

The reason I suggest this arrangement is that it seems it would require far less energy to operate, than trying to use speakers driven by inductors.

I actually made a "speaker" as I described, just as a sort of proof of concept thing. It worked, after a fashion (note that I wasn't trying for noise cancellation, just to see whether capacitive speakers were even possible).

I just Googled "capacitive speakers," and found that they're definitely possible and being sold.

So the main question is whether they'd be any good for canceling noise. I suppose they'd require a large range of motion and quick response time. I don't know whether that would be practical. But someone might want to give it a shot...

Agreed.

>

See my previous response about electrostatic speakers. I have no information on whether they are more or less efficient (in terms of energy use) than conventional voice coil speakers, especially after you take into account the high voltages needed. But they definitely have problems with large signals on the one hand and sensitivity on the other (without arcing).

Note that there is another problem with the simple 2-plate (membrane and grid) capacitor scheme you mention: The response is *way* less linear, actually a square law. This is not a deal-breaker for a servo system that is only trying to maintain a stationary position, but it's bad news for sound reproduction.

I used to work in a university lab where we did basic research on hearing mechanisms of the inner ear. We used guinea pigs, mostly. We needed to provide loud high-frequency sound (in the 20 kHz range) in order to stimulate the part of the inner ear that we were studying. (The high frequency sensory "hair cells" are the first ones the incoming sound encounters, with progressively lower frequencies handled by cells farther along.)

We needed to deliver the sound directly into the ear canal to avoid problems with head orientation and such. We used a

1/2 inch "condenser" (capacitor) microphone driven in reverse, driving a short ear tube. The mic is a thin metal membrane held very close to (but insulated from) a metal backing plate. The rated voltage was just over 300V before arcing. The attraction or repulsion forces acting on the membrane are proportional to the square of the applied voltage, such that the force is the same regardless of which plate is positive or negative.

The "standard" approach was to just apply 200V bias to the membrane, and apply the desired signal on top of that, up to

+/-100V. It wasn't perfect, since this was still a square-law system, but the resultant 2nd harmonic distortion was low enough to not cause a problem, usually.

I did build a special square root circuit to reduce the distortion, but it was more trouble than it was worth, as I recall.

For pure tones, which is what we mostly used, we could actually omit the bias and apply +/-300V directly at *half* the frequency, since the square of a sine is another sine at

2x the frequency. The problem was that when the tone started (ramped up carefully to control the spectral splatter) the membrane had to move from a resting position to a new average value, which caused a "thump" on top of the tone if we brought it up too fast. So, we used that only for long ongoing stimuli, not transient bursts.

Ahh, the good old days...

Best regards,

Bob Masta DAQARTA v6.02 Data AcQuisition And Real-Time Analysis

Scope, Spectrum, Spectrogram, Sound Level Meter Frequency Counter, FREE Signal Generator Pitch Track, Pitch-to-MIDI Science with your sound card!

t

es 92

It was originally understood as a single large area, i.e., 8' by 15', flate cap.

On the other hand, if a lot of "speakers" or non moving panels are required, it is probably cheap and easy to divide and wire the mylar into sections.

or

=A0

ht

Preferably it should come rolled up like wall paper.

Or some automatic noise threshold switch.

Bret Cahill