Note: Descriptions are shown in the official language in which they were submitted.
~2~
BACKGRûUND 0~ THE INNENTION
The invention relates to methods for the transmission of
acoustic information as perceivable vibrations, and to devices for
S the implementation of these methods. Methods and devices of this
general type are known from US Patent 4,289,~35.
In vibratory transmission of acoustic signals such as a Yoice,
the skin (e.g. at the wrist) is mechanically stimulatea. The
perception range of the skin for such mechano-cutane stimuli starts
at rather low frequencies, approximately on the order of magnitude
of 20 Hz. But the upper range (about 1.2 kHz) is relatively low
compared to the overall range of audible frequencies, (see e.g.
Funkschau 11 (198~) page 53). Unvoiced consonants such as s, sch,
ch, which acoustically predominantly have frequencies hi~her than 2
kHz, can therefore not be made perceivable by linear ~ransmission by
skin stimulation. In known instruments, the use of a microphone is
resorted to which is held close to the mouth so that the air blown
out of the mouth onto the microphone diaphragm generates a rumbling
noise characteristic for unvoiced sounds, which can then be made
perceivable.
In a portable device for the transmission of acoustic signals as
vibrations, e.g. in instruments which are to be used as
communication aids, i.e. a kind of hearing aia7 talking into an
extremely close microphone is not possible, first for cosmetlc
reasons and secondly because unvoiced sounds from both nearby an~
further away are to be transmitted. In such instruments, the
microphone is permanently joined to the amplifier unit as in
conventional pocket hearing aids, or else the microphone is
attachable to clothlng and located approximately 2û to 30 cm away
from the mouth.
Known from US Patent No. 4,289,935 is an instrument in which, in
order to obtain good intelligibility and simplification of the
equipmental design otherwise usual in hearing aids for the extremely
hearing impaired, the signals to be transmitted are divided into
S several frequency bands, the division being used to modulate
alternating voltages (sounds) which are then conduzted, together
with the signals coming from the microphone, to a receiver or else
to a vibrator (vibrotactile stimulus generator) after
amplification. For unvoiced sound this requires a switching from
the line spectrum to a noise spectrum or a line spectrum. The
vocoder provided, i.e. the high cost associated with such a
conversion, has so far been a hindrance to its introductlon ln the
hearing aid industry.
It is an object of the invention to provide a method and devices
for the implementation of this method which will work satisfactorily
under the usual hearing aid conditions.
SUMMARY OF THE INVENTI~N
The invention provides for a transformation of the
characteristic high frequency of the sibliants into the range of
perceivable frequencies. To do this, the acoustic signal can be
transformed in a microphone into a sequence of electrical signals
and fed to the vibrator amplified in the usual manner~ while the
signals are also branched off from the transmission path, and the
high frequencies corresponding to the sibliants are separated from
them by means of filters. These frequency components,
characteristic for the unvoiced sounds, are then fed to an
intermodulation amplifier so as to obtain sum and differential
products of the frequency components. The result of the
intermodulation amplification, i.e. the sum and differential
frequencies of the unvoiced sounds, is finally returned to the
conventionally amplified signal. To avoid overloa~iny the output
amplifier with the high-frequency sum frequencies, it is expedient
to isolate the low-frequency components below lOû Hz which can ~e
made perceivable by a vibrator and to pass only them on. There is
finally obtained, besides the perceivable vibrations occurriny in
known instruments, a vibration series characteristic for the
unvoiced sounds. This requires only the acquisition of the
sibliants themselves, and no blowing noise at the microphone, so
that an unequivncal identification of the sibliants becomes possible
also in portable instruments. On the other hand, the implementation
of the method merely requires a design in which a few simple
components operable ~ith the usual hearing aid batteries are
sufficient.
In one embodiment of the invention the acoustic signals a~e
received by a microphone and amplifie~ in a preamplifier that has a
limiter. The limiter assures that the signal at the preamplifier
output is not distorted due to harmonics originating at the limiter
which could artificially simulate the presence of unvoiced soun~s.
The signal thus amplified then reaches the vi~rator directly via a
volume control and output amplifier.
For the transformation of the high frequencies above 2 kHz which
are characteristic of unvoiced sounds, the voice signal is branched
off after the preamplifier and filtered first by a high-pass
filter. A frequency transfer at the filter limit of 12 dB/octave
suffices because the desired effect is clearly achieved at
relatively little cost. A frequency band characteristic for the
further processing of the unvoiced sounds is thus obtained by
isolating the frequencies approximately 2 kHz and higher. The
frequency components contained in the band are not discrete spectral
lines, but rather resemble a high-pass filtered, wide band noise
signal (e.g. high-pass filtered white noise). But the noise present
contains nevertheless a frequency distribution characteristic of the
respective sibiiants.
In an amplifier stage succeeding the high~pass filter the signal
is brought to a level sufficient for the modulation in the
subsequent intermodulation stage to be such that a linear signal is
obtained over as wide a range as possible. In this stage, which has
nonlinear characteristics, intermodulation products are produced
from the previously filtered noise, i.e. the high-frequency
components at the input of the intermodulation stage occur at its
output additionally as sum and differential frequencies. The
nonlinear characteristic of the intermodulation stage may correspono
to a quadratic transfer function. This is advantageous because only
sum and differential frequencies of the fundamentals are produced,
but not sum and differential frequencies of multiples of the
fundamentals. This would render differentiation of the unvoiced
sounds more difficult.
s
Some of the differential frequencies fall into the low-frequency
range below 1 kHz which can be made perceivable by means of a
vibrator. Therefore, these frequencies are expediently separated,
as mentioned above, in a low-pass filter, for which a second-order
filter also suffices. The signal thus obtained can then be matched
to the level of the original signal in an amplifier stage and added
to this original signal in a summation stage so that the vibrator
receiYes, via the loudspeaker and the output stage, a signal mix
containing the transformed frequencies of the sibilants, besides
more frequencies ~hich are present in the simple amplification.
Usable as hiyh and low-pass filters in the branched signal p~tns
are any active or passive second-order RC filters. THe limit
frequency of the hlgh-pass filter should be between 2 kHz to 4 kHz.
Characteristic frequency components of the unvoiced soun~s are
present in the voice frequency spectrum from 2 kHz up. At a limit
frequency higher than 4 kHz the various unvoiced sounds may no
longer be differentiable. A low-pass filter is not absolutely
necessary because the perceptibility ranye of the skin ends at about
1 kHz, but such a filter is useful to prevent overloading the output
stage with differential frequencies outside of the required
frequency range. The suitable limit frequency ranges from about 200
Hz to 1 kHz. The transmission of the differential sounds in the
range up to about 2Q0 Hz contains the important information
(comparable to the blow noises of a microphone near the mouth) and
must be able to take place. The transmission of frequencies higher
than 1 kHz is unnecessary because they are outside of tne
perceptibility range of the skin.
.. ......
To generate the intermodulation siynal, a stage can be
used which contains an operational amplifier that feeds the signal
to a diode~ There the bulk of the intermodulation signal is
generated in that a multiplication of the sum of the frequencies
by the nonlinear characteristic of the diode takes place. Since
a desired dynamic range, iOe. one within the operating range of
the diode, can lead to overproportionally great output signals
when strong signals are modulated, it is expedient to compensate
this rise of the characteristic by diodes poled in opposition to
each other. A linear characteristic of the intermodulation
products is achievable by appropriately choosing the rest of the
components.
Thus, in accordance with a broad aspect of the
invention, there is provided a method for transmitting acoustic
information, comprising: converting the information into an
electrical signal; isola-ting high-frequency components of the
signal; forming differential frequencies of said components;
and transmitting said differential frequencies to a vibrator.
In accordance with ano-ther broad aspect of the
invention there is provided a device for transmitting acous-tic
information, comprising: means for converting -the information
into an electrical signal; means for isolating high-frequency
components of the signal; means for forming differential
frequencies of said high-frequency components; and means for
transmitting said differential frequencies to a vibrator.
In accordance with another broad aspect of the
invention there is provided a device for transmitting acoustic
. ~
-6
~6~
information, comprising: means for convertiny the information
into an electrical signal; means for isolatiny high-frequency
components of the signal; an intermodulation means having a non-
linear transfer function for forming sum and differential
frequencies of said high-frequencies components; means for
isolating the differential frequencies; and means for transmitting
said differential frequencies to a vibrator.
BRIEF DESCRI TION OF THE DRAWIN5S
Exemplary and non-limiting preferred embodiments of the
invention are shown in the drawings, in which:
Fig. 1 is a basic block diagram of a preferred
embodiment of the invention; and
Fig. 2 an example of the circui-t for the intermodulation
stage contained in Fig. 1.
-6a-
~.
DETAILED DESCRIPTION GF_PREFER~ED EM8CUIMENTS
1 in Fig. 1 is a microphone connected to a limiter amplifier 2.
From it, a ~ine 3 goes to a headphone 6 via a volume control 4 and
an output amplifier 5. There is another connection of the
preamplifier 2 to a vibrator 10 via another line 7, a sum,nation
circuit 18 and toth a volume control 8 and output amplifier 9.
Finally a line 11 is present which leads through a switch 12 to a
high-pass filter 13 which only passes frequencies above 2 kHz.
These frequencies then reach an amplifier 1~ whence they arrlve at
an interModulation amplifier 15. Tne output signals of the
intermodulation amplifier 15 are then fed via a low-pass filter 16
and an amplifier 17 to a summation circuit 18 through which the
signal branched off by means of line 11 and processed in the
elements 13 through 17 is reunited with the original siynal
transmitted by line 7.
The signal path between the microphone 1 and the headphone 6
corresponds to that in conventional hearing aids. The signal is
acquired by the microphone 1, transformed lnto a sequence of
electrical signals and amplified in the preamplifier 2. After
having passed the volume control 5, this signal reaches the output
amplifier 5 whence the headphone 6 is then activated. This branch
with its components 3 through 6 can be omitted, if an instrument is
involved intended to serve persons with complete loss of hearing; it
then suffices to feed the signal processed in the preamplifier 2
directly via the line 7 and a volume control 8 to an output
amplifier 9 whence the vibrator 10 is operated.
In the instrument the signal is branched off between the
preamplifier 2 and the volume control 8 via a line 11. This signal
can then be connected to the processing stages 13 through 17 via the
switch 12, the processed signal then being added again in the
summation circuit 18 to the original signal path flowing through
line 7.
When the switch 12 is closed, the high-pass filter 13 whose
limit frequency is near 2 kHz isolates the signal components above
the limit frequency and then the signal is brought in the amplifier
14 to a level sufficient to drive the intermodulation stage 15. In
the low pass filter 1~ the summation signals of stage 15 are then
separated and, because of the 4ûO Hz limit frequency of filter 16,
only those components are passed which can be made perceivable in
the vibrator lOo Finally, these components are brought in the
amplifier 17 to a level which makes it possible without interference
to add in summation circuit 18 the signals processed between 13 and
17 to the original flow in line 7 again.
Used as high-pass filter 13 is an active second-order RC filter
whose limit frequency is near 2 kHz, and as low-pass filter 16 a
second order filter with a limit frequency near 400 Hz.
According to Fig. 2, the intermodulation stage 15 has after its
input 20 a resistor 21, an operational amplifier 22, a ~iode 23 and
a capacitor 24 before the connection ends at the outp~t 25. Between
the dioae 23 and the capacitor 24 is a branch to ground at the
chassis via a resistor 2~. Between the amplifier 22 and the diode
23 is a connection 29 to the positive power supply via resistor 28.
. ~,
~23,~
~he amplifier 22 has a cor~ection 34 to the positive
and a connection 3~ to the negative power supply.
The output and the inverting input of operational ampli-
fier 22 are bridged via a resistor 30. Also present is another
bridge of the amplifier 22 via a connection containing a
resistor 31 and dlodes 32 and 33 which are in parallel and
connected with opposed polarities. The other input of the
amplifier 22 is connected to ground at the chassis via a
resistor 36.
For an instrument according to Fig. 1 which is operated
with a power source of 4 to 6 V, the intermodulation amplifier
15 is expediently designed with an operational amplifier 22
characterized by a minimum supply voltage of 3 V and an upper
limit frequency of 20 kHz at 30 dB amplification, and equipped
with a 10 kOhm resistor 21, a 100 kOhm resistor 26, a 3.3 kOhm
reslstor 28, a 330 kOhm resistor 30 and a 22 kOhm resistor 31.
The capacitance of capacitor 24 is l ~f. The diodes 23, 32) 33
are commercial small signal silicon diodes with approximately
400 mW power ratings.
When the components of the intermodulation amplifier 15 are
designed as described above, the resistors 21 together effect
a 33 dB (or approximately 30 dB) amplification of the amplifier
22. Because of the nonlinear, partly square characteristic, the
diode 23 furnishes the actual intermodulation signal which is
then coupled out to a load resistance (not shown) by the capa-
citor 24. In order to keep the intermodulation signal as great
as possible, this load resistance should be in the order of
magniture of 100 kOhm. By connecting the positive supply vol-
tage at 29 via the resistor 28 it is achieved that the ampli-
fier 22 operates in its proper operating range. Connecting
the resistor 31 in series with the oppositely poled diodes 32
and 33 results in the compensa-tion of the overproportional
rise of the transmission characteristic of the intermodulation
amplifier at higher levels, thus linearizing the transfer
characteristic.
Those skilled in the art will undPrstand that changes can be
made in the preferred embodiments here described, and that these
embodiments can be used for other ,ourposes. Such changes and uses
are within the scope of the invention, which is limited only by the
S claims which follow.
--10--
