вход по аккаунту



код для вставкиСкачать
Patent Translate
Powered by EPO and Google
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
In some cases, only the filter 403 and the circuit 405 may be referred to as a 1-in-one R filter, for
the sake of context of this invention. Thus, the circuit of FIG. 26 shows an improvement of the
circuit of FIG. 21 by, for example, adding an IIR filter in series with the F-type R filter of FIG. The
F-R adaptive filter is believed to be suitable for compensating components of the feedback path
Hf1 o5 consisting of acoustic delay and attenuation. The engineering RJ-responsive filter (403,
405) is suitable for compensating for the resonance force present in the feedback path.
Resonances and very long feedback paths in hearing aids, loudspeakers and other electroacoustic
devices are better removed by internal feedback and memory provided by the AE filter method.
The filters 401 and 403 and the combining circuit 405 are first circuits (e.g. blocks 281.109, Ill).
115) in response to the signal derived from the signal supplied to this first circuit interconnected
with 115), the filtered signal (for example the output of Hs) and the second further signal (for
example se) A digital adaptive filter that generates an adaptive output X for the first circuit to
process the electronic infinite impalne response (AER) to substantially cancel the feedback
component in the electrical output of the hearing aid microphone An example is constructed. The
digital adaptive filter has first and second digital filters 403 and 405. The first digital filter 401
has an input 407 that receives a signal Y, which is an example of a filtered signal that is
combined with a second other signal. The output of the filter 401 comprises an output 409
connected to the second digital filter 403. 405 in the combining circuit 405. This second digital
filter 403, 405 produces an output X, part of which is fed back to itself at the input 411 of the
block 403. The output X of the second digital filter is provided to the first circuit for electronic
processing by the subtractor synthesis circuit 281 so as to substantially cancel the feedback
component in the electrical output of the hearing aid microphone . The filter coefficient update
operations in filter blocks 401 and 403 are similar to the process already described above,
except that the coefficients are included. The coefficient of the filter 401 is represented by W,
and the coefficient of the filter 403 is represented by ah. Here, X is an index for specifying each
specific coefficient (indicone, similar to 1 described above), and represents each specific
The regression expression form for updating the coefficients is: Wk (n + g = Wk (n) + sgn (En?Y
(n?k)) (13 ? ? ak (n + 1) = ak (n) + Sgn (En, X (n)) -: K)] (14, 1 In FIG. 26, the filter block 403
has the transfer function A which makes an adaptive change. This transfer function, represented
by the symbol "A", is a very complex function in the Laplace (or S or 2) domain. The filter block
403 has a transfer function 1 / (1-A) together with the synthesis circuit 405 and the feedback
thus formed. If the transfer function of the processing R filter 401 can be expressed by W, the
transfer function He of the entire filter configuration 401140.305 of FIG. 26 is W / (1-A). The
error filtering characteristic is further improved in FIG. The filter 401 of FIG. 26 is shown as the
filter 413 with the logic circuit 417 of FIG. Logic 417 is responsive to the output Y and the
filtered error 100 V that is now disturbing. The filter 423 and logic circuit 425 of FIG. 27
correspond to the filter block 403 of FIG. The combining circuit 427 adds the outputs of the
filters 413 and 423 to generate an output X. The output X is also fed back to the filter 423. Logic
circuit 425 is shown separate from filter 423 as it is convenient to also use it as a logic circuit for
additional error filter 431. The error filter 431 is connected at its input to receive the synthetic
signal En in the form of an error signal for the feedback filter He. A combination circuit 433 ij
shown as a subtractor is supplied with the combined signal input En and the output of the error
filter 431. The output V of the circuit comprising the filter 413 and the circuit 433 is a filtered
error signal expressed by the following equation: V = En (1-A) (15) The logic circuit 425 gives the
filters 423 and 431 the same factor. Logic circuit 425 is responsive to outputs ? and X and may
be comprised of any of the circuits shown in FIG. 12, FIG. 16, FIG. 22, FIG. 23 and FIG. Only the
polarity of the human power En or all the information of polarity and magnitude can be used.
The set of regression expressions for updating the coefficients of the three filters 413.423 and
431 in FIG. 27 is: WK (n + 1) = Wk (11) + sgn (V, 'Y (n?4cJ, l 16) a!
?????????????????????????????????? Since the
coefficients of both filters 423 and 431 are correspondingly identical, three equations are not
required. In each of FIGS. 261 and 27, it should be described that the digital adaptive filter
configuration corresponding to the rectangular frame He in the several block diagrams described
above is surrounded by the rectangular frame delineated by a broken line. . This additional error
filtering by filter 431 and subtractor circuit 433 of FIG. 27 provides a further advantageous
feature in that the operation of the adaptive aI wave circuit is smooth and effective. The
experimental and self-explanatory explanation of this invention is that elements 423 and 427
operate because of the l / (1-A) transfer function with the denominator (1-A) approaching zero at
the frequency It is believed that the loop (423, 427, 281, 425, 4417, 413) which lacks a little
stability can not be formed. By adding picture elements 431 and 433 with transfer function (1-A)
(which is the reciprocal of l / (1-A)), (1-A) Xi / (1 A multiplication of -A) occurs and the influence
of the denominator as described above disappears. This explanation is given by this example
though! It is not intended to exclude other excellent effects or actions not described in the above,
that is, to limit the description to only the matters described. The filter 431 and the combining
circuit 433 thus constitute an example of a third digital filter which filters the signal (for example
the combined signal input En) which is also supplied to the first means for electronic processing.
The filter 423 is an example of a digital filter section having a transfer function and generates an
electrical signal to be fed back to itself. The filter 431 and the combining circuit 433 form an
example of means for electronically filtering the combined signal in accordance with the transfer
function which is the inverse of the transfer function of the digital filter section, and which drives
the logic circuit. The logic circuits 425 and 417 together act as a type of logic circuit receiving a
signal from the third digital filter to control the first and second digital filters for infinite impulse
response (AE) processing Control. Various modifications of the connections to logic circuits 417
and 425 and the use of multiple synthesis circuits, etc. may also be used, in accordance with the
description already given in connection with FIGS. 15, 17 and 21. The process of processing the
electrical output of the microphone 13 to produce the combined signal human power En, to
generate a filtered signal according to the amplification and filtering Hs of the hearing aid, starts
with 5TART 501 and proceeds to step 503.
In the next step 505, false random as an example of a second alternative signal distinguishable
from the above-mentioned filtered signal? Generate a globe signal such as an I sound signal. In
the next step 507, the filtered signal is combined with the second other signal to the receiver of
the hearing aid. The second further signal is suitably weighted with a weight value W1 so that its
magnitude is smaller than that of the filtered signal, which is usually the loudness of a normal
speech. If the magnitude of the filtered signal changes, then, in step 509, the magnitude of the
second alternative signal (eg, noise) is changed as a function of the magnitude of the filtered
signal. For example, in one such embodiment, it is desirable for the magnitudes of both to have a
'Z constant ratio, by varying the magnitude of the noise directly corresponding to the magnitude
of the filtered signal. In the next step 511, the filtered signal is weighted by the weighting factor
W2 such that the loudness of the normal speech and the noise magnitude is generally greater
than the filtered signal magnitude, and Filtered combines the noise with the signal. Subsequently,
in step 513, a series of digital values with a polarity that is responsive to the dropout signal are
retrieved, such as by shifting the site random noise through the shift register. Then, in step 515,
the continuous sum is electronically held in the set of registers. Each successive sum is increased
or decreased depending on whether the polarity of the corresponding value of this series of
digital values is the same or not as compared to the combined signal input. In step 101'7, the
index N is checked to see if the index N has reached a predetermined value, such as 20. If it has,
then the continuous sum at step 515 is added to update the set of coefficients of the digital filter
and the index N is reset to one. Also, if the index N has not reached 20, step 517 causes the
operation branch to step 521 to increase the index N by one, and step 519 is skipped. In this
way, the continuous sum in the register set is electronically added to the coefficients at step 519,
respectively, less frequently than the frequency of occurrence of the increment / decrement
operation at step 515. After steps 519 or 521, operation proceeds to step 523 to adaptively
filter, for example, the second separate signal (or glove signal) synthesized in step 511 and the
filtered signal.
This filtering action is performed, for example, in response to digital coefficients that change only
as a function of the polarity, producing an adaptive output in the hearing aid. In this step 523 the
Grosene air R component adaptively filters the filtered signal and the second further signal to
produce an electrical signal, and this electrical signal is also adaptive in the same step 523 This
electrical signal is fed back as shown by arrow 525 to be filtered to produce an adaptive output.
Further, in step 527, the adaptive output is combined with the microphone output to
substantially cancel the acoustic feedback component in the microphone's electrical output to
processing step 503. If this process continues, the operation loops back to step 503 via test 529.
If not continuing, this operation branches from Tenu l-529 to E N D 531. The present invention
includes various embodiments incorporating software, hardware or firmware according to the
type and application using digital technology or analog technology. Applications, combinations
and glosses 4 intended for hearing aids, loudspeakers and other electroacoustic devices,
generally used in the atmosphere, in water or in other environments, are included within the
scope of the invention. From the above, it will be appreciated that the various objects of the
invention will be achieved and advantageous obtained. Since the configuration of the present
invention described above can be modified without departing from the scope of the present
invention, all the items included in the above description and shown in the accompanying
drawings are merely examples. It should not be construed that the nature of the invention limits
the scope of the invention. This invention is based on the United States Veterans Affairs (Y, A,)
contract V674-P-857, V2V5-P-1736, United States Federal Republic of China (NASA) Granted No.
N AG 10-0040. It was made with government support, and the copyright for the disclosure of
this specification and drawings belongs to the Central Institute for the Deaf. Thus, the copy is
subject to copyright, except as it relates to the patent application and registration of this
Brief description of the drawings
FIG. 1 is a simplified sketch of the hearing aid of the user of the invention including the
electronic filter according to the invention, showing this hearing aid partially in cross section,
and FIG. 2 is a side view of the hearing aid shown in FIG. A pictorial view, FIG. 3 is an electrical
block diagram of the hearing aid of FIG. 1 including the electronic filter circuit having all the
hardware circuits characteristic of the present invention, shown along with the acoustic feedback
path, FIG. FIG. 5 is an electrical block diagram of the entire hardware circuit comprising the
features of the present invention for the illustrated electronic filter, FIG. 5 is a serial interface in
the electronic filter of FIG. 3, random access memory and some Filter Limiter-Electrical block
diagram of the section of the filter, FIG. 6 shows a finite impulse response (F.F.) digital filter used
in some of the circuits comprising the features of the present invention FIG. 7 is a diagram
showing a part of the operation of a preferred embodiment of the digital filter for adaptively
simulating the characteristics of the feedback path. Figure 8 is a partial block diagram, Figure 8
is a diagram of the feedback path supplied with the globe signal, with fractions giving multiple
different delays to the probe signal, Figure 9 is a composite input signal A voltage vs. time
relationship diagram, including two curves delineating it and a probe signal synchronized thereto
and a third curve representing a product signal that maintains a positiveness to increase the filter
coefficients to simulate the feedback path, 10 The figure shows two curves showing the synthetic
input signal delayed with respect to the probe signal and sometimes negative when it is positive,
thereby simulating the feedback path A third set of curves showing voltage vs. time, consisting of
a third curve representing the product signal with almost no effect on the values of the other
filter coefficients, and FIG. 11 is a portion near the speech zero crossing point. Figure 12 is a pair
of voltage versus time diagrams showing the polarity of the composite input signal including nu
beacs that match the random noise probe signal and the polarity match, Figure 12 comprising
the features of the present invention for controlling the adaptive filter of Figure 4 FIG. 13 is a
schematic view of a data banu transmitting a digital signal having a size and a polarity, and FIG.
14 is a comparator for detecting the polarity signal. FIG. 15 is a block diagram of an entire
hardware circuit provided on a VLSI die having first and second synthesis circuits for noise
signals according to another embodiment of the present invention. , FIG. FIG. 17 is a schematic
diagram showing, in block form, another form of an adaptive filter for simulating a feedback path
and another form for controlling, FIG. 17 is another hard disk provided on another VLSI die
having the features of the present invention FIG. 18 is a block diagram of the wear circuit in
which the magnitude of the noise as the second separate signal is generally greater than the
magnitude of the filtered signal at the loudness of normal speech. A diagram showing the
relationship between speech and site random probe noise components versus time in the circuit,
both FIG. 19 and FIG. 20 show that the magnitude of noise as a second separate signal is the
same as that of a normal speech voice. Speech and site randomness in a separate synthesis
circuit where the magnitude is generally smaller than the magnitude of the filtered signal and the
magnitude of the noise is varied as a function of speech magnitude FIG. 21 is a diagram showing
all hardware provided with the feature of the present invention on a VLSI die in the form of using
a subtractor as a synthesis circuit for canceling the feedback component. FIG. 22 is a schematic
diagram showing a part of the logic circuit having the features of the present invention used in
the circuit of FIG. 21; FIG. A schematic diagram showing some blocks of the logic circuit
equipped with the feature of the present invention for adding this to the coefficients of the digital
filter from time to time, FIG. 24 shows a continuous sum and sometimes generates it of the
digital filter FIG. 25 is a schematic diagram showing a part of another logic circuit having the
feature of the present invention for adding coefficients, and FIG. 25 is compared with the
frequency characteristic (broken line) of the output of the hearing aid with feedback cancellation
function (KHz) vs. decibel (dB) curve showing the frequency vector (solid line) of the output of
the hearing aid without feedback cancellation function, FIG. 27 is a block diagram of an
electronic filter having the features of the present invention of FIG. FIG. 28 is a block diagram of
the logic circuit and FIG. 28 is a pro-sef flow chart showing a method having the features of the
present invention for operating the electronic filter and the hearing aid of the present invention.
11 иии Hearing aid, 13 и и и Microphone means, 15 и и и Ear-hook unit, 17 и и и Conversion means, 109
и и и 11th filter means, 111 и и и First combining means, 117 ... second filter means, 107 ... second
combining means, 113 ... circuit means. Patent issued by Central Intuition, The Deaf, Rihito
Shimizu, and others by two others a>, no change to the book t ?] Flo, 7 FIG, 9 FI G, l ? FIG, 16 Fl
(3, 18 F l (3, 18 F l) (3, 24 F IG, 81 Flo, 25--'
Без категории
Размер файла
17 Кб
Пожаловаться на содержимое документа