# Pulse modulator and pulse modulation method

FIELD: pulse modulators.

SUBSTANCE: proposed pulse modulator has subtraction stage that generates regulation error signal and depends for its operation on difference between input and feedback signals, and also signal conversion stage that functions to convert regulation error signal into control signal. The latter is multiplied by complex mixed signal at operating frequency ω_{0} in first multiplication stage. At least one of acting and imaginary parts shifted to signal higher frequency band is then quantized by quantization sage thereby generating actual pulse signal, This signal is then used to produce feedback signal for subtraction stage residing in feedback unit. Such pulse modulator provides for flexible regulation of quantization noise.

EFFECT: ability of shifting suppressed quantization noise toward desired operating frequency.

21 cl, 9 dwg

The invention relates to pulse modulators to convert a complex input signal into a pulse signal and method of pulse modulation of the complex input signal.

To convert a digital input signal into an analog signal, you can use digital to analogue Converter. However, these units are quite expensive and have a larger power consumption. Often requires significant power. Another disadvantage of the analog-to-digital converters is that they are difficult to integrate into a digital electrical devices and limit the miniaturization of the latter.

Consequently, in most applications, analog-to-digital converters replace the digital pulse modulators, such as Sigma-Delta converters. Part of the traditional Sigma-Delta modulator includes an integrator that integrates the difference signal between the input signal and the quantized pulsed feedback signal, and a quantizer, which quantum integrated signal. The quantized pulse signal can then be allocated to the output of the quantizer and returned as a feedback signal to the input of the Sigma-Delta Converter. Sigma-Delta modulators differ in typical noise characteristics, and the quantization noise is shifted from region n is skih frequencies, lying near ω=0, in the area of higher frequencies. The noise in the field of higher frequencies can then be suppressed by using a low pass filter. Sigma-Delta converters can be applied in the case of very modest financial capability, and can also be integrated with the digital electrical devices. However, for some applications, justified the embodiments preserving quantization noise in the region of higher frequencies.

One of the objectives of this invention is to provide a pulse modulator, and methods of pulse modulation, which would provide a flexible regulation of the spectral distribution of the quantization noise.

This object of the invention is solved by a pulse modulator for converting complex input signal into a pulse signal as described in claim 1, schemes of excitation in clause 16, the frequency generator according to claim 19, and method of pulse modulation of the complex input signal according to item 21. Claim 31 refers to a software product for the implementation of this method.

According to this invention, the pulse modulator is used to convert a complex input signal into a pulse signal has a cascade subtraction, which generates an error signal regulation on the basis of the difference between the complex input signal with a signal about atoi communication.
Next, the pulse modulator has a cascade of signal conversion, which converts the error signal regulation in the control signal. The control signal is multiplied in the first stage of the multiplication on the complex mixed signal with frequency ω_{0}obtaining thus at least one real or imaginary part of the control signal, is shifted in the upper frequency range ω_{0}. Pulse modulator has a cascade of quantization, in which the quantization of at least one real or imaginary part of the control signal, is shifted in the upper frequency range ω_{0}, obtaining thus a pulse signal; a pulse modulator also includes a block feedback, which uses a pulsed signal to obtain a feedback signal for cascade subtraction.

The following is a description of the method of operation of the pulse modulator according to this invention represents an improved version of the conventional Sigma-Delta Converter, for example embodiments with a constant input signal, which, however, does not restrict other uses of the invention. Cascade subtraction and conversion stage converts this input signal to the control with the persecuted,
which is exactly the same only slightly varies in time. Unlike conventional Sigma-Delta converters this control signal is multiplied by the first cascade multiplication on the complex mixed signal with frequency ω_{0}obtaining thus the control signal, is shifted in the upper frequency range with frequency ω_{0}. Real or imaginary part of this control signal with frequency ω_{0}then quantized cascade quantization, and thus at the output of the cascade of quantization obtained a valid pulse signal with a dominant frequency ω_{0}. This is valid pulse signal, together with a positive or negative pulses simulates a sinusoidal signal with frequency ω_{0}. A pulse signal at the same time represents the starting point for the calculation of the feedback signal, which is then returned to the cascade subtraction, where it is subtracted from the input signal to determine the error control.

To obtain a pulse signal does not necessarily count as a real and imaginary part of the control signal, is shifted in the upper frequency range ω_{0}. You can optionally display a pulse signal of the real part is displaced in the upper frequency range opravlyaushi what about the signal,
it is not necessary to create imaginary part of the mixed control signal.

The main advantage of the pulse modulator of the present invention compared with the conventional Sigma-Delta modulators is that the range of low quantization noise is shifted from the low-frequency range ω=0 to the operating frequency ω_{0}. This is achieved through a complex mixing control signal in the first stage of multiplication. Thus, generates a pulse signal with a low noise level adjacent to ω_{0}spectral range.

The starting point in the consideration of the noise characteristics is that the cascade of signal conversion, which may, for example, be implemented on the basis of the integrator, has a high transmittance at the lower frequencies. This means that relatively high-frequency components are suppressed by the cascade of signal conversion. In traditional Sigma-Delta converters is the suppression of high-frequency components in the control loop leads to an increase in quantization noise on the above-mentioned higher frequencies. And low-frequency noise, by contrast, is very low. In the case of a pulse modulator, relevant to this invention, the control signal, which may be removed from the output of the cascade change the FL signal,
offset in the upper frequency range to the frequency ω_{0}by multiplying the complex mixed signal to the frequency ω_{0}. Thus, the range of low-frequency noise is also offset from the frequency ω=0 to the frequency mixing ω_{0}even if the conversion stage input is still processing unbiased to this signal. The result is a pulse signal with a low noise level in the area of ω_{0}.

The pulse modulator of the present invention is a low cost, relatively low power consumption and can be easily integrated with digital electrical devices.

Preferably the pulse modulator has a tract in-phase signal for processing of the real part of the input signal and the path of the quadrature signal for processing the imaginary part. It is also desirable that the error signal regulation, the control signal and the feedback signal was integrated signals, i.e. each of them had the real and imaginary components. To ensure compliance with the valid phase pulse signal real or imaginary part of the control signal, is shifted in the upper frequency range ω_{0}cascade subtraction, the cascade of signal conversion, the first stage of the multiplication and the unit opposite from the connection must be a complex blocks of signal processing,
in each of which there are paths in-phase signal and quadrature signal. However, to obtain a valid pulse signal after the first stage multiplication via cascade quantization requires only the real part of (or otherwise imaginary part of the output signal. Cascade quantization may be, therefore, a valid path processing. In fact, a valid pulse signal is then again converted into a complex signal feedback unit feedback. Design pulse modulator allows to synthesize an exact coincidence of the phase of a valid pulse signal reproducing harmonic oscillations at the frequency ω_{0}at low phase and amplitude noise.

According to one of preferred embodiments of the invention in the composition of the conversion stage signal includes a cascade integrator, in which the integration of the error signal regulation and the integrated output signal as the control. The integration of the error signal control allows you to continuously synchronize (integrated) integrated signal with a complex input signal. Since the cascade integrator has a characteristic similar to the low-pass filter, the control signal at the output of the cascade in which ugratara has reduced the noise level in the area of ω
_{0}. And by moving this control signal, the first cascade multiplication with subsequent quantization receive a pulse signal with the required noise performance.

It is advisable to have in the cascade integrator of the two integrators: a first integrator for tract in-phase signal and a second integrator for tract quadrature signal; a first integrator works with real part of the error signal regulation, and the second integrator, respectively, with the imaginary. The complex cascade of integrator complex error signal of the regulation may be thus implemented using two separate integrators.

It is also desirable that the composition of the cascade of signal conversion consisted of the amplification stage. The gain in this case is selected so as to match the input level of the quantizer.

In a more preferred embodiment of the invention the first cascade multiplication has two multipliers, one for tract in-phase signal and a second path of the quadrature signal. The first multiplier multiplies the real part of the control signal on the real part of a complex mixed signal with frequency ω_{0}obtaining thus the first result signal. The second multiplier multiplies the imaginary frequent the control signal to the imaginary part of a complex mixed signal with frequency ω
_{0}obtaining thus the second resulting signal. According to another preferred variant of the invention, the pulse modulator integrated adder, folding the first result signal from the first multiplier and the second result signal from the second multiplier to obtain the total signal, which is necessary to determine the real part of the mixed control signal.

If we assume that the integrated control signal has the form R+j·I and as an example of what a comprehensive converted signal can be represented by the formulathe first result signal from the first multiplier has the form R·cos(ω_{0}t). Suppose that the second result signal from the second multiplier has the form I·sin(ω_{0}t), and the adder gives the total signal R·cos(ω_{0}t)+I·sin(ω_{0}t). However, this signal corresponds exactly to the real part (R+j· (I)·Thus, the real part of a complex multiplication of the control signal and the mixed signal can be defined by the first multiplier, the second multiplier and adder.

According to one of preferred embodiments of the invention, the total signal produced by the adder, and the ZAT is quantized cascade quantization to obtain thus a valid pulse signal.

In this case it is convenient to add noise to the input signal cascade quantization. Pulse modulator delivers clock pulses at the sampling frequency ω_{A}that should be significantly higher frequency mixing ω_{0}. When a certain ratio ω_{0}it ω_{A}in the pulse modulator arise relaxation oscillations that are visible in the spectrum of the frequency pulse signal as an additional peaks. Since the noise signal is added to the input signal of the quantizer, the result of the quantization process is statistically rounded. This solution helps to prevent the formation of the relaxation oscillations.

Preferably, the cascade of quantization performed the quantization of the corresponding input signal on two or three variables. In the case of quantization by two variables can be assumed that the pulse signal can have only values 0 or 1. Thus is produced a pulse signal, which contains only positive voltage pulses. In the case of a pulse signal, quantized in three variables, it can have the values -1, 0, 1. Such a pulse signal includes both positive and negative voltage pulses. Thus, the quantization of the three variables is carried out in case the need to obtain a pulse with whom drove with positive and negative pulses.

Preferably the actuator feedback includes a second cascade multiplication, which performs multiplication pulse signal by a complex conjugate mixed signal having the oscillation frequency ω_{0}with education so feedback signal for vicites shifted in the lower frequency range ω_{0}. A pulse signal obtained by the quantization of the real part is displaced in the upper frequency range of the control signal, and, accordingly, the dominant frequency component is a frequency ω_{0}. Before the pulse signal can be used as the feedback signal, it must be, therefore, again shifted into low range before the bandwidth of the original signal. To accomplish this, a pulse signal is multiplied by a complex conjugate mixed signal with frequency ω_{0}getting thus shifted in the low frequency range complex feedback signal.

Preferably, the second cascade multiplication there was a third multiplier for generating a valid part of the feedback signal, and in addition, a fourth multiplier for generating the imaginary part of the feedback signal, and the third multiplier multiplies the pulse signal by the real part of a complex-conjugate mixed signal frequency is th ω
_{0}and the fourth multiplier multiplies the pulse signal on the imaginary part of the complex-conjugate mixed signal with frequency ω_{0}. To offset this frequency component of the pulse signal with the frequency ω_{0}in the right direction multiplication pulse signal in the mixed signal must be performed in an integrated form. Pulsed signal y(t) is a valid signal, whereas the complex-conjugate mixed signal can be represented in the formBy complex multiplication, thus, generated a comprehensive feedback signal with the real part of y(t)·cos(ω_{0}t) and the imaginary part y(t)·sin(ω_{0}t).

Preferably, the pulse modulator filed clock pulses at the sampling frequency ω_{A}that would be from 2 to 1000 times higher frequency mixing ω_{0}. It is necessary to meet the stability criterion Nyquist for mixed signals.

According to a more preferred variant of the invention, the pulse modulator implemented using a digital signal processor DSP (digital signal processor, DSP). All operations necessary for the operation of the pulse modulator can be programmed using conventional signal processing methods.

According to D. nomu the invention the circuit micromechanical resonator must have at least one pulse modulator of the type specified above.
A pulse signal generated by the at least one pulse modulator, preferably used for electrostatic excitation of oscillations of the resonator. The generated pulse signal can be connected directly to the electrodes of the excitation of the resonator. In this case, it is preferable that the frequency mixing ω_{0}pulse modulator corresponding to one of the resonant frequencies of the resonator because it is required for efficient excitation of the oscillator.

The frequency generator according to this invention used for the synthesis of a pulse signal at a preset frequency and phase, has at least one pulse modulator of the type specified above. The pulse modulator according to this invention can be used to generate a corresponding pulse signal y(t) at a predetermined frequency and phase. In this case, the phase angle of the generated pulse signal can be with high precision is set by setting the ratio of the real and imaginary parts of the input signal x(t). The result is a pulse signal with a low noise level in the area of ω_{0}.

According to another preferred variant of the invention, after the pulse modulator is placed bandpass fil is R,
preferably quartz or (piezo)ceramic. This is located next bandpass filter retains those frequency components that are removed from the field ω_{0}and in high noise levels.

The invention is described hereinafter in more detail by the example of the preferred embodiments with references to the drawings illustrating the described illustrative embodiments of the invention, namely:

figure 1 shows a comprehensive block diagram of the pulse modulator of the invention;

figure 2 shows a block diagram of a pulse modulator, indicating separately the in-phase and quadrature paths;

figure 3 shows the quantized three variables pulsed signal y(t);

figure 4 shows the frequency spectrum of a pulsed signal y(t)generated at the output of the quantizer;

figure 5 shows the frequency spectrum of figure 4 after filtering micromechanical oscillator;

figure 6 shows the frequency spectrum of a pulsed signal y(t), is constructed with a ratio of frequency mixing to the sampling frequency, equal to ω_{0}/ω_{A}=0.25;

7 shows a pulse modulator with statistical rounding;

on Fig shows the frequency spectrum of 6 is performed with the statistical rounding; and

figure 9 shows the block diagram womern the th pulse modulator.

figure 1 shows the integrated form of the block diagram of the pulse modulator, the relevant provisions of the present invention. The complex input signal x(t) includes the real and imaginary parts, both of which are represented in digital form. A comprehensive feedback signal 2 is subtracted from the complex input signal x(t) in the node, adding 1, and the difference between these two complex signals is determined by the error control. Furthermore, (again complex) the content of the delay element 3 is summed with the specified difference node adding 1. The contents of delay element 3 is supplied through the signal line 4 to the node adding 1. Delay element 3 together with the signal line 4 forms a complex cascade integrator, which is the integration of complex error control, that is the difference between the input signal and the feedback signal. Integrated signal 5 is amplified by a factor "a" in the amplifier 6, and the amplified signal 7 is fed to the first stage of multiplication 8, where the amplified signal 7 is multiplied by a complex mixed signalwith signal 9, which is offset in the upper frequency range with frequency ω_{0}. Unit 10 determines the real part of a complex summed signal 9, and thus obtained actually the considerable part 11 the summed signal is then available for quantizer 12.

In the embodiment of the invention, is shown in figure 1, the quantizer 12 is represented by a quantizer in three variables, which converts the corresponding input signal in three possible values -1, 0, +1 pulse signal by means of the Comparators. Thus obtained quantized pulse signal y(t) can be connected to the output of the quantizer 12. A valid pulse signal y(t) is multiplied in the second cascade multiplication 13 to complex-conjugate mixed signalto obtain comprehensive feedback signal 2. Obtained by multiplying real numbers and complex numbers the complex feedback signal 2 is supplied to the node adding 1 to the input schema.

The sequence of functional blocks, shown in figure 1, can be implemented by a DSP or hardware designed specifically for this purpose. DSPS would need to work on the sampling frequency ω_{A}that is significantly higher frequency ω_{0}integrated mixed signal. For example, the sampling frequency ω_{A}may exceed the frequency mixing ω_{0}from 2 to 1000 times.

Figure 2 again shows the pulse modulator shown in figure 1, with the separately paths in-phase signal and the quadrature signal is.
In the upper half of figure 2 shows the path 14 in-phase signal, which handles the real part R of the input signal x(t). In the lower half of figure 2 shows the path 15 quadrature signal processing imaginary part I. the Real part of the error regulation is determined in block 16 in addition, the common mode path as the difference between the real part R of the input signal and the real part 17 of the feedback signal. The integrator value is stored up to this point in the delay element 18, is folded by regulatory failure and then is transmitted through the signal line 19 to the block adding 16. Together with the signal line 19 delay element 18 forms in this operation, the integrator. The addition of the real part of the error regulation and the previous value of the integrator results in a new value integrator, which again is stored in the delay element 18. Integrated signal 20 in the path of the common mode signal damnaged on the coefficient "a" amplifier 21 and is passed as an amplified signal 22 to the first multiplier 23. The first multiplier 23 multiplies the actual amplified signal 22 for a valid signal cos(ω_{0}t), i.e. the real part ofThe first multiplier 23 determines the resulting R·cos(ω_{0}t), which gives the I signal 24 to the adder 25.

Tract 15 quadrature signal of the pulse modulator has a node adding 26 in which is calculated the difference between the imaginary part of the I input signal and the imaginary part 27 of the feedback signal. This difference corresponding to the imaginary part of the error regulation, is added to the previous value of the delay element 28, which is supplied to the power summing 26 via the signal line 29. The new value is obtained as the sum of the previous value and the imaginary part of the error control, it is stored in the delay element 28. Together with the signal line 29 delay element 28 forms in this operation, the integrator. Integrated signal 30 from the path of quadrature signal at the output of this integrator and then multiplied by the coefficient "a" of the amplifier 31. The amplified signal 32 obtained in this way in the path of the quadrature signal can then be multiplied by the signal sin(ω_{0}t) in the second multiplier 33. The resulting I·sin(ω_{0}t)obtained in this way, serves as a signal 34 to the adder. 25. The adder 25 adds the signals R·cos(ω_{0}t) and I·sin(ω_{0}t) to obtain the output signal R·cos(ω_{0}t)+I·sin(ω_{0}t) as the signal 35. However, this signal 35 corresponds exactly to the real part of the mixed signal as a complex multiplication of x(t)
and gives:

=(R+j· (I)·(cos(ω_{0}t)-j·sin(ω_{0}t))=

=[ R·cos(ω_{0}t)+I·sin(ω_{0}t)]+j·[I·cos(ω_{0}t)-R·sin(ω_{0}t)]

and the real part of this signal is equal to R·cos(ω_{0}t)+I·sin(ω_{0}t). The signal 35, thus, represents the real part of a complex mixed signal and in this sense corresponds to the signal 11 shown in figure 1.

Digital real signal 35 is fed to the quantizer 36, which converts this input signal into a quantized pulse signal y(t). Triple junction (quantization three variables) quantizer shown in the examples of figure 1 and figure 2, quantum input algorithm y(t)∈{-1; 0; +1}. For this purpose, the quantizer 36 has Comparators, in which there is a continuous comparison of the signal level 35 with preset threshold values. Depending on the results of the comparison output signal y(t) are in each case assigned to one of the values -1; 0; +1 as the value of the current signal. Instead of a triple junction (three variables) quantization depending on the application can be used with other types of quantization, for example two-stage (two variables) or multi-stage quantization.

The actual frequent is 17 and imaginary part 27 of the complex feedback signal is obtained from the quantized pulse signal y(t). This pulsed signal y(t) are multiplied by the complex conjugate mixed signal

=y(t)·cos(ω_{0}t)+j·y(t)·sin(ω_{0}t)

The real part of y(t)·cos(ω_{0}t) of the complex feedback signal is generated by a third multiplier 37, performs multiplication of the pulse signal y(t) cos(ω_{0}t). The real part 17 of the feedback signal thus generated as an output signal of the third multiplier 37 and returns to block adding 16. To generate the real part of y(t)·sin(ω_{0}t) of the complex feedback signal is a pulse signal y(t) is multiplied by sin(ω_{0}t) in the fourth multiplier 38. Imaginary part 27 of the feedback signal thus generated as an output signal of the fourth multiplier 38 and returns to block adding 26.

In the example embodiments of the invention, shown in figures 1 and 2, on the side of the entrance there are integrators that integrate the regulatory failure between the input signal and the feedback signal, thus generating an integrated signal. The transformation function H(z) integrator can be written as. On the input side instead of the integrators are also allowed to use the other cascades of signal conversion with other conversion functions H(z). For example, a high-level conversion function H(z) can be used if:

lim H(z)=∞

z→1.

The transformation function H(z), thus, tends to infinity, if the frequency ω tends to zero (z→1). Additional free parameters H(z) can be used to optimize certain characteristics of the modulator (for example, the ratio signal/noise) or the entire system.

Figure 3 shows the shape of the pulse signal y(t), which can be obtained at the output of the quantizer in the case of quantization in three variables with y(t)∈{-1; 0; +1}, obtained by simulation. The value of the real part R of the complex input signal was set equal to 0.3, whereas the value of the imaginary part of I took zero. The input signal x(t) thus was constant and was not a function of time. The sampling frequency ω_{A}five times higher than the frequency mixing, i.e. ω_{0}/ω_{A}=0.2. The clock pulses at the sampling frequency ω_{A}pending the abscissa and sequentially numbered from 5000 to 5100. During each beat of the pulse signal y(t) takes one of three possible values: -1; 0; +1. The corresponding value of y(t) separately for each stroke when the sampling frequency is deposited on the ordinate axis.

The results of spectral analysis FFT) of the pulse signal in figure 3 is shown in figure 4.
The frequency of the corresponding spectral components shown in arbitrary units FFT along the x-axis, and the intensity of the signal in dB is plotted on the ordinate axis. In the spectral distribution of visible peaks on the frequency ω_{0}. It is also seen that the noise level in the area adjacent to the frequency ω_{0}significantly lower than in the rest of the spectrum. Unlike traditional Sigma-Delta modulator noise is significantly reduced it in the low frequency range, i.e. in the area adjacent to the frequency ω_{0}. In the case of the pulse modulator, the relevant provisions of the present invention, integrated and amplified signal is shifted in frequency mixing ω_{0}by complex multiplication. As a consequence, the spectral range, which is the most effective noise reduction, is also displaced in the direction of the frequency mixing ω_{0}obtaining the noise characteristics shown in figure 4.

The pulse modulator, the appropriate provisions of this invention, can be used to digitally generate a pulse signal, and in this case, the main spectral component of the pulse signal can be set through frequency mixing ω_{0}. The phase angle of the generated pulse signal can be precisely set through the your installation ratio of the real and imaginary parts of the input signal,
thus ensuring the stability of the phase pulse signal. When using the pulse modulator corresponding to the conditions of this invention for the synthesis frequency pulse signal y(t) should pass through an electrical band-pass filter, the middle of the bandwidth which is at a frequency of ω_{0}. This band-pass filter, implemented, for example, as quartz band-pass filter or (piezo)ceramic filter, enables to suppress spectral ranges, remote from ω_{0}that is quite a high noise level. Such band-pass filter also allows to significantly improve the ratio of signal to noise.

The pulse modulator, the appropriate provisions of this invention are suitable, inter alia, for the excitation of the Electromechanical oscillators to generate harmonic oscillations. In particular, the electrostatic force required for the excitation of oscillations can be generated by quantized in three variable pulse signal applied to the excitation electrodes of the micromechanical resonator. Frequency ω_{0}pulsed signal y(t) in this case, it is preferable to choose equal to the resonant frequency of the micromechanical oscillator. If a pulse signal similar to that shown in figure 3 and figure 4, it is used for the harmonic excitation of the oscillator with a high q-factor (e.g.,
factor q-factor equal to 10^{4}), the resonance frequency which corresponds to the excitation frequency ω_{0}then the main part of the quantization noise will be filtered out by the oscillator. In particular, the quantization noise in the spectral ranges remote from the tuning frequency ω_{0}suppressed by the oscillator. Thus obtained filtered spectrum is shown in figure 5.

There are some relationships frequencies ω_{0}/ω_{A}for which pseudonoise product quantization in y(t) is converted into a number of more or less periodic functions. As an example, figure 6 shows the frequency spectrum obtained for the relationship ω_{0}/ω_{A}=0,25. In addition to the peak at the frequency ω_{0}see the range of spectral lines 39, 40, 41, etc. Cause the formation of these spectral lines is that the quantizer is a highly nonlinear element of the control loop, and thus excited relaxation oscillations in the control loop under certain relations frequencies. The response of the control loop is well known for traditional Delta-Sigma converters.

To prevent the relaxation oscillations can be improved Central linearity of the quantizer by adding the noise signal to the input signal Kwan is the user.
For this task it is preferable to use spectral uniformly distributed noise signal. 7 shows a block diagram corresponding to an improved pulse modulator. Compared with the block diagram in figure 2, the pulse modulator, shown in Fig.7, further comprises a noise generator 42 42 that generates a noise signal 43. In addition, the integrators shown in figure 2, the diagram shows in General form as cascades 44, 45 conversion signal conversion function H(z). In the rest of the nodes shown in Fig.7 correspond to the elements of the block diagram of figure 2. The noise signal 43 is supplied to the adder 25, where it is the addition of the signals 24 and 34. Consequently, the signal 35 at the input of the quantizer 36 is superimposed noise signal, which eventually leads to static rounding in the quantization process. On Fig shows the frequency spectrum of a pulsed signal y(t)obtained using improved pulse modulator shown in Fig.7. Although the ratio of the frequencies ω_{0}/ω_{A}also equal to 0.25, relaxation oscillations are not formed.

The pulse modulator, the appropriate provisions of the present invention may be used, in particular, to electrostatic excitation of micromechanical oscillators. To do this, as primerano electrodes excitation of micromechanical resonator may be filed quantized in three variables a pulse signal indicated on figure 3 type.
A pulse signal shown in figure 3, is a sinusoidal signal at the frequency ω_{0}. Such a pulse signal can be used for excitation of the harmonic oscillation of the micromechanical resonator at the frequency ω_{0}more specifically , in the case when the frequency ω_{0}pulse signal at least approximately corresponds to the resonant frequency of the oscillator.

Resonators, capable of producing vibrations in two mutually perpendicular directions y_{1}and y_{2}used speed sensors and gyroscopes. Shown in Fig.9 two-dimensional pulse modulator can be used for electrostatic excitation of the resonator with two degrees of freedom. Two-dimensional pulse modulator contains a first pulse modulator 46, which generates a pulse signal y_{1}(t) on the basis of the complex input signal R_{1}, I_{1}when this received pulse signal is used to excite the resonator in the direction of y_{1}. A pulse signal y_{2}(t) is generated based on the complex input signal R_{2}, I_{2}the second pulse modulator 47 and is used to excite the resonator in the direction of y_{2}. As the first pulse modulator 46 and the second pulse modulator 47 represent the pulse m is gulatory with statistical rounding,
similar to those shown in Fig.7. The description of the design and methods of work of the first and second pulse modulator 46, 47 can therefore be found in the description of figure 2 and 7. However, two-dimensional pulse modulator, shown in figure 9, has one two-dimensional quantizer 48, which is divided into two channels and which converts the signal 49 of the first pulse modulator 46 in the quantized pulse signal y_{1}(t)and an output 50 of the second pulse modulator 47 - quantized pulse signal y_{2}(t). The use of two-dimensional quantizer 48, is divided into two channels, allows quantization of signals 49, 50 to consider additional conditions, which is an important advantage when operating a micromechanical sensors. For example, one such additional terms is that in each case only one channel can generate pulses different from zero. Another possible condition is that the output signals y_{1}(t), y_{2}(t) may at any time be changed. Such additional conditions may be of real value if the bias currents supplied to the electrodes of the double resonator, measured in total that allows you to deduct the deviation of the oscillator. Additional conditions allow one to associate a bias current with a separate electrode. Thus, poaul who is able to perform the decomposition of the signal between the signals
due to the deviation of the oscillator in the direction of y_{1}and deviation in the direction of y_{2}.

1. A pulse modulator for converting complex input signal (x(t)) into a pulse signal (y(t)), characterized in that it contains

cascade subtracting (1)generating an error signal regulation on the basis of the difference between the complex input signal (x(t)) and the feedback signal (2),

one conversion stage that converts the error signal regulation in the control signal (7);

the first stage of multiplication (8), which is the multiplication of the control signal (7) for complex mixed signal having the oscillation frequency ω_{0}, obtaining thus at least one of the valid (11) and imaginary parts of the control signal, is shifted in the upper frequency range ω_{0};

cascade quantization (12)where is the quantization of at least one of the real and imaginary parts of the control signal, is shifted in the upper frequency range ω_{0}, obtaining thus a pulse signal (y(t));

block feedback, which is based on the pulse signal (y(t)) generates a signal (2) feedback for cascade subtraction.

2. The pulse modulator according to claim 1, characterized in that it has a path in phase C is Nala for processing the real part of the input signal, and also the path of the quadrature signal for processing the imaginary part of the input signal.

3. The pulse modulator according to claim 1, characterized in that the error signal regulation, the control signal and the feedback signal are complex signals, that is, each has a real and imaginary component of the signal.

4. The pulse modulator according to claim 1, characterized in that the cascade of signal conversion includes cascade integrator, which is the integration of the error signal regulation and generating an integrated signal as the control signal.

5. The pulse modulator according to claim 4, characterized in that the cascade integrator includes a first integrator for system (14) is in-phase signal and a second integrator for system (15) of the quadrature signal, and the first integrator processes the real part of the error signal regulation, and the second the imaginary part.

6. The pulse modulator according to any one of claims 1 to 5, characterized in that the cascade of signal conversion includes cascade (6) gain.

7. The pulse modulator according to any one of claims 1 to 5, characterized in that the first cascade multiplication contains the first multiplier (23) for path in-phase signal and the second multiplier (33) for tract quadrature signal, and the first multiplier handles the real part (22) of the control signal, cleverly who nd it on the real part of a complex mixed signal,
having the oscillation frequency ω_{0}, obtaining thus the first result signal (24)and the second multiplier (33) multiplies the imaginary part (32) of the control signal on the imaginary part of a complex mixed signal having the oscillation frequency ω_{0}, obtaining thus the second resulting signal (34).

8. The pulse modulator according to claim 7, characterized in that it contains an adder (25), which adds the first result signal (24) from the first multiplier and the second result signal (34) from the second multiplier to obtain the total signal (35), needed to determine the real part is displaced in the upper frequency range of the control signal.

9. The pulse modulator of claim 8, characterized in that the cascade quantization quantum the total signal generated by the adder.

10. The pulse modulator according to any one of claims 1 to 5, characterized in that the noise is added to the input signal cascade quantization.

11. The pulse modulator according to any one of claims 1 to 5, 8, 9, characterized in that the cascade performs quantization the quantization of the corresponding input signal on two or three variables.

12. The pulse modulator according to any one of claims 1 to 5, 8, 9, characterized in that the actuator feedback includes a second cascade (13) multiplication, which is multiplying the pulses of the signal by a complex conjugate mixed signal,
having the oscillation frequency ω_{0}, obtaining thus the signal (2) feedback, which is offset in the low frequency range to the frequency ω_{0}and supplied to myCitadel.

13. The pulse modulator according to item 12, wherein the second cascade multiplication includes a third multiplier (37) to obtain the real part (17) of the feedback signal and the fourth multiplier (38) to obtain the imaginary part (27) of the feedback signal, and a third multiplier (37) multiplies the pulse signal by the real part of a complex-conjugate mixed signal having the oscillation frequency ω_{0}and the fourth multiplier (38) multiplies the pulse signal on the imaginary part of the complex-conjugate mixed signal with frequency ω_{0}.

14. The pulse modulator according to any one of claims 1 to 5, 8, 9, 13, characterized in that the pulse modulator operates at the sampling frequency ω_{A}that at 2-1000 times the frequency mixing ω_{0}.

15. The pulse modulator according to any one of claims 1 to 5, 8, 9, 13, characterized in that the pulse modulator is implemented as a digital signal processor.

16. Control circuit for a micromechanical resonator having at least one pulse modulator, characterized in that it includes a pulse modulator according to any one of claims 1 to 15.

17. The circuit panel is the effect on P16, characterized in that the pulse signal generated by the at least one pulse modulator is used for electrostatic excitation of oscillations of the resonator.

18. The control circuit according to item 16 or 17, characterized in that the frequency mixing ω_{0}pulse modulator corresponds exactly to one resonance frequency of the resonator.

19. Frequency generator for synthesizing a pulse signal at a preset frequency and phase with at least one pulse modulator, wherein the pulse modulator is a modulator according to any one of items 1 to 15.

20. The frequency generator according to claim 19, characterized in that after the pulse modulator is selected bandpass filter, preferably quartz or (piezo)ceramic.

21. The method of pulse modulation of the complex input signal, characterized in that it comprises the following operations:

generating the error signal regulation on the basis of the difference between the complex input signal (x(t)) and signal (2) feedback;

converting the error signal regulation in the control signal (7);

the multiplication of the control signal (7) for complex mixed signal having the oscillation frequency ω_{0}, obtaining thus at least one of the valid (11) or imaginary is part of the control signal,
offset in the upper frequency range ω_{0};

quantization of at least one of the valid (11) and imaginary parts of the control signal, is shifted in the upper frequency range ω_{0}with the receipt of a pulse signal (y(t));

generating a feedback signal (2) based on the pulse signal (y(t)).

22. The method according to item 21, wherein the error signal regulation, the control signal and the feedback signal are complex signals, that is, each has a real and imaginary component of the signal.

23. The method according to item 21 or 22, characterized in that the error signal regulation is converted into a control signal by integrating the error signal regulation.

24. The method according to item 21 or 22, characterized in that the real part of the control signal is multiplied by the real part of a complex mixed signal with frequency ω_{0}education is thus the first result signal, and the imaginary part of the control signal is multiplied by the imaginary part of a complex mixed signal with frequency ω_{0}with the establishment of the second resulting signal.

25. The method according to paragraph 24, wherein the first and second resultant signals are added with the formation of the total signal to determine the actual customising in the upper frequency range of the control signal.

26. The method according A.25, characterized in that the total signal quantuum with the receipt of pulse signal.

27. The method according to any of p, 22, 25 or 26, characterized in that the addition of noise produced before quantization at least one of the real and imaginary parts of the control signal, is shifted in the upper frequency range ω_{0}.

28. The method according to any of p, 22, 25 or 26, characterized in that the feedback signal is generated by multiplying the pulse signal by a complex conjugate mixed signal with frequency ω_{0}.

29. The method according to any of p, 22, 25 or 26, characterized in that a pulse signal is used for electrostatic excitation oscillation of the micromechanical resonator.

30. The method according to clause 29, wherein the frequency mixer ω_{0}exactly matches the frequency of the micromechanical resonator.

31. The method according to any of p, 22, 25, 26, or 30, characterized in that the operation of the method performed on a computer, digital signal processor or similar device.

**Same patents:**

FIELD: radio transmitters, possible use in wireless data transmission terminals, for example, cell phones.

SUBSTANCE: in the method of operation of two-mode radio transmitter with multiple channel intervals, a set of control signals is set for radio transmitter in accordance to second modulation format used during second channel interval, while one of control signals sets the operation mode of power amplifier. Two-mode radio transmitter with multiple channel intervals contains programmable power amplifier, control block with multiple channel intervals which outputs control signals to programmable power amplifier in accordance to first and second modulation formats. Computer software product, realized in computer-readable information carrier, provides for execution of all necessary operations in two-mode transmitter with multiple channel intervals.

EFFECT: linear alternation of output power between neighboring channel intervals in accordance to modulation type of the next channel interval.

6 cl, 7 dwg

FIELD: pulse modulators.

SUBSTANCE: proposed pulse modulator has subtraction stage that generates regulation error signal and depends for its operation on difference between input and feedback signals, and also signal conversion stage that functions to convert regulation error signal into control signal. The latter is multiplied by complex mixed signal at operating frequency ω_{0} in first multiplication stage. At least one of acting and imaginary parts shifted to signal higher frequency band is then quantized by quantization sage thereby generating actual pulse signal, This signal is then used to produce feedback signal for subtraction stage residing in feedback unit. Such pulse modulator provides for flexible regulation of quantization noise.

EFFECT: ability of shifting suppressed quantization noise toward desired operating frequency.

21 cl, 9 dwg

FIELD: technology for transmitting information across a distance, possible use in systems for wired and wireless communications, for encoding and decoding information.

SUBSTANCE: aforementioned method for transmitting and receiving information and devices are based on modulation of carriers with spatial-frequency modes of oscillations of substance molecules and demodulation, based on extraction of aforementioned modes of oscillations by means of registration of galvanic-magnetic effects, for example, Hall effect.

EFFECT: possible transmission of both digital and analog information and possible adjustment of receiver for type of substance used in transmitter due to usage of type of modulation and demodulation of electromagnetic waves or alternating current.

2 cl, 3 dwg

FIELD: electricity.

SUBSTANCE: microwave modulator includes extended structure on the basis of magnon crystal from ferrite film, microstrip converters for excitation and reception in ferrite film of surface magnetostatic waves, which are arranged on opposite ends of structure, constant magnetic field source, element for control of distribution parameters of magnetostatic waves. Magnon crystal is made on the basis of film from yttrium iron garnet with surface periodic structure in the form of parallel grooves the depth of which is 0.01-0.2 of film and which are arranged perpendicular to axis of extended structure. Element for control of distribution parameters of magnetostatic waves is arranged on the side of grooves and represents metallic shield movable in the direction perpendicular to plane of periodic structure and connected to piezoelectric actuator connected to electric generator of modulating signal.

EFFECT: providing the possibility of controlling the rejection level of microwave signal in frequency band without control direct current flowing along metal film.

4 cl, 4 dwg

FIELD: radio engineering, communication.

SUBSTANCE: Mobius surface of the multilayer metal-dielectric-metal-p-i-n-diodes-metal-dielectric-metal is used according to the invention in the code-switching modulator of microwave oscillations, and in each section of such surface there is a capacitive component formed by the rectangular dielectric on two opposite sides coated with the metal plates. It is possible to connect or disconnect the additional plates through the p-i-n-diodes, so that the capacity is discretely increased or decreased (herewith the inductance will remain unchanged). The closure in the form of the one-sided Mobius surface of the two metal plates forms one metal plate in the form of the two short-circuited inductor turns.

EFFECT: changing the resonance characteristics of the travelling and standing wave mode by connecting or disconnecting the additional plates to the container in the modulator using the p-i-n-diodes.

3 dwg

FIELD: pulse modulators.

SUBSTANCE: proposed pulse modulator has subtraction stage that generates regulation error signal and depends for its operation on difference between input and feedback signals, and also signal conversion stage that functions to convert regulation error signal into control signal. The latter is multiplied by complex mixed signal at operating frequency ω_{0} in first multiplication stage. At least one of acting and imaginary parts shifted to signal higher frequency band is then quantized by quantization sage thereby generating actual pulse signal, This signal is then used to produce feedback signal for subtraction stage residing in feedback unit. Such pulse modulator provides for flexible regulation of quantization noise.

EFFECT: ability of shifting suppressed quantization noise toward desired operating frequency.

21 cl, 9 dwg

FIELD: pulse modulators.

_{0} in first multiplication stage. At least one of acting and imaginary parts shifted to signal higher frequency band is then quantized by quantization sage thereby generating actual pulse signal, This signal is then used to produce feedback signal for subtraction stage residing in feedback unit. Such pulse modulator provides for flexible regulation of quantization noise.

EFFECT: ability of shifting suppressed quantization noise toward desired operating frequency.

21 cl, 9 dwg

FIELD: information technology.

SUBSTANCE: codec (30) includes at least one coder (10) and at least one decoder (20). Encoder includes data processing circuit for application to input data (D1) of one of forms of differential and/or summing coding to form one or more corresponding coded sequences, which is subjected to cyclic shift relative to maximum value and/or cyclic shift relative to minimum value to generate encoded output data (D2 or D3). Decoder includes data processing circuit for processing one or more parts of encoded data (D2 or D3) configured to use one of difference and/or summing decoding types to one or more corresponding coded sequences specified one or more parts, wherein one or more encoded sequences are subjected to cyclic transition operation relative to maximum value and/or cyclic transition relative to minimum value for formation of decoded output data (D5).

EFFECT: high degree of data compression.

44 cl, 3 dwg, 2 tbl

FIELD: radio transmitters, possible use in wireless data transmission terminals, for example, cell phones.

SUBSTANCE: in the method of operation of two-mode radio transmitter with multiple channel intervals, a set of control signals is set for radio transmitter in accordance to second modulation format used during second channel interval, while one of control signals sets the operation mode of power amplifier. Two-mode radio transmitter with multiple channel intervals contains programmable power amplifier, control block with multiple channel intervals which outputs control signals to programmable power amplifier in accordance to first and second modulation formats. Computer software product, realized in computer-readable information carrier, provides for execution of all necessary operations in two-mode transmitter with multiple channel intervals.

EFFECT: linear alternation of output power between neighboring channel intervals in accordance to modulation type of the next channel interval.

6 cl, 7 dwg

FIELD: pulse modulators.

_{0} in first multiplication stage. At least one of acting and imaginary parts shifted to signal higher frequency band is then quantized by quantization sage thereby generating actual pulse signal, This signal is then used to produce feedback signal for subtraction stage residing in feedback unit. Such pulse modulator provides for flexible regulation of quantization noise.

EFFECT: ability of shifting suppressed quantization noise toward desired operating frequency.

21 cl, 9 dwg

FIELD: combinations of signals for wired or wireless communication with several carriers, such as OFDM or MC-CDMA systems.

SUBSTANCE: combination of signals is provided in form of points, set apart from each other for distance of maximized minimal difference between distributions of conditional probabilities, such as Kullback-Leibler divergence. Preferably, combination points are positioned in concentric circles with or without a point in the coordinate origin, while adjacent circles are turned to maximize angular distance between points on adjacent circles. Pilot-signal symbols, inserted into signal being transmitted, are used by receiving device to evaluate channel of system with several carriers. Different combinations of signals ensure optimal communication productivity for various states of communication channel, especially in environment with fast fading. For estimating a channel, signal-noise ratio is used, determined on different number of branches. Number of used channel branches may be less than number of pilot-signals per OFDM or MC-CDMA symbol.

EFFECT: increased efficiency.

4 cl, 34 dwg

FIELD: technology for transmitting and receiving digital information.

SUBSTANCE: in accordance to the method during transmission a set of data bits, meant for transmission, is generated in a set of protocol data blocks, header bits are inserted into distributed positions within limits of protocol data blocks, where header bits are shifted relatively to one another, and protocol data blocks are used to modulate a set of carriers, generating the output signal. During receipt of digital signal of audio radio broadcasting, protocol data blocks are demultiplexed and transformed to audio signal of main program service and integrated signal of data transmission service, which are transformed to outgoing audio signal and outgoing data signal.

EFFECT: increased dynamic range in digital audio radio broadcasting system, and reduced noise in transmitted signal.

4 cl, 18 dwg, 2 tbl

FIELD: information technologies.

SUBSTANCE: when subbands are used, M of used subbands are initially ordered to form multiple subband group where each of subband groups contains different subset of used subbands. Each of T transmitting antennas is associated with one or more subband groups for transmitting pilot-signal and usually with one subband group for data transmission. Pilot-signal and data may be transmitted by each of antennas in relevant subbands. For each transmitting antenna its transmitting power in each assigned to it subband can be proportionally higher so that full power available for the antenna is for transmission. Pilot-signal and/or data can be transmitted by all T antennas in all used subbands concurrently without any interference.

EFFECT: elimination of interference caused by concurrent signal transmission by multiple antennas.

32 cl, 13 dwg

FIELD: radio engineering, communication.

SUBSTANCE: method for iterative detection and decoding of a signal in communication systems with a MIMO channel includes two steps: at the first step, generating a Minimum Mean Square Error (MMSE) estimate of a transmitted symbol using a linear MMSE filter; at the second step, via linear transformation of each n-th (where n=1, 2, (N)) component

EFFECT: high efficiency and quality of data transmission.

5 cl, 6 dwg