JP6396158B2 - Frequency information detection device and high frequency power supply device - Google Patents

Description

  The present invention relates to a frequency information detection device for detecting frequency information of an analog sine wave signal and a high frequency power supply device using the frequency information detection device.

  The frequency information detection apparatus can be used in a high frequency power source that supplies power to a load such as a plasma processing apparatus that performs plasma etching and plasma CVD, for example.

FIG. 5 is a schematic configuration example of a high-frequency power supply system.
The first high-frequency power source 110 and the second high-frequency power source 120 amplify a high-frequency signal output from the common oscillator 100 by a traveling wave power output unit (not shown) and output high-frequency power. The high-frequency power output from the first high-frequency power source 110 and the second high-frequency power source 120 is supplied to two locations of the electrodes (131, 132) in the plasma processing apparatus 130 serving as a load. 5 shows an example in which high-frequency power is supplied to each of the electrode 131 and the electrode 132. For example, as shown in FIG. 6, high-frequency power is supplied to two locations of the upper electrode 131. Is also possible.

  The high frequency power directed from the first high frequency power supply 110 and the second high frequency power supply 120 to the plasma processing apparatus 130 is referred to as traveling wave power, and the high frequency power reflected by the plasma processing apparatus 130 and returning to the high frequency power supply side is reflected wave power. That's it.

  In the plasma processing apparatus 130, the plasma 133 is generated using the high-frequency power supplied from the first high-frequency power source 110 and the second high-frequency power source 120, and etching is performed.

  An impedance matching device that matches the impedances of the first high frequency power supply 110 and the plasma processing apparatus 130 may be used between the first high frequency power supply 110 and the plasma processing apparatus 130. Similarly, an impedance matching device may be used between the second high-frequency power source 120 and the plasma processing apparatus 130.

  In the high frequency power supply system of FIGS. 5 and 6, the first high frequency power supply 110 and the second high frequency power supply 120 are controlled by separate control circuits. If the control circuits are different, the output frequencies of the first high frequency power supply 110 and the second high frequency power supply 120 may be slightly different even if a high frequency signal is given from the common oscillator 100. As this cause, for example, an error due to a difference in the control clock may be considered. In particular, if the first high-frequency power source 110 and the second high-frequency power source 120 are from different manufacturers, a slight difference may occur.

In the high frequency power supply system as shown in FIGS. 5 and 6, when the output frequencies of the two high frequency power supplies 110 and 120 are slightly different, the electrodes (131, 132) are supplied with high frequency power having slightly different frequencies. Mixed. That is, since the frequencies of the two types of high-frequency voltages applied to the electrodes (131, 132) are slightly different, the amplitude of the voltage waveform generated by overlapping the two types of high-frequency voltages is periodically changed as if modulated. fluctuate. For example, if the frequencies of two types of high-frequency voltages differ by 1 Hz in the vicinity of 13.56 MHz, the amplitude varies with a period of about 1 s. As a result, the potential of the plasma generated in the plasma processing apparatus 130 also varies periodically, which has an adverse effect.
For example, when the peak of one high-frequency voltage waveform and the peak of the other high-frequency voltage waveform are in the same phase, the waveform of the overlapping voltage is the maximum value, but the peak of one high-frequency voltage waveform and the valley of the other high-frequency voltage waveform When the two have the same phase, the waveform of the voltage generated by overlapping becomes the minimum value. Since the difference between the maximum value and the minimum value is large, a waveform of a voltage generated by overlapping two types of high-frequency voltages periodically varies with a large change in amplitude.

  Therefore, the output frequencies of the first high frequency power supply 110 and the second high frequency power supply 120 need to be the same. For this purpose, the above problem can be solved by detecting the output frequency of one high-frequency power supply and matching the output frequency of the other high-frequency power supply with the detected frequency. Therefore, also in the field of the high frequency power supply system, a frequency information detection device for detecting the frequency is necessary. Moreover, the use of monitoring the output frequency of a high frequency power supply etc. can be considered.

JP 63-66472 A

  Conventionally, as a frequency information detection device in a high frequency region, for example, a device as disclosed in Patent Document 1 (Japanese Patent Laid-Open No. 63-66472) has been proposed. However, in this patent document 1, in order to detect a frequency, a differentiation circuit, a phase shift circuit, an integration circuit, etc. were required, and the calculation was complicated. In addition, a lookup table for storing the correspondence relationship with the frequency in advance is necessary, and it is necessary to prepare in advance.

  An object of this invention is to provide the frequency information detection apparatus which can detect a frequency more easily than before.

The frequency information detection apparatus provided by the first invention is
A frequency information detecting apparatus for detecting the frequency information of the analog sine wave signal V _in,
A sine wave signal V _in analog, (the reciprocal of the sampling frequency f _s: 1 / f _s) a predetermined sampling period by the instantaneous value V of the input signal V _in at the time [t] obtained by converting a digital signal an A / D converter that converts _in [t] sequentially into a sine wave signal composed of a plurality of sampling data;
Time cosine value obtained by differentiating the instantaneous value _V in [t] of the input signal _{V in} in [t] (cos (α [ t])) an estimate of the instantaneous value of the input signal _{V in} at the time [t-1] A cosine value estimator that uses V _in [t−1], an instantaneous value V _in [t + 1] of the input signal V _in at time [t + 1], and a known value;
Time the instantaneous value _V in [t] of the input signal _{V in} at the [t] and a sine component (sin component), the cosine value at time [t] estimated by the cosine value estimating section (cos (alpha [t]) ) As a cosine element (cos element), and a phase estimation unit that calculates a phase α [t] of a cosine value (cos (α [t])) at time [t] using an arctangent function (tan ⁻¹ ) ,
Based on the phase α [t−1] at time [t−1] and the phase α [t] at time [t] calculated by the phase estimation unit, the phase displacement amount Δα [t] at time [t] is calculated. A phase displacement amount calculation unit to be calculated;
A frequency estimation unit for calculating the estimated frequency information of the input signal V _in in on the basis of the phase shift amount [Delta] [alpha] [t] in the calculated time in the phase shift amount calculation unit [t], the time [t],
A moving average unit calculates the moving average of the estimated frequency information of the input signal V _in at the time [t] calculated by the frequency estimation unit, and outputs the estimated frequency moving average information,
Frequency information obtained by adding the correction frequency information to the set frequency information of the input signal V _in when the synchronizing frequency information, calculates the error of the synchronizing frequency information for the estimated frequency moving average value information at the time [t], the error An error frequency information output unit for outputting as frequency information;
A frequency correction unit that determines and outputs the correction frequency information so as to reduce the error frequency information;
It has.

The frequency information detection apparatus provided by the second invention is
The cosine value estimation unit calculates a cosine value (cos (α [t])) according to the following equation, where K is a constant calculated using the synchronization frequency information.
cos (α [t]) = K · {V _in [t + 1] −V _in [t−1]}

The frequency information detection apparatus provided by the third invention is
The constant K is calculated by the following equation. However, the synchronization frequency information is “F _syc ”.
K = 1 / {2 sin (2π · (F _syc / f _s ) [t])}

The high frequency power supply device provided by the fourth invention is
The synchronization frequency information output from the frequency information detection apparatus according to any one of the first to third aspects is used as output frequency setting frequency information.

In the present invention, an estimated value of the cosine value (cos (α [t])) obtained by differentiating the instantaneous value V _in [t] of the input signal V _{in at} the time [t] necessary for calculating the frequency information is expressed as time [t-1] input signal _{V in} instantaneous value _V in [t-1] of the is calculated using the instantaneous value _V in [t-1] and the known values of the input signal _{V in} at the time [t + 1]. Thereby, an arithmetic expression can be simplified and a calculation load can be reduced.

It is a block diagram which shows the example of a detection of a high frequency voltage. 1 is a configuration example of a frequency information detection device 1. If the frequency _{f in} of the input signal _{V in} is assumed to have shifted from the set frequency _{f The set,} the moving average value _{f Dif_ave} estimate _{f dif} difference frequency output from the moving average unit 71 [t] [t] This is a simulation result. If the frequency _{f in} of the input signal _{V in} is assumed to have shifted from the set frequency _{f The set,} the moving average value _{f Dif_ave} estimate _{f dif} difference frequency output from the moving average unit 71 [t] [t] This is a simulation result. It is an example of schematic structure of a high frequency electric power supply system. It is another schematic structural example of a high frequency electric power supply system.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, the same code | symbol is attached | subjected to the same or similar structure as the past.

FIG. 1 is a block diagram illustrating an example of detecting a high-frequency voltage.
The first high-frequency power supply 110 is a device for amplifying a high-frequency signal output from the oscillator 100, outputting traveling wave power PF having an output frequency in a radio frequency band, and supplying the traveling wave power PF to a plasma processing apparatus 130 serving as a load. . The traveling wave power PF output from the high frequency power supply 110 is supplied to the plasma processing apparatus 130 via the transmission line 140. In general, this type of high-frequency power supply outputs traveling wave power PF having a frequency in a radio frequency band of several hundred kHz or more (for example, a frequency of 13 MHz, 40 MHz, etc.).

  The plasma processing apparatus 130 serving as a load is an apparatus for processing (etching, CVD, etc.) a workpiece such as a wafer and a liquid crystal substrate that is provided with a processing section and is carried into the processing section. Note that an electrode (see FIGS. 5 and 6) is provided in the processing portion. When a plasma discharge gas is introduced into the processed portion and the traveling wave power PF output from the high frequency power supply 110 is supplied to the electrode, discharge occurs and plasma is generated. The plasma processing apparatus 130 processes the workpiece using this plasma.

The directional coupler 150 is inserted between the first high frequency power supply 110 and the plasma processing apparatus 130 to detect a traveling wave voltage VF from the high frequency power supply 110 to the plasma processing apparatus 130 and detect the detected signal as a traveling wave detection. Output as signal _Vf . The directional coupler 150 also has a function of detecting the reflected wave voltage reflected by the load, but is omitted because it is not necessary in this example.

The filter 160 is a low-pass filter or a band-pass filter, removes harmonic components from the traveling wave detection signal V _f output from the directional coupler 150, and passes the fundamental frequency component. As a result, the signal output from the filter 160 becomes a sine wave signal and is sent to the frequency information detection apparatus 1. In this specification, a signal output from the filter 160 is an input signal V _in to the frequency information detection device 1.
The input signal V _in is because it is an analog voltage signal, the output frequency f, and time t, the phase offset theta, when the angular frequency is ω (= 2π · f), the input signal V _in is of the formula It can be expressed as (1). It should be noted that, here, is set to "1", the amplitude of the input signal V _in.
V _in = sin (2π · f · t + θ)
= Sin (ω · t + θ) (1)

  Further, a level conversion circuit for converting the signal level so as to be suitable for the input range of the A / D conversion unit 10 to be described later may be provided at the subsequent stage of the filter 160, but this is omitted in FIG.

  FIG. 2 is a configuration example of the frequency information detection apparatus 1. As shown in FIG. 2, the frequency information detection apparatus 1 includes an A / D conversion unit 10, an estimated phase information output unit 20, an estimated frequency information output unit 30, a moving average unit 40, an error frequency information output unit 70, and a frequency correction unit. 90.

<A/D converter 10>
A / D converter 10, a predetermined sampling period the input signal _{V in} analog output from the filter 160 (the reciprocal of the sampling frequency _{f s:} 1 / _{f s)} to convert into a digital signal. The A / D converted sampling data is sequentially output as an instantaneous value V _in [t] of the input signal V _in at time [t]. Thereby, the alternating analog signal waveform is converted into a digital signal waveform composed of a plurality of sampling data. Note that if the input signal V _in outputted from the filter 160 is a sine wave signal, a digital signal waveform output from the A / D converter 10 is also a sine wave signal.
The A / D converter 10 includes an A / D converter and performs the A / D conversion as described above. Further, a low-pass filter or a band-pass filter may be provided at the subsequent stage of the A / D converter so as to pass the fundamental frequency component.

Here, frequency _{f in} of the input signal _{V in,} the sampling frequency _{f s,} the phase offset theta ', when the relative angular frequency is _{ω in = 2π · (f in} / f s), the input at time [t] The instantaneous value V _in [t] of the signal V _in can be expressed as shown in Equation (2). It should be noted that, here, is set to "1", the amplitude of the input signal V _in. The time data “t” is a variable that is incremented every sampling period.
V _in [t] = sin (2π · (f _in / f _s ) [t] + θ ′)
= Sin (ω _in [t] + θ ′) (2)

<Estimated phase information output unit 20>
The estimated phase information output unit 20 outputs the estimated value α [t] of the phase information of the input signal Vin at time [t] as estimated phase information. The estimated phase information output unit 20 includes a cosine value estimation unit 21 and a phase estimation unit 22.

[Cosine value estimation unit 21]
The cosine value estimation unit 21 obtains the estimated value of the cosine value “cos (ω _in [t] + θ ′)” obtained by differentiating the instantaneous value V _in [t] of the input signal V _{in at} the time [t] at the time [t− 1] instantaneous value _V in of the input signal _{V in} at the [t-1], is calculated using the instantaneous value _V in [t + 1] and the known values of the input signal _{V in} at the time [t + 1]. The estimated cosine value “cos (ω _in [t] + θ ′)” is sent to the phase estimation unit 22. This will be specifically described below.

When “α [t] = ω _in [t] + θ ′ = 2π · (f _in / f _s ) [t] + θ ′”, the cosine value (cos (α [t])) at time [t] is It can be expressed by equation (3).
cos (α [t])
= {(2 sin (ω _in [t]) · cos (α [t])} / (2 sin (ω _in [t]))
= {Sin (α [t] + ω _in [t]) − sin (α [t] −ω _in [t])} / (2 sin (ω _in [t]))
= {Sin (2ω _in [t] + θ ′) − sin (θ ′)} / (2 sin (ω _in [t])) (3)

Referring to equation (2), because the time [t] instantaneous value _V in [t] of the input signal _{V in} in is a _{"sin (ω in [t] +} θ ') ", the frequency _{f in} of the input signal _{V in} If does not change, the displacement amount of the relative angular frequency ω _in for each sampling period (1 / f _s ) is constant at “ω _in [t]”. Thus, molecules of formula (3), the time from the instantaneous value _V in of the input signal _{V in} at the [t + 1] [t + 1], time instantaneous value _V in of the input signal _{V in} at the [t-1] [t- 1] Represents subtracting.

In addition, “sin (ω _in [t])” of the denominator of Expression (3) is obtained by omitting the phase offset θ ′ from the instantaneous value V _in [t] of the input signal V _in at time [t]. Yes.

Therefore, the cosine value (cos (α [t])) at time [t] is the instantaneous value V _in [t−1] of the three input signals V _in at time [t−1], [t] and [t + 1]. ], V _in [t] and V _in [t + 1].

However, as described above, “sin (ω _in [t])” is obtained by omitting the phase offset θ ′ from the instantaneous value V _in [t] of the input signal V _in at time [t]. An error occurs in the denominator of Equation (3). Here, since the value of the sine function (the value of sin) changes within a range of ± 1 centering on 0, the error may be positive or negative.

As will be described later, after estimating the cosine value (cos (α [t])) at time [t], the phase α [t] is calculated, and the phase displacement amount Δα for each sampling period of the phase α [t]. It calculates an estimate _{f est} of the frequency of the input signal _{V in} on the basis of the [t] [t] or estimate omega _est of the angular frequency [t]. After that, since the moving average value of the calculated frequency estimated value f _est [t] or the angular frequency estimated value ω _est [t] is calculated, the denominator error in the equation (3) is almost canceled.

Therefore, even if the frequency f _in of the input signal V _in constituting the denominator of the equation (3) having an error is replaced with the set frequency f _set as in the equation (4), there is almost no influence. Here, the setting frequency _{f set,} because it is setting the frequency of the input signal _{V in,} the error is small. For example, in the examples of FIGS. 5 and 6, the frequency of the high frequency signal output from the oscillator 100. Since this set frequency f _set is known in advance, it may be input to the cosine value estimation unit 21, for example. Of course, the set angular frequency ω _set (= 2π · f _in ) may be used.
2 sin (ω _in [t]) = 2 sin (2π · (f _in / f _s ) [t])
→ 2 sin (2π · (f _set / f _s ) [t]) (4)

In this specification, the set frequency f _set and the set angular frequency ω _set are collectively referred to as “set frequency information F _set ”.

Here, all the elements constituting the formula (4) are all known values. Therefore, when “1 / {2 sin (2π · (f _set / f _s )] [t])}” is expressed by a constant K, Equation (3) can be transformed into Equation (5). Therefore, originally, when estimating the cosine value (cos (α [t])) at time [t], it is necessary to use the equation (3), and thus a complicated calculation is required. By using (5), the calculation formula can be simplified and the calculation load can be reduced.
cos (α [t]) = K · {sin (2ω _in [t] + θ ′) − sin (θ ′)}
= K · {V _in [t + 1] −V _in [t−1]} (5)

Synchronization frequency information F _syc [t] described later is obtained by adding correction frequency information F _rev [t] (described later) to _set frequency information F _set to obtain synchronization frequency information F _syc [t]. F _set is approximated to estimated frequency moving average value information F _{est_ave} [t] output from the moving average unit 40 described later. Further, since the synchronization frequency information F _syc [t] is known, the synchronization frequency information F _syc [t] can be input instead of inputting the _set frequency information F _set .

As described above, the cosine value estimating unit 21, the cosine value at time [t] by using the data of a plurality of input signal _{V in} which a digital signal by the A / D converter 10 (cos (α [t] ) ) Is calculated. For this reason, the cosine value estimation unit 21 has a memory (not shown) and sequentially stores the instantaneous value V _in [t] of the input signal V _in output from the A / D conversion unit 10 _in the memory.

Incidentally, as described above, the cosine value estimating unit 21, a time [t-1] and the instantaneous value _V in of the input signal _{V in} [t-1] at time [t + 1] instantaneous value V of the input signal _{V in} at _in [t + 1] is used, but the instantaneous value V _in [t] of the input signal V _in at time [t] is not used. However, since the instantaneous value V _in [t] of the input signal V _{in at} the time [t] is used for the calculation performed by the phase estimation unit 22 as described later, at least the instantaneous value V _in [t−1] is stored in the memory. , Three consecutive data of the instantaneous value V _in [t] and the instantaneous value V _in [t + 1] are stored.

In the above example, after the instantaneous value V _in [t + 1] of the input signal V _in at time [t + 1] is input to the cosine value estimation unit 21, the cosine value (cos (α [t] at time [t] is )) Can be computed.

[Phase estimation unit 22]
The phase estimation unit 22 uses the instantaneous value V _in [t] of the input signal V _in at time [t] as a sine element (sin element), and uses the cosine value (cos (α [t])) at time [t] as a cosine. As an element (cos element), the phase α [t] of the cosine value (cos (α [t])) at time [t] is calculated using an arctangent function (tan ⁻¹ ) as shown in Expression (6). Calculate. The phase α [t] is calculated within a range of ± π [unit: rad]. The calculated phase α [t] is sent to the phase displacement amount calculation unit 31.
α [t] = tan ⁻¹ (V _in [t] / cos (α [t])) (6)

As described above, after the instantaneous value V _in [t + 1] of the input signal V _{in at} the time [t + 1] is input to the cosine value estimation unit 21, the cosine value (cos (α [t] at the time [t]) is input. because)) is calculated, phase alpha [t] is also time instantaneous value _V in of the input signal _{V in} at the [t + 1] [t + 1] is calculated after being input to the cosine value estimating unit 21. Further, the instantaneous value V _in [t] of the input signal V _{in at} the time [t] necessary for the calculation of the phase α [t] may be read from the memory provided in the cosine value estimation unit 21.

<Estimated frequency information output unit 30>
The estimated frequency information output unit 30 outputs an estimated value of frequency information of the input signal Vin at time [t] as estimated frequency information F _est [t]. The estimated frequency information output unit 30 includes a phase displacement amount calculation unit 31 and a frequency estimation unit 32.

[Phase displacement calculation unit 31]
The phase displacement amount calculation unit 31 is based on the phase α [t−1] at time [t−1] and the phase α [t] at time [t] calculated by the phase estimation unit 22 at time [t]. The phase displacement amount Δα [t] is calculated. The calculated phase displacement amount Δα [t] is sent to the frequency estimation unit 32.

Here, the phase displacement amount calculation unit 31 calculates the phase displacement amount Δα [t] generated during the sampling period from time [t−1] to time [t]. Since the phase α [t] of the cosine value (cos (α [t])) at the time [t] is calculated in the range of ± π [unit: rad] as described above, {“α [t ] −α [t−1]} is positive, the phase displacement amount Δα [t] is calculated using Expression (7), and {“α [t] −α [t−1]} is negative. In this case, the phase displacement amount Δα [t] is calculated using Expression (8).
Δα [t] = α [t] −α [t−1] (7)
Δα [t] = (α [t] −α [t−1]) + 2π (8)

  As described above, the phase displacement amount calculation unit 31 uses the phase α [t−1] at time [t−1] and the phase α [t] at time [t] calculated by the phase estimation unit 22. Since the phase displacement amount Δα [t] at time [t] is calculated, the phase displacement amount calculation unit 31 has a memory (not shown), and the phase α [t] output from the phase estimation unit 22 is stored in the memory. It memorizes sequentially. This memory stores at least the phase α [t−1] at time [t−1] and the phase α [t] at time [t].

Further, as described above, after the instantaneous value V _in [t + 1] of the input signal V _in at time [t + 1] is input to the cosine value estimation unit 21, the phase α [t] at time [t] is calculated. Therefore, the phase displacement amount Δα [t] is also calculated after the instantaneous value V _in [t + 1] of the input signal V _in at time [t + 1] is input to the cosine value estimation unit 21.

[Frequency estimation unit 32]
Frequency estimation unit 32, based on the phase shift amount [Delta] [alpha] [t] in the calculated time in the phase shift amount calculation unit 31 [t], the time estimate of the angular frequency of the input signal V _in at the [t] ω _est [ t] or the estimated frequency f _est [t]. Estimate omega _est of the angular frequency of the operation input signal _{V in} [t] or frequency estimate _{f est} [t] is sent to the error frequency information output section 70.

Here, since the phase displacement amount Δα [t] is a phase displacement amount generated during the sampling period, as shown in the equation (9), when the phase displacement amount Δα [t] is multiplied by the sampling frequency f _s. The amount of phase displacement that occurs in one second can be determined. That is, the time estimate of the angular frequency of the input signal _{V in} at the _{[t] ω est [t]} [ Unit: rad / s] can be calculated. Further, as shown in the equation (10), the phase displacement amount Δα [t] at the time [t] is multiplied by the sampling frequency f _s and divided by “2π” to obtain the estimated value of the frequency at the time [t]. f _est [t] [unit: Hz] can be calculated.
ω _est [t] = Δα · f _s (9)
f _est [t] = Δα · f _s / (2π) (10)

Further, the estimated angular frequency value ω _dif [t] between the estimated angular frequency value ω _est [t] and the set angular frequency ω _set can be calculated as in Expression (11). The set angular frequency ω _set is “2π · f _set ”. Moreover, the estimated frequency f _dif [t] of the difference frequency between the estimated frequency f _est [t] and the set frequency f _set can be calculated as in Expression (12). Therefore, the frequency estimation unit 32 can also output an estimated value ω _dif [t] of the difference angular frequency and an estimated value f _dif [t] of the difference frequency. In this case, for example, the set frequency f _set may be input to the frequency estimation unit 32. Of course, the set angular frequency ω _set may be input to the frequency estimation unit 32. Also it may be calculated to set angular frequency omega _{The set} from the set frequency _{f The set,} and vice versa, may be calculated set frequency _{f The set} of setting angular frequency omega _{The set.} Of course, instead of inputting the _set frequency information F _set as described above, the synchronization frequency information F _syc [t] can be input.
ω _dif [t] = ω _est [t] −ω _set [t] (11)
f _dif [t] = f _est [t] −f _set [t] (12)

In this specification, the estimated value omega _est of the angular frequency of the input signal _{V in} [t], the estimate _f est of the frequency [t], the estimated value omega _dif difference angular frequency [t] and the estimated value of the frequency f _dif [t] of the input signal _{V in} at the time [t] are collectively referred to as "estimated frequency information _{F est} [t]".

Also, the estimated value omega _est of the angular frequency of the input signal _{V in} [t], the estimate _f est of the frequency [t], the estimated value omega _dif difference angular frequency [t] and the estimated value _f dif difference frequency [t] May output any one of them, but may output a plurality of them.

Incidentally, as described above, the time phase displacement of [Delta] [alpha] [t] in the [t], the time [t + 1] instantaneous value _V in of the input signal _{V in} at the [t + 1] is operation after being input to the cosine value estimating unit 21 since the time estimated frequency information _{F est} [t] of the input signal _{V in} at the [t] is also time instantaneous value _V in of the input signal _{V in} at the [t + 1] [t + 1] is input to the cosine value estimating unit 21 It is calculated after.

[Moving average unit 40]
The moving average unit 40 calculates a moving average value of the estimated frequency information F _est [t] of the input signal Vin _in the time [t] output from the frequency estimation unit 32, and estimates frequency moving average value information F _{est_ave} [t ] Is output. The moving average value is calculated using a predetermined number of data.

As described above, the moving average unit 40, based on the estimated frequency information F _{est [t]} of the input signal V _in at the time [t] outputted from the frequency estimation unit 32, a predetermined number of data moving average Therefore, the moving average unit 40 has a memory (not shown) and sequentially stores the estimated frequency information F _est [t] output from the frequency estimation unit 32 in the memory. This memory stores at least a predetermined number of data necessary for calculating the moving average value.

Further, as described above, the time after the instantaneous value _V in of the input signal _{V in} [t + 1] is input to the cosine value estimation unit 20 in the [t + 1], the estimated frequency information F of the input signal _{V in} at the time [t] since _est [t] is computed, the moving average value omega _ave of the angular frequency [t], the moving average value _{f ave} frequency [t] is also instantaneous value _V in of the input signal _{V in} at the time [t + 1] [t + 1 ] Is input to the cosine value estimation unit 20 and then calculated.

Further, immediately after starting sampling of data, a predetermined number of data necessary for calculating the moving average value is not stored in the memory. Therefore, while the predetermined number of data is not stored in the memory, the moving average value may be calculated using data smaller than the predetermined number. Alternatively, the moving average value may be calculated after a predetermined number of data is stored in the memory. How to calculate the moving average value may be determined in advance.

The moving average unit 40, the estimated value omega _dif difference angular frequency [t] moving average _{omega Dif_ave} [t] or difference frequency estimate _{f dif} [t] moving average _{f Dif_ave} [t] the output of the You can also

Further, in this specification, the moving average value ω _{est_ave} [t] of the estimated value ω _est [t] of the angular frequency and the moving average value f _{est_ave} [t] of the estimated value f _est [t] of the frequency at _time [t]. are collectively moving average _{omega Dif_ave} estimate omega _dif difference angular frequency [t] [t] and the moving average value _{f Dif_ave} estimate _{f dif} difference frequency [t] [t], the time [t] referred to as "estimated frequency moving average value information _{F est_ave} [t]" of the input signal _{V in} in.

In addition, the moving average value ω _{est_ave} [t] of the estimated value ω _est [t] of the angular frequency, the moving average value f _{est_ave} [t] of the estimated value f _est [t] of the frequency, and the estimated value ω _dif [ _{Any one} of the moving average value ω _{dif_ave} [t] of t] and the moving average value f _{dif_ave} [t] of the estimated value f _dif [t] of the difference frequency may be output. May be.

Estimated frequency moving average value information _{F Est_ave} of the input signal _{V in} which were determined as described above [t], for example, sent to a second high-frequency power source 120 shown in FIGS. 5 and 6, the second RF power supply The output frequency of 120 can be used to match the output frequency of the first high-frequency power supply 110. In this case, in order to make the control clock of the frequency information detection device 1 and the control clock of the second high-frequency power source 120 the same, the same control is performed, for example, by incorporating the frequency information detection device 1 into the second high-frequency power source 120. A clock may be used.

Incidentally, the estimated frequency moving average value information F _{Est_ave} of the input signal V _in outputted from the moving average unit 40 the _[t] with the digital data, the estimated frequency moving average oscillator of the second high-frequency power source 120 is a digital data It is desirable that a high frequency signal having a frequency indicated by the value information F _{est_ave} [t] can be generated. A direct digital synthesizer (DDS) can be used as such an oscillator.

<Reasons for calculating the moving average value>
As described above, since the cosine value (cos (α [t])) at time [t] includes an error, the phase α [t] also includes an error. Of course, the cosine value (cos (α) [t−1]) and the phase α [t−1] at time [t−1] also include errors. As a result, the phase displacement amount Δα [t] also includes an error. Therefore, also included an error in the time estimated frequency information _{F est} of the input signal _{V in} at the [t] [t].

That is, the error included _in the estimated frequency information F _est [t] of the input signal Vin at time [t] is due to the cosine value (cos (α [t])) at time [t].
Error of the cosine value in this time [t] (cos (α [ t])) , so also be negative if also become positive, as described above, the input signal V _in at the time [t] The error of the estimated frequency information F _est [t] may be positive or negative. Therefore, if the moving average unit 40 performs a moving average of a plurality of data, the error is canceled out, so that the error can be reduced and accurate frequency information can be obtained.

Therefore, when estimating the cosine value (cos (α [t])) obtained by differentiating the instantaneous value V _in [t] of the input signal V _in at time [t], the set frequency information F _set or the synchronization frequency information F _By using _syc [t] and _setting a part of equation (5) as a constant, it is possible to reduce the calculation load for obtaining frequency information and obtain accurate frequency information.

<Error frequency information output unit 70>
The error frequency information output unit 70 represents the error of the synchronization frequency information F _syc [t] with respect to the estimated frequency moving average value information F _{est_ave} [t] at time [t], and the error frequency information F _err [t] at time [t]. Output as.

The estimated frequency information F _est [t] and the synchronization frequency information F _syc [t] need to be compared with the same type of information. For example, the estimated frequency information F _est [t] is expressed in frequency. If present, the synchronization frequency information F _syc [t] also needs to be expressed in frequency. The same applies to the angular frequency.

<Frequency correction unit 90>
When the error frequency information F _err [t] at the time [t] is outside the predetermined range, the frequency correction unit 90 corrects the frequency information at the time [t] so that the error frequency information F _err [t] is reduced. Determine and output F _rev [t]. Hereinafter, the frequency correction unit 90 will be further described.

Frequency of the input signal _{V in} is obtained by estimating frequency information _{F est} outputted from the frequency estimation unit 32 [t]. Further, in order to reduce the error and obtain an accurate estimate value of the frequency information, a moving average unit 40 is provided after the frequency estimation unit 32. In this case, as the error between the frequency of the set frequency information F _{The set} and the input signal V _in it is low, it is possible to obtain an estimate of a good frequency information accurate.
Therefore, obtains synchronization frequency information _{F syc} [t] by adding the correction frequency information _{F rev} [t] to be described later to set frequency information _{F The set,} the frequency of the input signal _{V in} the synchronizing frequency information _{F syc} [t] Control to approach. If the synchronization frequency information F _syc [t] is used instead of the _set frequency information F _set , it is possible to obtain an accurate estimate value of the frequency information.

<Simulation results>
3, when the frequency _{f in} of the input signal _{V in} is assumed to have shifted from the set frequency _{f The set,} the moving average value f of the estimated value _{f dif} difference frequency output from the moving average unit 71 [t] It is a simulation result of _{dif_ave} [t]. 3 (a) is a simulation result of the estimated value _{f dif} difference frequency of the input signal _{V in} [t], FIG. 3 (b), the estimated value _f dif difference frequency of the input signal _{V in} [t] FIG. 3C is a partially enlarged view of FIG. 3B. FIG. 3C is a simulation result of the moving average value f _{dif_ave} [t] of FIG. 3A and 3B show data for about 150 to 160 μs after the start of sampling, and FIG. 3C shows data for about 159 to 160 μs after the start of sampling.

The simulation conditions are as follows.
(1) Sampling frequency f _s : 50 MHz
(2) Setting frequency f _set : 13.56 MHz
(3) of the input signal _{V in} frequency _f in: 13.563MHz
(Assuming that the _set frequency f _set is shifted by 3,000 Hz)
(4) Number of data used for calculating the moving average value: 500 (moving average value of 10 μs)

4, when the frequency _{f in} of the input signal _{V in} is assumed to have shifted from the set frequency _{f The set,} the moving average value f of the estimated value _{f dif} difference frequency output from the moving average unit 71 [t] It is another simulation result of _{dif_ave} [t]. Figure 4 (a) is a simulation result of the estimated value _{f dif} difference frequency of the input signal _{V in} [t], FIG. 4 (b), the estimated value _f dif difference frequency of the input signal _{V in} [t] FIG. 4C is a partially enlarged view of FIG. 4B. FIG. 4C is a simulation result of the moving average value f _{dif_ave} [t] of FIG.
Incidentally, FIG. 4, except that the frequency _{f in} of the input signal _{V in} is 13.5603MHz (assuming are shifted by setting the frequency _{f The set} and 300 Hz), is the result of simulation under the same conditions as FIG. 3 .

As shown in FIG. 3 (a), the input signal _V when the frequency of the _in is assumed to have shifted 3,000Hz the set frequency _f set (13.56MHz), the estimated value f of the difference between the frequency of the input signal _{V in dif} [t] varies in the range of about ± 400 Hz with respect to the frequency to be detected (3,000 Hz), but as shown in FIGS. 3B and 3C, the estimated value of the difference frequency f _dif [t] moving average _{f dif_ave} [t] of, it is seen that within the range of about ± 0.6 Hz for the frequency to be detected (3,000 Hz).
As described above, even when the frequency of the input signal V _in is significantly deviated from the set frequency f _set , the moving average of the estimated value f _dif [t] of the difference frequency output from the moving average unit 71 is obtained. The value f _{dif_ave} [t] has a small error with respect to the frequency to be detected.

Further, as shown in FIG. 4 (a), when the frequency _{f in} of the input signal _{V in} is assumed to have shifted 300Hz the set frequency _f set (13.56MHz), estimation of the difference frequency of the input signal _{V in} The value f _dif [t] varies in a range of about ± 40 Hz with respect to the frequency to be detected (300 Hz), but as shown in FIGS. 4B and 4C, the estimated value of the difference frequency. f _dif [t] moving average _{f dif_ave} [t] of, it is seen that within the range of about ± 0.05 Hz for the frequency to be detected (300 Hz). Therefore, it can be seen that can detect frequency accurately the input signal V _in.

That is, as the error between the frequency _{f in} of the input signal _{V in} and the set frequency _{f The set} is small, it can be seen that can detect a frequency of more accurately the input signal _{V in.}

The frequency information detection apparatus of this embodiment, the sine wave signal V _in analog, (the reciprocal of the sampling frequency f _s: 1 / f _s) a predetermined sampling period by A / D-converted digital signal This is a frequency information detection device that detects frequency information of the obtained digital sine wave signal, and its application is not particularly limited. Therefore, although the example which uses a frequency information detection apparatus for a high frequency electric power supply system like FIG.5 and FIG.6, for example was shown above, it is not limited to this. For example, it can be applied even in a commercial frequency band (50 Hz to 60 Hz).

DESCRIPTION OF SYMBOLS 1 Frequency information detection apparatus 10 A / D conversion part 20 Estimated phase information output part 21 Cosine value estimation part 22 Phase estimation part 30 Estimated frequency information output part 31 Phase displacement calculation part 32 Frequency estimation part 40 Moving average part 70 Error frequency information Output unit 90 Frequency correction unit 100 Oscillator 110 First high frequency power source 120 Second high frequency power source 130 Plasma processing apparatus (load)
131 Electrode 132 Electrode 133 Plasma 140 Transmission Line 150 Directional Coupler 160 Filter

Claims (4)

A frequency information detecting apparatus for detecting the frequency information of the analog sine wave signal V _in,
A sine wave signal V _in analog, (the reciprocal of the sampling frequency f _s: 1 / f _s) a predetermined sampling period by the instantaneous value V of the input signal V _in at the time [t] obtained by converting a digital signal an A / D converter that converts _in [t] sequentially into a sine wave signal composed of a plurality of sampling data;
Time cosine value obtained by differentiating the instantaneous value _V in [t] of the input signal _{V in} in [t] (cos (α [ t])) an estimate of the instantaneous value of the input signal _{V in} at the time [t-1] A cosine value estimator that uses V _in [t−1], an instantaneous value V _in [t + 1] of the input signal V _in at time [t + 1], and a known value;
Time the instantaneous value _V in [t] of the input signal _{V in} at the [t] and a sine component (sin component), the cosine value at time [t] estimated by the cosine value estimating section (cos (alpha [t]) ) As a cosine element (cos element), and a phase estimation unit that calculates a phase α [t] of a cosine value (cos (α [t])) at time [t] using an arctangent function (tan ⁻¹ ) ,
Based on the phase α [t−1] at time [t−1] and the phase α [t] at time [t] calculated by the phase estimation unit, the phase displacement amount Δα [t] at time [t] is calculated. A phase displacement amount calculation unit to be calculated;
A frequency estimation unit for calculating the estimated frequency information of the input signal V _in in on the basis of the phase shift amount [Delta] [alpha] [t] in the calculated time in the phase shift amount calculation unit [t], the time [t],
A moving average unit calculates the moving average of the estimated frequency information of the input signal V _in at the time [t] calculated by the frequency estimation unit, and outputs the estimated frequency moving average information,
Frequency information obtained by adding the correction frequency information to the set frequency information of the input signal V _in when the synchronizing frequency information, calculates the error of the synchronizing frequency information for the estimated frequency moving average value information at the time [t], the error An error frequency information output unit for outputting as frequency information;
A frequency correction unit that determines and outputs the correction frequency information so as to reduce the error frequency information;
A frequency information detection apparatus comprising: The cosine value estimator calculates a cosine value (cos (α [t])) by Equation (1), where K is a constant calculated using the synchronization frequency information. Item 2. The frequency information detection device according to Item 1.
cos (α [t]) = K · {V _in [t + 1] −V _in [t−1]} (1) The frequency information detecting apparatus according to claim 2, wherein the constant K is calculated by the equation (2). However, the synchronization frequency information is “F _syc ”.
K = 1 / {2sin (2π · (F _syc / f _s ) [t])} (2) The high frequency power supply device which uses the synchronous frequency information output from the frequency information detection apparatus in any one of Claims 1-3 as setting frequency information of an output frequency.

Images