WO2010094202A1

WO2010094202A1 - Lumped parameter rectangular band pass filter

Info

Publication number: WO2010094202A1
Application number: PCT/CN2009/075673
Authority: WO
Inventors: 何连成
Original assignee: He Liancheng
Priority date: 2009-02-17
Filing date: 2009-12-17
Publication date: 2010-08-26
Also published as: CN101488732A; CN101488732B

Abstract

A lumped parameter rectangular band pass filter mainly for aerial tuning network has at least two stages of the serial and parallel arm units, wherein, each stage of the serial and parallel arm units consists of one serial arm and one parallel arm, the serial arm is formed of a capacitance and an inductance connected to each other in series, the parallel arm is a reactive element.

Description

Lumped parameter rectangular bandpass filter

The present invention relates to a filter, which is mainly applied to a high frequency power transmission network, and more particularly to a lumped parameter rectangular band pass filter mainly used for a medium wave sky-tuning network in a wireless transmission system.

Background art

In the field of small signal processing, high-performance band-pass filters can be designed with active circuits. By introducing feedback to improve the rectangular characteristics of the filter, digital filters can also be used. Frequency shifting can also be used to utilize ceramic filters. High rectangularity pass frequency characteristics of a device or surface acoustic wave filter.

However, in the case of high power transmission, the above methods are impossible. When the frequency is very high (above TV, FM band), the narrow bandpass filter can be designed to take advantage of the performance of the cavity and transmission line, but at a low frequency (below the short band)

In this case, only inductive and capacitive components can be used, which can be implemented passively.

In high-power (several hundred watts to several hundred kilowatts) wireless transmission systems, a band-pass filter with a pass-frequency characteristic close to a rectangle is required between the transmitter and the antenna, allowing both the in-band frequency to pass smoothly and greatly attenuate. Out-of-band frequencies to avoid out-of-band emissions from the transmitter and other frequencies. In the low frequency range (hundreds of kilohertz to several megahertz) of medium, long and short waves, this bandpass filter can only be implemented with lumped parameter components (inductor coils, capacitors). For example, the medium-wave network of the medium wave should ideally use this band-pass filter to achieve impedance matching, anti-interference and lightning protection. However, due to the design of the high-rectangle lumped-parameter band-pass filter, it has always been a difficult problem in the academic world. Therefore, the traditional wave-adjusting network uses impedance matching for the carrier center frequency plus point frequency blocking for the interference frequency. Or absorb to achieve.

The traditional medium wave network design method has the following problems:

1. Using the traditional design method, when there are many transmission frequencies in the station (or high-power stations nearby), it is necessary to add many targeted blocking or absorbing units to the network, which will inevitably narrow the bandwidth of the network. The band reflection increases dramatically. Therefore, it is a waste of land and human resources to build new stations to disperse these frequencies.

2. At the low frequency end of the mid-band, when the effective height of the antenna is much smaller than a quarter wavelength (for example, 558 kHz carrier frequency, antenna height 76 m), the bandwidth of the antenna near the carrier frequency is very narrow. The traditional design method cannot expand the bandwidth. Therefore, it is impossible to obtain a qualified network design result (in the carrier frequency range of ±9 kHz, the standing wave ratio is less than 1.2).

3. When three or more frequencies are common to the tower, if the ratio of adjacent frequencies is less than 1.2, the network bandwidth will be narrowed due to the access of the common tower blocking unit, and the traditional design method cannot be qualified. Design results. References to traditional mid-wave-tone network design are given below:

1. Zhang Yuzao et al. The principle and maintenance of all-solid PDM medium wave transmission system [M]. Beijing: China Radio and Television Publishing House, November 1999, first edition, 266-336.

2, Zhang Yuzao and other editors. All solid-state medium wave transmission system adjustment and maintenance [M]. Xiamen: Xiamen University Press, July 2007, the first edition, 317-412.

Summary of the invention

The object of the present invention is to provide a not only better impedance conversion to meet impedance matching requirements, but also good in-band passband characteristics, out-of-band attenuation characteristics, bandwidth expansion capability, and lightning protection capability. The circuit form is mainly applied to the lumped parameter rectangular bandpass filter of the medium wave sky-tuning network.

The technical solution of the invention is composed of a lumped parameter inductor and a capacitor element.

The present invention is provided with at least two stages of parallel arm units, each of which is composed of one string arm and one parallel arm, the string arm is formed by connecting a capacitor and an inductor in series, and the arm is a reactance element.

The capacitance in the string arm of the first stage parallel arm unit is preferably 200 to 2000 pf, the inductance is preferably 1 to 100 μΗ, and the arm reactance value is preferably 5 to 150 Ω _; from the second stage serial arm unit The capacitance in the string arm is preferably 100 to 5000 pf, the inductance is preferably 1 to 100 μm, and the reactance of the arm is preferably 5 to 200 Ω.

The reactance component of the arm can be an inductor or a capacitor, or a series or parallel connection of the capacitor and the inductor, but it is not necessary to use it because it is complicated compared with a simple capacitor or an inductor. There is no difference in function.

Preferably, the parallel arm in the last stage of the parallel arm unit uses an inductive coil having an inductance of 2 to 12 μΗ. The last stage uses the inductor coil just to facilitate the adjustment of its inductance during debugging. Variable inductors with high power capacity are much less expensive than variable capacitors.

Compared with the existing medium wave network, the present invention has the following outstanding advantages:

1) In-band pass frequency characteristics:

The center frequency point fo has a standing wave ratio of 1.0, and the in-band standing wave ratio of ±9 kHz (18 kHz bandwidth) is less than 1.2, so that the problem of excessive sideband reflection can be completely solved. If used in other frequency bands, this bandwidth is approximately 1% to 4% of ft.

2) Out-of-band attenuation characteristic - When the circuit of the present invention includes three serial-parallel arm units, while ensuring the in-band pass-frequency characteristic, the attenuation amount can reach 20 dB or more as long as the band is externally separated from the center frequency by about 40 kHz. The pass frequency characteristic curve is close to a rectangle and has a cutoff characteristic for high power emission interference at or near the station. For example, the center frequency is 801 kHz, the out-of-band frequency is greater than 841 kHz and the attenuation less than 761 kkHz is greater than 20 dB, which gives the network ideal anti-interference capability without the need for any additional blocking or absorbing unit circuits. Therefore, the present invention can well solve the first problem of the above conventional method. If used in other frequency bands, a 20 dB bandwidth is approximately 4% to 15% of the carrier frequency f _Q .

3) Bandwidth expansion capability - The present invention can expand the bandwidth up to 4 times, so that the second and third problems of the above conventional methods can be easily solved. For a conventional antenna (different from a short antenna), as long as the ratio of adjacent carrier frequencies reaches 1.1, the two-frequency common tower can be smoothly realized by the present invention; as long as the ratio reaches 1.15, a three-frequency and four-frequency common tower can be realized; When the ratio reaches 1.2 (if the carrier frequency is less than 700 kHz, the ratio is required to reach 1.3), the problem of the five-frequency and six-frequency common towers can be easily solved. This technology makes the industry dream of manufacturing "medium wave multiplexer" a reality. Promoting in the country, at least reducing the footprint of the launch site by several hundred thousand mu, has great practical significance.

4) Lightning protection capability:

The main frequencies of lightning energy are DC and low frequency (below tens of kHz), and high frequency bands (several MHz or more), because the design of the present invention only allows narrow-band energy near the carrier frequency to pass, and exhibits cut-off characteristics for other frequencies, so Greatly attenuating the lightning energy, the actual effect is very good.

5) Unified circuit form:

The design of the present invention enables the medium wave sky tone network in any case to have a unified circuit form, which is advantageous for standardized production. At different carrier frequencies, the network only needs to adjust the component parameter values without changing the circuit form.

If the use of two serial form and the arm unit (see FIG. 1, that is, remove L2, C2 and C5), the bandwidth expansion capability only 2 times, from the point of 20dB attenuation frequency f _Q is about 100kHz.

DRAWINGS

1 is a schematic structural view of an embodiment of the present invention;

Figure 2 is a schematic diagram of the composition of the positively variable characteristic circuit;

Figure 3 is a schematic diagram of the circuit composition of a negatively variable characteristic reactance in a certain frequency range;

Figure 4 is a 1^, C, R series circuit (a) and its impedance characteristics (b);

Figure 5 is a circuit combination (a) with negative micro-variable reactance and its negative micro-variation characteristic (b);

Figure 6 shows the structure of the medium wave sky tone network;

Figure 7 is a Z _B track diagram;

Figure 8 is a Z _D trace diagram of critical compensation;

Figure 9 is a Z _D trace diagram of overcompensation;

Figure 10 is an under-compensated Z _D trace diagram;

Figure 11 is a schematic diagram of the impedance compensation mode of the critical compensation mode;

Figure 12 is an overcompensation mode matching impedance diagram;

Figure 13 is a 558kHz single frequency single tower network diagram; Figure 14 is a 900 kHz two-frequency common tower network diagram;

Figure 15 is a 1296 kHz six-frequency common tower network diagram.

detailed description

The invention will be further illustrated by the following examples in conjunction with the accompanying drawings.

The design principle of the present invention is described below.

The traditional method only focuses on the impedance matching of the center frequency, that is, using several reactive components (inductance, capacitance), and the series matching of the antenna load in parallel with the antenna load makes the total matching impedance of the center frequency point equal to the characteristic impedance of the transmission line. In addition, there is no targeted means to guarantee the impedance matching of other frequencies in the frequency band.

Based on the traditional method, the invention increases the analysis of the trend of the total deployment impedance change when the frequency changes near the center frequency (so-called micro-variation parameter analysis), and adopts an effective targeted means to offset and mitigate this. A variety of trends, thereby achieving the purpose of broadening the frequency band while increasing the out-of-band attenuation. This is the overall design idea of the microvariate parameter design method.

Because the present invention is a completely new design concept, in order to illustrate the principles, three new concepts and definitions need to be introduced:

The positive and negative micro-variation characteristics of the reactance—a reactance variable, within a certain frequency range, when the frequency increases, if the reactance value changes in the direction of the inductive reactance or the capacitive reactance decreases, then the reactance has f«^, and vice versa. The mathematical expression is as follows - set = , then ^^

Df df if ^ _{> 0} , then ^ has positive micro-variation characteristics;

Df If ^ _<0 , then ;^ has a negative micro-variation characteristic.

Df

Characteristic bandwidth of the load - if a load Z(f) = R + jX(f) ( Ω ), where X(f) has a positive micro-variation near f _Q and ^ ZL _W ( Ω / kHz), then The characteristic bandwidth of the load Z(f) in the vicinity is ±H (kHz). The meaning of the physical Af m is that this bandwidth is obtained when the load Z(f) is subjected to impedance transformation by a simple circuit (r network, inverted r network, T network, π network, etc.) as a transmission line terminal load. The maximum bandwidth of the standing wave ratio ρ ≤ 1.2. This conclusion can be proved by the theory of high-frequency transmission lines, where the proof process is omitted.

1) Reactance circuit with positive micro-variation characteristics

The reactances of a single inductor and capacitor are: Xc = -j— X _L = j2 ^ fL Because of j2nL , a single inductor and capacitor have positively variable characteristics.

Similarly, it can be proved that the reactance between the circuit combination person and B shown in FIG. 2 has positive micro-variation characteristics.

In Figure 2, the reactance of Figure (a) in the full frequency range has positive micro-variation characteristics, and other circuits (Figures b, c, d, e, f) have the mutations at the individual frequency points, and the other frequencies have Positive micro-variation characteristics.

In fact, in the absence of a resistive component, a circuit consisting of only inductive and capacitive components has positively variable characteristics in addition to individual frequency points.

2) Circuits with negative micro-variable reactance in a certain frequency range

To obtain a negative micro-variable reactance in a certain frequency range, it must contain three components: resistor, capacitor, and inductor, and have the circuit form shown in Figure 3.

A common feature of the four circuit forms in Figure 3 is that a reactive component (inductor or capacitor) is connected in parallel with a string of resistors, capacitors, and inductors (series connected in series).

In the analysis of the above circuit, the analytic complex mathematical expression is too complicated and cumbersome. In order to more intuitively highlight its physical meaning, the following is an example of Figure 3 (b), using the Smith impedance circle diagram.

In Figure 3 (b), the component parameters are shown in the figure. The traces of the Rl, C6, and L4 branch impedances on the Smith chart are shown in Figure 4.

In Figure 4, when the frequency changes, the branch impedance moves on a resistance circle of Ι = 30 Ω, and as the frequency increases, it moves in a clockwise direction, exhibiting a reactance characteristic of positively variable characteristics. The branch reactance is 0 at the crossover frequency fo= _ _ ^l 125kHz of L4 and C6, as indicated by the triangle mark in Fig. 4.

After the branch is connected in parallel with L5, the impedance diagram of Ζ _如图 is shown in Figure 5.

The impedance map of Ζ _依然 is still circular, but has moved to the upper left of the Smith chart, and the diameter of the circle is reduced, as shown in the figure. As the frequency increases, the impedance point remains in a clockwise direction along the small circle.

In the LQ segment of the small circle, when the frequency increases, the moving direction of the impedance point is L→Q→H, which is clockwise for the small circle, but for the adjacent resistance circle (white dotted line) in the Smith chart. In this case, this direction is counterclockwise. That is to say, in the LQH segment, when the frequency is increased, the inductive reactance portion of Z _AB is instead reduced, and its reactance exhibits a negative micro-variation characteristic. In this example, f _L = 1087 kHz, f _H = 1132 kHz, f _Q = 1109 kHz (center frequency).

1 is a schematic structural diagram of an embodiment of the present invention. Rl and jXl are load impedances, and three serial arm units are L1, Cl, C4, L2, C2, C5, and L3, C3, and L4, respectively. In the figure, A, B, C, D, E, and F are test reference points. String arm The unit is mainly used to adjust and compensate the reactance micro-variation parameters, and the main function of the arm is impedance matching (the last stage is the arm such as L4) and the negative reactance of the circuit is biased into the central frequency symmetry. Levels and arms such as C4, C5)

All the steps of the mid-wave-tunnel network design using the method of the present invention are given below. The medium wave sky tone network designed according to the method of the present invention has a general structure as shown in FIG. 6. In FIG. 6, R, L0, and CO are antenna equivalent impedances, L5 and L6 are preconditioning networks, and multi-frequency common tower blocking units. Used to block other common tower frequencies except the local frequency, from blocking fl (consisting of L7, C7) to blocking fii (consisting of Lk, Ck), T, Q are the test reference points of the circuit, and other parts are the same as in Figure 1. .

The following method requires computer simulation software to participate in the auxiliary design, taking the M _U ltisim2001 circuit simulation software (this software can be easily downloaded via the Internet) as an example. The network analyzer in the software is set to: normalized impedance value 50Ω, scan width set to about 200~400kHz around fo, linear frequency sweep, take the corresponding number of scan points, so that the minimum frequency increment during observation is lkHz.

Let the impedance of the antenna at the carrier center frequency fo (kHz) be two +^^. ( Ω). Within the range of ±10 kHz, reactance X. It has a positive micro-variation characteristic, and its average rate of change with frequency is X _WQ ( Ω / kHz), and the characteristic impedance of the transmission cable is Ζ ₀ (Ω).

Step 1: Calculate the equivalent circuit model of the microvariable parameter of the antenna load.

Use X. And X _w . Find L0, COo

The above equations can be used to find XLO, XCO (unit: Ω), where f _Q is in kHz, X _WQ is in Ω / kHz, and Xo is in 0. Further find

L0= -^; CO ¹

² o ²⁷ o ^X co

Step 2: Determine the antenna pre-tuning network component parameters.

When Xo is capacitive reactance and is above 150 Ω, 5=^; otherwise, if £5 = 0, L5 is canceled.

² o

The value of L6 must meet the following requirements:

1) Whether it is a single-frequency single tower or a multi-frequency common tower, the real part of the impedance seen from the T point to the antenna is 20 Ω or more (if the real part R of the original antenna impedance is less than 20 Ω, it is as close as possible to the original antenna impedance. Real);

2) If it is a multi-frequency common tower, make the real part of the impedance of all frequencies at point T as close as possible.

3) When the frequency changes at the carrier frequency ±10 kHz, the real change of the impedance seen from the T to the antenna is as small as possible. As shown in Fig. 6, the value is generally 5 to 50 μΗ. , Step 3: Determine the component parameters of the blocking unit when multi-frequency common tower. As shown in FIG. 6, each blocking unit is composed of a parallel circuit of Lk and Ck, and Lk and Ck are connected in parallel to the blocked common tower frequency fn, wherein the value of Lk is generally 5 to 30 μΗ.

The larger the value of Lk, the higher the blocking reactance value of the Lk and Ck parallel circuits to fn, and the better the effect.

The value of Lk must meet the following principles:

1) For the two-frequency common tower, the blocking factor of the blocking unit at the two frequencies of positive and negative 9 kHz is not less than the common frequency for the three-frequency and four-frequency common towers, which is not less than 1.5 kQ; , this value is not less than 2kQ.

A 1 o n

2) Make "~~ ^ 2.5kHz. The real part of the impedance at the Q point at the operating frequency f _Q , . The average rate of change of the imaginary part of the zero point impedance at fo ± 10 kHz, in Ω / kHz.

Step 4: Determine C l and coarsely select the value of L1.

Let the measured impedance of Q point at fo be ZQ = RQ + jX _Q , and the average microvariable of XQ within 10 kHz around f _{Q be} determined by the following equation:

0.1 e B ( 1 )

2X, Cl

X, WQ +■

f.

XLI = XCI - XQ + 10 ( 2 ) Then find: Cl= LI-

B takes 2.4 to 3, fo is in kHz, X _W Q is Ω / kHz, and RQ and XQ units are Ω.

Step 5: Find the exact values of Ll and C4.

L1 value taken by the step 2, small-scale adjustment, while selecting C4 (f _Q in the capacitance is generally 5~15Ω) the value of the real part is maximized at _Beta [zeta] f _Q when impose 6 ~8Ω.

Adjustment rule: the larger the capacity C4, the smaller the maximum value of the real part of [zeta] _Beta. For a certain value of C4, fine-tuning L1 can make Ζ _{Β the} real part at ft maximum. The frequency scan obtained by Z _B in the simulation network analyzer is shown in Fig. 7.

After adjustment, f _Q should be at the triangle mark in the upper right corner of the circular path in the figure, as shown in Figure 7. This is also the tangent point between the track circle and one of the resistance circles of the Smith chart (the resistance circle in Figure 7 is R = 0.15, normalized resistance).

Step 6: Determine C2 and rough select the L2 value.

First, Z _B = R _B + jX _B when the frequency is fo, and the absolute value X _WB of the average micro-variation parameter of X _B in the range of f _Q positive and negative 5 KHz. The reactances X _C 2 and XL ₂ at a frequency of fo are determined by the following equation: L2 = c2 ~XB + 10

The unit of kHz of fo, the unit of X _WB is Ω/kHz, n is the compensation coefficient, and the empirical value range is 1.3~1.5. The obtained X< XL ₂ unit is Ω.

Step 7: Find the exact values of L2 and C5.

The capacity of C5 is estimated according to the capacitive reactance of 7~10Ω (at f _Q frequency), and the intermediate value is used as the test value. When the other parameters are not moving, fine-tune the L2 inductance so that the frequency change trajectory of Z _D is as shown in Fig. 8.

The trajectory should present a double peak pointing to the upper right, and the double peak is just tangent to the same Smith resistance circle. The bottom of the valley between the two peaks is the impedance point corresponding to fo. The real part of the impedance should be between 8 and 15 Ω. Otherwise, the capacitance of C5 should be changed and the adjustment process of L2 should be repeated until the requirements are met.

If the value of η is relatively large in step 6, the tuned Z _D trajectory map is the overcompensation mode, see Figure 9. However, if the value of n is small in step 6, the adjusted Z _D track map is the under-compensation mode, see Figure 10. The critical compensation mode can obtain the maximum bandwidth, but the out-of-band attenuation performance is not the best; the over-compensation mode can obtain the best out-of-band attenuation performance, but the bandwidth is not the largest; the under-compensation is the mode that should be avoided. Both the available bandwidth and the out-of-band attenuation performance are worse than the first two.

Step 8: Determine the exact value of L4, and the values of L3 and C3.

After completing step 7, it is measured at f. Where Z _D = R _D + jX _D (Ω).

As long as C3 and L3 get one of the parameters, the other can be obtained from the above equation. The following is the debugging method: Take C3 C2, and take several values every 50~100pf around it, and calculate the corresponding L3, get the test data of several sets of C3, L3, then input the simulation diagram, and finally take the largest bandwidth. That group. At this time, the trajectory of Z _F is shown in Figures 11 and 12.

Let Z _F = R _F + jX _F , and as the transmission line (characteristic impedance Z _Q ), the reflection coefficient and standing wave ratio caused by the terminal load are r and P, respectively.

The above relationship can be used to determine the modulation impedance trace map to achieve a bandwidth of P 1.2 of f _Q ±9 kHz.

In the actual debugging process, it is necessary to adjust it several times between steps 7 and 8, in order to achieve the desired effect.

A design example is given below.

Example 1. An iron tower antenna with a side width of 50 cm and a height of 76 m has an impedance of 15 _"' 173 (; Ω ) at a frequency of 558 kHz, and the imaginary capacitance of each 1 kHz impedance is reduced by 0.5 Ω at a frequency of around 558 kHz.

See Figure 13 for the design results of this example.

Comparing the general form of Figure 6, Figure 13 shows that there are fewer blocking units because there is no common tower frequency.

In Figure 13, the characteristic bandwidth of Ζ _被 is set to 3.1 kHz, the critical compensation mode, the total bandwidth of the modulation is slightly larger than ± 11 kHz, and the frequency of the carrier frequency (558 kH _Z ) is above 28 kHz, and the attenuation is greater than 20 dB.

Example 2. An iron tower antenna with a side width of 50 cm and a height of 64 m, and a 900 kHz and 819 kHz two-frequency common tower. The antenna impedances are: 31. 29.07 ( Ω ) and 24.00-j68.01 (Q). The antenna impedance varies little within the carrier frequency ± 10 kHz.

The 900kHz design parameters are as follows:

The antenna equivalent impedance CO is 6083pF, R is 31.1Ω; blocking 819: C7 is 4000pF, L7 is 9.44μΗ, L6 is 7μΗ; L1 is 44.5μΗ, C4 is 10000pF, C2 is 600pF, L2 is 56.55μΗ, C5 is 23000pF , C3 is 667pF, L3 is 44.8μΗ, L4 is 4.01μΗ; inlet F: 900kHz.

Compared with the general form of Fig. 6, Fig. 14 omits L0. Since the antenna impedance changes little around the carrier frequency, the reactance component in the equivalent antenna can be directly used with C0. Since the antenna capacitive reactance is small, L5 is omitted; C1 is also omitted because the above-mentioned C1 capacitive reactance calculated in step 4 is very small.

In the figure, the characteristic bandwidth of Z _A is set to 2.47 kHz, and the reactance values of the blocking unit at 900 kHz plus or minus 9 kHz (ie, 89 lkHz, 909 kHz) are around 2.2 kΩ, and the critical compensation mode is adjusted to a total bandwidth slightly larger than ± 10 kHz. The frequency of the carrier frequency above 39 kHz, the attenuation is greater than 20 dB.

Example 3. An iron tower antenna with a side width of 50 cm and a height of 64 m. 6 frequency common tower. The common tower frequency is (unit: kHz) 567, 747, 900, 1080, 1296, 1557.

The design of 1296kHz is as follows:

In Fig. 15, the value of L6 is large, which is selected according to the requirement of step 2 and satisfying the six common tower frequencies at the same time; since it is a 6-frequency common tower, five blocking units are provided. The blocking reactance value of all common tower frequency bandwidths (±9kHz) is above 4kQ; the characteristic bandwidth of Z _A is set to 2.42kHz, the critical compensation mode, the total bandwidth is slightly larger than ± l lkHz, and the carrier frequency is above 36kHz. At the frequency point, the attenuation is greater than 20dB.

Claims

Claim

1. A lumped parameter rectangular bandpass filter, characterized in that at least two stages of parallel arm units are provided, and each stage of the parallel arm unit is composed of one string arm and one parallel arm, and the string arm is composed of a capacitor and an inductor string. Connected, and the arm is a reactive component.

2. The lumped parameter rectangular band pass filter according to claim 1, wherein the capacitance C in the string arm of the first stage parallel arm unit is 200 to 2000 pf, and the inductance 1 is 1 to 10 (^11) The arm has a reactance value of 5 to 150 Ω.

3. The lumped-parameter rectangular band-pass filter according to claim 1, wherein the capacitance C in the string arm from the second-stage string parallel arm unit is 100 to 5000 pf, and the inductance 1 is 1 to 10 ( ^11, the arm has a reactance value of 5 to 200 Ω.

4. The lumped parameter rectangular band pass filter according to claim 1, wherein the parallel arm is a reactance element, the reactance element is an inductor coil or a capacitor, or the capacitor is connected in series with the inductor coil.

5. The lumped parameter rectangular band pass filter according to claim 1, wherein the arm of the last stage parallel arm unit is an inductor, and the inductance of the inductor is 2 to 12 [mu].