KR20030069140A - Unequal Wilkinson Power Dividers using Simple Rectangular Defected Ground Structure - Google Patents

Abstract

PURPOSE: An unequal Wilkinson power divider is provided to increase a characteristic impedance more than a standard microstrip line by inserting a defected ground structure in a ground surface of a microstrip transmission line. CONSTITUTION: A defected ground structure of a quadrangle is placed at a ground surface which is put below a signal line(4) of a microstrip transmission line so as to have a characteristic impedance more than a standard microstrip line. A plurality of small-sized defected ground structures are used instead of a large defected ground structure. An etching structure of the ground surface is one selected from a group of a circle, a triangle, an eddy type, and a geometric shape.

Description

Unequal Wilkinson Power Dividers using Simple Rectangular Defected Ground Structure

The limit of reliable implementation of a typical MS transmission line is known to be 100 ~ 120 kHz. The existing asymmetric Wilkinson power divider using a standard MS transmission line had a characteristic impedance of 103 kW even though N = 2 (1: 2 power divider), which was not easy to implement. Since N = 3 requires a characteristic impedance of 132 kHz, an asymmetric Wilkinson power divider with N ≧ 3 has been known to be practically impossible to implement. For the characteristic impedance of 120 kHz or more, the line width of the MS transmission line was extremely thin, exceeding the implementation limit, and thus the line width error of the implemented transmission line was so large that mass production was impossible.

Therefore, in order to implement an asymmetric Wilkinson power divider with N≥3, a 90 ° branch hybrid coupler, a 180 ° ring hybrid coupler, a directional coupler, and the like have been used. However, since the basic function is a coupler circuit rather than a divider, and thus the characteristics of the coupler are applied to an asymmetric divider, the divider has the following fatal problems.

The first problem with applying a hybrid coupler as an asymmetric splitter is that the phase difference between the two splitter outputs is large. This occurs because the coupler itself has a structure in which a phase difference of 90 degrees or 180 degrees can occur. The phase difference between the two outputs can cause a very big problem in some circuit applications in asymmetric dividers, and when applied to balanced circuits, the reverse phase difference must be compensated for.

The second problem is that in the case of directional couplers, the two output terminals are DC-coupled from each other, making them unusable in asymmetrical divider applications in balanced circuits where DC current must flow simultaneously. Therefore, in this case, a hybrid coupler was used. At this time, the above-mentioned phase difference problem occurs, and thus, there was a problem that the effort to compensate for the reverse phase difference later was required.

In the present invention, a very simple rectangular DGS is etched into the ground plane of the MS transmission line to implement an MS transmission line having a characteristic impedance of 200 Ω or more, which far exceeds the general implementation limit. We present a method for designing a 1: N high ratio asymmetric Wilkinson power divider with N> 3, which has been considered. In the present specification, a 1: 6 splitter is used as an example, and the actual measured results are presented. The characteristic impedance of 207 필요한 is required to be implemented as an MS transmission line using DGS.

1 is a side structure diagram of a general printed circuit board (PCB) for implementing a microstrip (“MS”) transmission line.

2 is a three-dimensional structural diagram of a standard MS transmission line.

FIG. 3 shows a simple rectangular defect ground structure (DGS) etched into the ground plane of the MS transmission line.

[Fig. 4] is a diagram showing the rectangular DGS and MS transmission lines etched on the ground plane shown in [Fig. 3] as if viewed from above.

FIG. 5 is a simulation of an MS transmission line (hereinafter referred to as "DGS MS transmission line") having a rectangular DGS on the ground plane, as shown in FIG. 4, using an electromagnetic simulation tool. The results are expressed in S-parameters.

FIG. 6 illustrates the actual fabrication of the DGS MS transmission line shown in FIG. 4 and measurement of S-parameter characteristics.

7 is a transmission line model set to obtain a changed characteristic impedance of a DGS MS transmission line.

8 is a theoretical circuit diagram of a typical 1: N asymmetric Wilkinson power divider.

FIG. 9 is a circuit diagram of a practical 1: N asymmetric Wilkinson power divider in which the theoretical circuit of FIG. 8 is converted to the desired characteristic impedance Z _o to enable actual measurement.

Table 1 summarizes the circuit components of a 1: N asymmetric Wilkinson power divider from N = 1 to N = 6.

(A) and (b) are representative views. First, (a) is a manufactured circuit diagram of a 1: N asymmetric Wilkinson power divider including a large rectangular DGS in the ground plane. The pattern shown in FIG. 4 was used here, and N = 6 was taken as an example of the distribution ratio. (Ie 1: 6 asymmetric Walkinson power divider) [Fig. 10] (b) shows that the same effect can be obtained by dividing the large square DGS into several smaller square DGS.

FIG. 11 illustrates the theoretical performance of a 1: 6 asymmetric Wilkinson power divider using a circuit performance analysis tool (Suparameter). Since it is theoretical, it is the ideal performance.

FIG. 12 shows the S-parameter by actually fabricating a 1: 6 asymmetric Wilkinson power divider and measuring its performance.

FIG. 13 shows the phase at the two output terminals of a theoretical 1: 6 asymmetric Wilkinson power divider. Since it is a theoretical circuit, it is the ideal case where the phase difference is zero.

FIG. 14 shows the phase difference at the two output terminals of a 1: 6 asymmetric Wilkinson power divider actually fabricated and measured.

FIG. 1 is a side view of a typical printed circuit board (PCB) used to implement a microstrip (MS) transmission line. The upper conductor surface 1 and the lower ground conductor surface 3 are distributed in parallel with the dielectric 2 interposed therebetween. A signal line of a microstrip transmission line is implemented on the conductor surface 1, and a signal line having a certain characteristic characteristic impedance is implemented to have a constant line width. In the figure, T is the thickness of the upper and lower conductor surfaces (1 and 3), and H is the thickness of the dielectric (2). The dielectric 2 serves as a path that pulls the propagated signal energy toward itself as much as possible and delivers it in the direction of the signal's progress. The dielectric constant is expressed by ε _r and has a size of 1 in air only, and greater than 1 in all other materials. (In the present specification, a substrate having ε _r = 2.2, T = 0.036 mm, and H = 0.79 mm will be described as an example.)

FIG. 2 is an MS transmission line composed of the PCB substrate of FIG. 1. After determining the line width to have a certain characteristic characteristic impedance, the MS transmission line is realized by leaving only the signal line pattern 4 of the desired line width from the upper conductor surface 1 and removing the rest.

FIG. 3 is a view showing a state in which a simple rectangular defect ground structure (DGS) is removed by etching the bottom ground plane 3 of the MS transmission line of FIG. 2. It is a state where the conductor surface of the large grounding conductor surface 3 of the bottom was removed as much as the area of the large square 5. In the present specification, this is referred to as "DGS MS transmission line" for convenience. If there is no DGS on the ground plane, it is a standard MS transmission line.

FIG. 4 is a drawing drawn as viewed from above for convenience in explaining the structure of FIG. 3. WM is the line width of the MS transmission line implemented on the upper conductor surface 4, and W1 and W2 are the dimensions of the rectangular DGS (5) removed by etching from the bottom conductor surface. In this specification, a 1: 6 asymmetric Wilkinson power divider is designed and presented as an example at a center frequency of 1.5 GHz, wherein the determined dimensions are WM = 0.4mm, W1 = 22mm, and W2 = 12mm. Dimensions of WM = 0.4 mm correspond to 120 Ω characteristic impedance in standard MS transmission lines without DGS in the ground plane when using a substrate with ε _r = 2.2, T = 0.036mm, H = 0.79mm to be. The WM in a standard MS transmission line to achieve the characteristic impedance of 207 kHz required in a 1: 6 asymmetric Wilkinson power divider is 0.035 mm, a dimension that cannot be reliably implemented, and therefore not mass-produced.

[Figure 5] is a diagram showing the characteristics of the electrical characteristics of the DGS MS transmission line of Figure 4 by the S-parameter analysis. In the standard MS transmission line, S21 is close to 0dB and S11 is less than -20dB because of excellent transmission and reflection coefficient characteristics. However, since the DGS MS transmission line of FIG. 4 deliberately adds DGS to the ground plane, the S11 exhibits near total reflection at a specific frequency (for example, 1.5 GHz in this specification). (Around 0dB), S21 also has very poor transmission characteristics. These bad characteristics are deliberately induced to obtain high characteristic impedance due to DGS.

FIG. 6 is a diagram showing the characteristics of a DGS MS transmission line having an impedance of 207 kHz as shown in FIG. It can be seen that it is very similar to the electromagnetic analysis characteristics of FIG. 5, and a deliberately induced bad transmission characteristic (S21) is shown.

7 is a transmission line model for obtaining a new characteristic impedance of an MS transmission line when a DGS is inserted. DGS added to the ground plane equivalently reduces the capacitance (C) and at the same time greatly increases the inductance (L), which is one of the characteristic impedance expressions of the transmission line. The characteristic impedance is greatly increased by the relation of.

The increased impedance is denoted Z _DGS in FIG. 7. Z _o is typically 50 Ω as port impedance or termination impedance. The DGS MS transmission line shorted with Z _o has a characteristic impedance of Z _DGS , where the impedance seen at the input is expressed as Z _in .

We will now explain how to find the increased characteristic impedance. The S11 value, which represents the reflection coefficient characteristic of the transmission line, is the size of the actual reflection coefficient ) And equation (1). The magnitude of the reflection coefficient and Z _in are related to Eq. (2). Therefore, the characteristic impedance Z _DGS of the DGS MS line can be finally obtained by using a very simple λ / 4 transformer equation such as Equation (3).

5 and 6, S11 is -1 dB at the center frequency of 1.5 GHz, which is calculated using Equation (1). Is 0.891. Since Z _o is 50 mW, the final calculated Z _DGS is 208 mW. Importantly, as mentioned above, when WM = 0.4mm, the characteristic impedance of the standard MS transmission line is 120 Ω, but the characteristic impedance is increased to 208 Ω by inserting the rectangular DGS. If the standard MS transmission line implements 208Ω, WM = 0.035mm, which is impossible to implement.

8 is a theoretical circuit diagram of a typical 1: N asymmetric Wilkinson power divider. The input power incident to terminal 1 (P1) is divided into terminal 3 (P3) and terminal 2 (P2) at a ratio of 1: N and appears as an output. To be the most basic bipolar power divider (1: 1), N = 1.

9 is a practical circuit diagram of a 1: N asymmetric Wilkinson power divider. In FIG. 8, impedances R2 and R3 of the output terminals vary according to arbitrary N. In general, characteristic impedances of general measuring instruments, circuits, and systems are unified to Z _o = 50 kPa for the convenience of actual use or measurement. Therefore, if R2 and R3 having arbitrary values are converted to λ / 4 by Z _o = 50 kV through equations (4) and (5), the output terminal impedance is Z _o = 50 kW, as shown in [Fig. 9]. The actual 1: N asymmetric Wilkinson power divider is designed.

Table 1 shows a 1: N asymmetric Wilkinson power divider with the values of the circuit elements up to N = 6. N can be raised in the same way, but in the present specification, an asymmetric splitter in the case of N = 6 is designed and manufactured to introduce measurement results.

Figure 10 (a) is a layout of a 1: 6 asymmetric Wilkinson power divider with a center frequency of 1.5 GHz, designed for example purposes. One simple square DGS mentioned in FIG. 4 was used, and the characteristic impedance of each line of the 1: 6 asymmetric Wilkinson power divider shown in Table 1 was implemented. The large square DGS of FIG. 10 (a) can also be replaced by a number of small square DGSs as shown in FIG. 10 (b). At this time, the spacing between the rectangles and the size of the rectangles are adjusted to show the desired electrical characteristics as shown in FIG. 5 or 6.

FIG. 11 shows the theoretical performance of a 1: 6 asymmetric Wilkinson power divider calculated using a circuit performance analysis tool. Because of the 1: 6 distribution ratio, the input power incident on terminal 1 must be delivered 1/7 to terminal 3 and 6/7 to terminal 2. Therefore, theoretically, S31 = -8.45dB at terminal 3 and S31 = -0.67dB at terminal 2. At this time, the matching characteristics of the three terminals S11, S22 and S33 and the isolation characteristic S32 (or S23) between the terminals 2-terminal 3 are very excellent in a certain band including the center frequency.

FIG. 12 shows the performance of a 1: 6 asymmetric Wilkinson power divider actually fabricated and measured. In the actual fabrication process, there are some fabrication errors that cannot be considered in the theoretical circuit, so that the measurement results show some frequency shift from ideal performance, but the difference is negligible. Since the measured performance is very consistent with the theoretical characteristics as a whole, the design method of the high-rate asymmetric Wilkinson power divider using DGS, which is presented here, is very valid.

13 shows the phases of two output signals. Theoretically, the ideal result shows no phase difference between the two signals.

14 is a phase difference between two output signals actually measured. Although there may be an error in the manufacturing process or a measurement terminal reference setting in the measurement process, the ideal value is not 0, but the phase difference is very good at about 5 °. By making minor adjustments to the line length of the measuring terminals, the phase difference can easily be made close to zero.

One of the most widely used methods to implement an asymmetric power divider is to apply a 90 ° or 180 ° hybrid coupler as an asymmetric power divider. Since the phase difference between the two outputs is basically large, it is necessary to compensate for it later by making an inverse phase difference. There was a problem.

Another widely used method was to apply directional couplers as asymmetrical power dividers. In this case, since the two output terminals are DC-blocked to each other, it is inconvenient to connect the DC supply circuits to the two terminals when an active circuit requiring DC supply is connected to the two output terminals.

Recently, in designing high efficiency linearizers and power amplifiers in various mobile communication systems, broadcasting systems, and other communication systems, there are many phase ratio-free, high ratio asymmetric power dividers with DC current flowing. I am doing it.

The high-rate asymmetric Wilkinson power divider using the simple square DGS proposed in the present invention has a structure that can solve the problems of the above-mentioned coupler. That is, there is no phase difference between the two outputs, and the two output terminals are DC-connected, so that when applied to a high efficiency linearizer or a power amplifier, the system can be configured relatively simple and at low cost. Considering that the parts market for linearizers and power amplifiers is very large, the application of the method proposed in the present invention can produce a sufficiently large economic effect.

Claims (3)

An asymmetric Wilkinson power divider implemented with a power splitter microstrip, characterized by a square etched structure (DGS) on the ground plane below the signal line (4) of the microstrip transmission line, which is larger than a standard microstrip transmission line. Asymmetric Wilkinson Power Divider with Impedance The asymmetric Wilkinson power divider of claim 1, which is implemented using several smaller square DGSs instead of one large square DGS. The asymmetric Wilkinson power divider according to claim 1, wherein the etch structure of the ground plane is circular, triangular, eddy, or other geometric shape in addition to the quadrangle.