WO2003088410A1

WO2003088410A1 - Electric four-wire network with a transformation line

Info

Publication number: WO2003088410A1
Application number: PCT/DE2003/000950
Authority: WO
Inventors: Andreas Przadka
Original assignee: Epcos Ag
Priority date: 2002-04-18
Filing date: 2003-03-21
Publication date: 2003-10-23
Also published as: KR20040108743A; EP1495513A1; EP1495513B1; JP4058004B2; KR101025233B1; DE10217387A1; JP2005523598A; DE10217387B4

Abstract

A network is provided comprising a transformation line in order to electrically adapt an electric component which is embodied in or on a preferably ceramic substrate. Said component has a predetermined electrical length for obtaining a desired phase displacement and comprises a folded electrical conductor (LE), whereby the straight sections thereof (A) are connected together in a right-angled manner and the widths of individual conductor sections are selected differently for the compensation of couplings.

Description

description

Electrical adaptation network with a transformation line

Electrical components often require an electrical matching network to adapt to their circuit environment. Such a device can include inductors, capacitors and transformation lines and essentially serves to adapt the impedance of a component to the external environment. Such matching networks are often designed as passively integrated networks, in which the discrete elements forming the network are integrated together in a substrate, which preferably forms the carrier substrate for the component. It is even possible to form a ceramic component in a ceramic, in whose ceramic body or on the ceramic body the matching elements are applied and integrated with the component.

Electrical transformation lines as components of matching networks are often implemented in a multilayer ceramic substrate in which, as mentioned, other elements can also be integrated. Transformation lines are used, for example, in front-end modules for mobile communication devices, where they can be used as part of pin-diode switches and have to achieve a phase shift of approx. 90 °, for example. Furthermore, such a transformation line should have the best possible adaptation under the given source and load impedances. A further exemplary use can be found in a transformation line in a duplexer, which, likewise used in a mobile communication terminal, connects the antenna to both the transmitting and the receiving path of the terminal.

Another requirement for transformation lines, especially in mobile communication devices, is that requires little space and space. In the case of a front-end module, for example, the external dimensions are significantly smaller than the fraction of the wavelength in the ceramic substrate by which the phase shift is to take place. Since the phase shift can only take place with a conductor that has a certain geometric length, the transformation lines used today are unfolded and in some cases have multiple layers. Both the folding and the multi-layer design, which leads to overlap of conductor sections, result in capacitive and inductive couplings between different sections of the line. This leads to a change in the adaptation and to an additional phase shift compared to an ideal line of the same geometric length, which is of single-layer and unfolded design. In addition, the available area and the position of the connection points at which the transformation line is connected to the component or the further matching network cannot be selected arbitrarily, since they depend on the other components of the circuit parts to be integrated.

An exemplary embodiment of a transformation line is a so-called tri-plate line, in which a, for example, folded conductor is guided between two shielding ground layers, that is to say between two metallized levels, and is separated from each of these by a ceramic layer. The distance to the upper and lower shielding ground plane influences the characteristic impedance and is therefore chosen accordingly. Due to the technology and the need for integration with further elements in the common substrate, the thicknesses of the ceramic layers cannot be chosen arbitrarily, but must be selected from a limited number of available and suitable layer thicknesses, so that optimal adaptation is not possible in this way. In a space-saving known transmission line, the conductor is, for example, meandered and made of two layers. A symmetrical arrangement of the two levels in which the conductor runs is selected so that the characteristic impedance of the line is the same in the two conductor levels and corresponds to the impedance of source and load. The coupling between the individual sections of the conductor is minimized in that sections of the conductor lying in parallel have a sufficient distance from one another which is generally greater than the width of the conductor. The coupling between conductor sections in different conductor levels is reduced in that either superimposed sections in both layers are arranged at right angles to one another or by placing conductor sections of one conductor level between the projection of the conductor sections of the other level. The geometric length of the conductor can be increased to increase the phase rotation of the transmission line. If the area is limited, this is only possible by moving the individual sections of the conductor closer together. As a result, however, the coupling of the line parts to one another increases, the adaptation between source and load being deteriorated.

The object of the present invention is therefore to provide a network with a transformation line which is also suitable for further miniaturized components and with which a desired adaptation of ^, for example, better than 10 dB is achieved.

This task is accomplished through a network with the characteristics of

Claim 1 solved. Advantageous embodiments of the invention emerge from further claims.

The invention specifies a network which has a transformation line of a predetermined electrical length formed in or on a substrate. To make better use of the available for the transformation management The surface of the conductor is folded, the sections being rectilinear and being connected to one another at right angles. The resulting disadvantageous coupling of adjacent sections of the conductor is taken into account according to the invention in that the width of the conductor in the sections is designed differently. The inventors have recognized that the coupling can be influenced by deliberately changing the width of individual conductor sections, so that the desired adaptation can be achieved in individual sections by suitable choice of the conductor width. If, for example, two conductor sections that couple capacitively and inductively with one another are considered, the inductive coupling can be reduced, for example, by increasing the conductor width in one of the two conductor sections. By increasing the conductor width in one section, the parasitic and intrinsically disruptive capacitive coupling to adjacent conductor sections can also be increased. The electrical adaptation of the transmission line can be improved by varying the conductor width of an individual conductor section. Through a suitable and independent selection of the widths of all conductor sections, the adaptation can be optimized and set exactly to a desired value. Conventional circuit environments may require adaptation to 50 Ω, for example.

The invention makes it possible in a simple manner to optimize the electrical adaptation of the transformation line and thus overall the network for adapting the electrical component exactly to the desired values without this leading to an increased area requirement of the transmission line. On the contrary, arrangements are also possible with the invention which, in known transmission lines, have led to unauthorizedly high couplings and thus to poor adaptation, but which are now compensated according to the invention.

This allows a further reduction in the area requirement of the transmission line and, alternatively or additionally, one geometric shape of the transmission line, which could not be realized without further disadvantages. In this way, a surface available on the substrate can be better utilized with the invention. An increased area requirement of the invention is excluded solely by the fact that the invention does not significantly change the geometrical and therefore generally also the electrical length of the conductor, which is largely responsible for the extent of the phase shift.

Section of the conductor is understood to mean any section of the conductor with a given length. As a rule, and both for the calculation and for the construction of the transmission line, it is easier to define sections between two corner points of the folded line.

Like the conventional transmission line, the transmission line according to the invention can also be designed with a conductor folded in two conductor levels. The two conductor levels are separated from one another by an insulator, preferably a ceramic layer. Another insulating layer, in particular another ceramic layer, separates the conductor planes from the shielding plane connected to ground.

The transmission line can also be designed as a tri-plate line, in which the conductor levels are arranged between two ground levels. With the invention it is possible to make the insulation layer that separates the two conductor planes thinner than in known transformation lines. The resulting disruptive couplings can be compensated for with the invention. The two parts of the conductor running in different conductor levels are connected to one another by plated-through holes.

In the two conductor levels, the sections are guided in such a way that there are no parallel sections in the two conductor levels come to lie on top of each other. Sections parallel to one another are offset from one another at least by a minimum length in the two planes. Crossings between sections in different conductor levels are preferably at a distance from the end of the section and preferably in the middle of the

Conductor portions. When varying the width of the conductors in individual sections, boundary conditions are advantageously maintained. In particular, the widths of the conductor sections as well as the spacing between parallel conductor sections should have a mostly technologically related minimum value, which is chosen, for example, at 100 μm. However, these minimum distances and minimum widths are not the subject of the invention, but are only taken into account as boundary conditions in the optimization process and are accordingly reflected in the precise design of the transformation line. Other boundary conditions and minimum values can also be observed.

The geometric length of the conductor of the transformation line is chosen so that its electrical length corresponds to a λ / 4 line. A λ / 4 line is required in many cases where the circuit state has to be changed from "SHORT" to "OPEN". However, the transformation control of a network according to the invention can bring about a phase shift deviating from λ / 4.

A preferred impedance adjustment is 50 Ω, since this value is required in many circuit environments. However, it is also possible to adapt the transformation cable and thus the network to other circuit environments that deviate from 50 Ω. The impedance can be adjusted in a tri-plate line by varying the distances between the shielding levels and the conductor levels. However, it is also possible, especially if the available layer thicknesses in a given substrate are not sufficient to set a desired impedance, an additional one perform separate impedance transformation and provide a corresponding element in the network.

The network according to the invention is preferably integrated in a multilayer ceramic, for example an LTTC

Ceramics that are optimized for minimal shrink, for example. Such a low-shrink ceramic in LTTC design (= low temperature cofired ceramic) allows a high level of integration of network elements and, if necessary, additionally the integration of the actual components in the ceramic, since with this technology a high-quality ceramic and low-loss metallic conductors are possible at the same time reproducible component geometry or network geometry can be obtained. Usually, however, the substrate of the network is the carrier substrate for the component on which it is attached and with which the component is contacted, for example in one step by means of an SMD process. If the component is a component working with acoustic waves, a flip-chip arrangement can be selected, for example.

The substrate for the network, which can be an integrated network, can simultaneously represent the substrate for a module in which several components and the associated network are integrated.

The invention and a method for optimizing a network according to the invention are explained in more detail below on the basis of exemplary embodiments and the associated figures.

FIG. 1 shows a schematic plan view of a conductor of a known transmission line folded in two planes,

FIG. 2 shows a schematic cross section of a transmission line designed as a tri-plate line, FIG. 3 shows a Smith diagram of a known transmission line,

FIG. 4 shows the conductor of a transmission line according to the invention in a schematic plan view,

Figure 5 shows the Smith chart of the transmission line according to the invention.

A known transmission line will be explained in more detail with reference to FIGS. 1 and 2. The figures are only for explanation and are not to scale. The known tri-plate arrangement consists of a first and a second conductor level LEI, LE2, which are separated from one another by a ceramic intermediate layer. Above and below the first and second conductor levels, a shielding level ME1, ME2 is also separated by a ceramic intermediate layer, for example a metallization level (see FIG. 2). The conductor levels and the shielding levels are preferably arranged symmetrically to one another, so that the distances of the shielding levels ME from the adjacent conductor level LE are uniformly equal to dE. The distance dE can differ from the distance dL of the two conductor levels LEI, LE2. In a known transmission line, for example, dE = 125 μm, while dL = 95 μm. Figure 1 shows the folding of the conductor LEI in the first conductor level and shown in dashed lines the projection of the folded conductor LE2 in the second conductor level. The conductor consists of straight sections that are assembled at right angles. The sections are arranged with respect to one another in the two conductor levels LEI and LE2 in such a way that rectilinear conductor sections which are parallel to one another do not come to lie one above the other. The two parts LEI, LE2 of the overall conductor are connected to one another in the two levels via the through-connection DK. At the two connection points T1 and T2, the conductor or the transmission line is connected to an external circuit environment, for example the network or a component connected. The conductor has a uniform width dO.

FIG. 3 shows the adaptation calculated from this known transmission line, shown in the Smith chart. The adaptation of the known transmission line is significantly worse than 15 dB, the impedance adaptation is approx. 35 Ω.

According to the invention, the width of individual conductor sections of one or both conductor levels LEI, LE2 is now varied and in particular increased. As a result, the coupling of the corresponding conductor sections AI to A6 with adjacent conductor sections of the same conductor level or the underlying conductor level LE2, not shown in FIG. 4, is reduced or changed in character. By widening a conductor section A, for example, the inductive coupling can be reduced, while the capacitive coupling can be increased. The widths of the conductor track sections d ₃ , d ₄ , d ₅ and d ₆ are given by way of example only for the corresponding conductor sections A3, A4, A5 and A6. With d ₀ a virtual "original" width of the conductor is indicated. An optimal adaptation of the conductor normally results in the widths d _{x of} all the varied conductor sections Ax assuming different values from one another. However, it is also possible for individual conductor sections to have the same width. This applies in particular to the conductor sections unchanged from the original structure. In FIG. 4 only the conductor level LEI is shown, the underlying second conductor level LE2 can and will be changed accordingly, so that conductor sections of different widths are also present there.

FIG. 5 shows the Smith diagram belonging to the transmission line shown in FIG. 4. A comparison with FIG. 3 shows that the electrical adaptation of the transmission line according to the invention is significantly improved. It is close to 50 Ω and has a phase shift of, for example, exactly λ / 4. The extent of the phase shift can however, they can be varied accordingly in one or both of the planes by increasing or decreasing the geometric and thus also the electrical length of the conductor. A phase shift by values deviating from λ / 4 is also possible.

The optimization procedure for adapting the transmission line according to the invention can be carried out as follows. A conductor with sections of uniform width is assumed and its electrical characteristics are calculated or simulated. The width of a section is then varied and the electrical parameters are recalculated. The effect thus achieved (= shift of the adjustment in the Smith chart as a vector) is saved as an adjustment measure for the varied section. Then, based on the start structure, another section in the

Width varies and the electrical parameters are recalculated. So you get another adjustment measure. Depending on the problem at hand and the effects achieved with the individual adaptation measures, a desired or required adaptation can possibly already be achieved with two adaptation measures, the effectiveness of which can be varied by interpolating the effect and accordingly changing the width of the respective section. For demanding adaptations, it may be necessary to calculate further adaptation measures for other sections or for all sections and to additively add the desired adaptation from the individual adaptation measures. Finally, further adjustments may be necessary for the structure obtained in this way, since the individually calculated adjustment measures can influence one another.

A network according to the invention with the novel transformation line can be used to adapt any electrical components. It is advantageously used for passively integrated networks, which are absolutely necessary for the further miniaturization of electrical components. A particularly advantageous use for the Network for the electrical adaptation of components of front-end modules in wireless communication devices, for example in cell phones. Here, the passive integration to achieve the desired or already achieved external dimensions must be integrated into the component substrate or the front-end module substrate.

In order to accommodate additional network components and to fulfill its function as a component substrate, the substrate is reinforced by additional layers compared to the layer sequences shown in FIG. 2. The thickness of the substrate or the number of layers required for this depends on the number of network elements and components to be integrated in the substrate. Depending on the component to be implemented in the substrate ceramic, the material for the corresponding ceramic layers is also selected.

In the present case, an electrically insulating ceramic is used for the intermediate layer between the two conductor levels LEI and LE2, the preferably low dielectric constant of which also determines the impedance of the line. A lower dielectric constant of the intermediate layer also reduces the coupling between the conductor levels. With the invention, however, such couplings can be reduced or used advantageously. The ceramic layers between a conductor level LEI and a shielding level ME1 connected to ground are also set in an electrically insulating manner, although the value of the corresponding dielectric constant must also be taken into account here. The same ceramic is usually used for all ceramic layers, including the intermediate layer. According to the invention, however, it is also possible to use a ceramic different from the other ceramic layers for the intermediate layer, in particular in order to set the coupling, which according to the invention may be desired again, to a desired value. The areas available for the individual components are generally determined by vias and other elements that are present or implemented on the same level. With the invention, a particularly good adaptation to an available, arbitrarily shaped surface can be realized.

Claims

claims

1. Network for the electrical adaptation of an electrical component, with a transformation line formed in or on a substrate of a predetermined electrical length in which the transformation line comprises a folded electrical conductor (LE), linear sections (A) being connected at right angles to one another and the width (d) the conductor is different in the sections.

2. Network according to claim 1, • • in which the width (d) of the conductor (LE) in the individual sections (A) is selected such that interfering couplings between different sections of the conductor are compensated and an impedance adaptation to the given environment of better than 25 dB is reached.

3. Network according to claim 1 or 2, wherein the folded conductor (LE) runs in a first plane, which is separated by at least one ceramic layer from a shielding plane connected to the first plane and connected to ground.

4. Network according to one of claims 1 to 3, wherein the conductor (LE) comprises two parts (LE1, LE2), which extend over two levels separated by a ceramic intermediate layer, the two parts via a via (DK) with each other are connected.

5. Network according to claim 4, wherein the sections (A) are guided in the two planes so that mutually parallel sections (A) do not lie one above the other and at least a minimum length are offset from one another.

6. Network according to claim 5, in which a minimum length of 100 μm is maintained both for the offset of mutually parallel sections (A) arranged in different planes and for the distance of adjacent and mutually parallel sections arranged within a plane.

7. Network according to one of claims 1 to 6, in which all sections (A) of the conductor at least one of

Minimum length corresponding width (d).

8. Network according to one of claims 3 to 7, in which the transformation line is designed as a triplate line with two shielding, ground-connected levels (ME), in which the two ceramic layers which are arranged between a conductor level and the shielding levels , have the same thickness (dE).

9. Network according to one of claims 1 to 8, in which the transformation line is designed as a λ / 4 line.

10.Network according to one of claims 1 to 9, in which the transformation line is adapted to 50 Ω.

11.Network according to one of claims 1 to 10, in which the impedance matching to the external environment is ensured with the aid of an additional element for impedance transformation.

12.Network according to one of claims 1 to 11, wherein the substrate is a multilayer ceramic and forms the carrier for a component or a module.

13.Network according to claim 12, wherein the component or the module comprises at least one component working with acoustic waves.