Ultra-wide band measurement bridge

Abstract

A measuring bridge (1) provides a first matching pad (2), a second matching pad (3) and a third matching pad (4), wherein all matching pads (2, 3, 4) comprise at least three resistors (2.sub.1, 2.sub.2, 2.sub.3, 3.sub.1, 3.sub.2, 3.sub.3, 4.sub.1, 4.sub.2, 4.sub.3) which are arranged in a T-structure. A second resistor (3.sub.2) of the second matching pad (3) is connected to a second resistor (2.sub.2) of the first matching pad (2), and a third resistor (4.sub.3) of the third matching pad (4) is connected to a third resistor (2.sub.3) of the first matching pad (2). A second resistor (4.sub.2) of the third matching pad (4) can be connected to a device under test (7). A third resistor (3.sub.3) of the second matching pad (3) can be connected to a calibration standard (5), and a first resistor (3.sub.1, 4.sub.1) of the second and the third matching pad (3, 4) are connected in each case to a signal input of an element (11) which suppresses a common-mode component on its two signal inputs.

Claims

1. A measuring bridge comprising: a first matching pad; a second matching pad; and a third matching pad, wherein all matching pads comprise at least three resistors, which are arranged in a T-structure, wherein a second resistor of the second matching pad is connected to a second resistor of the first matching pad and a third resistor of the third matching pad is connected to a third resistor of the first matching pad, wherein a second resistor of the third matching pad is configured to be connected to a device under test, wherein a first resistor of the second and the third matching pad are connected to two signal inputs of an element which suppresses a common-mode component on its two signal inputs, and wherein the measuring bridge comprises a bias unit which is connected to the second resistor of the third matching pad.

2. The measuring bridge according to claim 1, wherein a third resistor of the second matching pad is configured to be connected to a calibration standard.

3. The measuring bridge according to claim 2, wherein the element is a balun or a differential amplifier.

4. The measuring bridge according to claim 1, wherein a first resistor of a first matching pad is configured to be connected to a signal generator and/or wherein a second resistor of the first matching pad is combined with a second resistor of the second matching pad to form a first equivalent resistor and/or wherein a third resistor of the first matching pad is combined with a third resistor of the third matching pad to form a second equivalent resistor.

5. The measuring bridge according to claim 1, wherein the resistors of the first, second and third matching pads are embodied as thin-film resistors on a substrate, and/or wherein the substrate is a ceramic or quartz substrate.

6. The measuring bridge according to claim 4, wherein the calibration standard is formed by at least one thin-film resistor on the substrate and/or wherein a conductor track which connects the at least one calibration standard embodied as a thin-film resistor to the second matching pad provides a V-shaped recess at its end facing towards the calibration standard, which reduces a capacitive scattering field, and/or wherein the measuring bridge provides a first connector, via which the second matching pad is configured to be connected to a separate calibration standard, and wherein the first connector provides the same electrical properties as a second connector which connects the third matching pad to the device under test.

7. The measuring bridge according to claim 2, wherein the measuring bridge comprises a dummy-bias unit which is connected to the third resistor of the second matching pad and wherein the dummy-bias unit is constructed in the same manner as the bias unit, so that the measuring bridge is symmetrical.

8. A balun comprising: a first portion, wherein the first portion comprises a substrate on the upper side of which a first signal line and at least one second signal line are guided, and on the lower side of which a thin-film resistor layer is embodied beneath the signal lines, wherein the thin-film resistor layer provides a DGS structure which splits the thin-film resistor layer into two mutually separate thin-film resistor layers, wherein the width of the DGS structure is matched to the substrate material used and the frequency range to be covered, and wherein the DGS structure is embodied directly between the first signal line and the at least one second signal line.

9. The balun according to claim 8, wherein the balun comprises a second portion, wherein the second portion provides a coaxial line, of which the inner conductor is connected to the first signal line and of which the outer conductor is connected at a first end of the coaxial line to the second signal line, and wherein at least one first ferrite surrounds the coaxial line.

10. The balun according to claim 8, wherein the first signal line and the second signal line are arranged close to one another and/or that the substrate is a ceramic or quartz substrate.

11. The balun according to claim 8, wherein the DGS structure is embodied directly between the two signal lines on the lower side of the substrate.

12. The balun according to claim 9, wherein the outer conductor of the coaxial line is connected at a second end to the reference ground and/or wherein the balun provides a compensation line, of which the first end is connected to the first signal line and/or wherein the compensation line is a wire.

13. The balun according to claim 12, wherein the compensation line together with the coaxial line is surrounded by the at least one first ferrite or that the compensation line is surrounded by at least one further ferrite, wherein the at least one further ferrite provides the same properties as the at least one first ferrite and/or wherein a second end of the compensation line is connected to the reference ground.

14. The balun according to claim 12, wherein the end of the coaxial line which is distant from the first portion of the balun is connected to a first cup core, wherein the first cup core comprises a coaxial line coiled within a ferrite or wherein the coaxial line is coiled within the first cup core and/or wherein the compensation line is connected to a second cup core, wherein the second cup core comprises a line coiled within a ferrite, or wherein the compensation line is coiled within a second cup core and wherein the second cup core provides the same electrical properties as the first cup core.

15. The measuring bridge according to claim 1, wherein a balun according to any one of claims 8 to 14 is inserted into the measuring bridge as a common-mode suppressing element.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Various exemplary embodiments of the invention are described in the following by way of example with reference to the drawings. Identical subject matters provide the same reference numbers. In detail, the corresponding Figs. of the drawings show:

(2) FIG. 1A an exemplary embodiment of an equivalent circuit diagram of the measuring bridge;

(3) FIG. 1B a further exemplary embodiment of an equivalent circuit diagram of the measuring bridge;

(4) FIG. 1C a further exemplary embodiment of an equivalent circuit diagram of the measuring bridge;

(5) FIG. 2A an exemplary embodiment for the realisation of three matching pads for the measuring bridge;

(6) FIG. 2B an exemplary embodiment for the realisation of a calibration standard of the measuring bridge;

(7) FIG. 2C an exemplary embodiment of the measuring bridge according to the invention which is suitable for connection to a separate calibration standard;

(8) FIG. 3 an exemplary embodiment of a chocked balun;

(9) FIG. 4 an exemplary embodiment of a resistive balun;

(10) FIG. 5 a presentation which describes the field distribution for common-mode signals and odd-mode signals in a resistive balun, according to one embodiment;

(11) FIG. 6 an exemplary embodiment of a resistive balun with a DGS structure;

(12) FIG. 7 an exemplary embodiment of a combination of a balun with ferrites and a compensation line and a resistive balun;

(13) FIG. 8 an exemplary embodiment of a combination of a balun with ferrites and a compensation line and a resistive balun;

(14) FIG. 9 a further exemplary embodiment of a combination of a balun with ferrites and compensation line and a resistive balun with a DGS structure;

(15) FIG. 10 an exemplary embodiment of a combination of a balun with ferrites and a compensation line and a resistive balun with a DGS structure;

(16) FIG. 11 an exemplary embodiment of a combination of a balun with ferrites and a compensation line and a resistive balun with a DGS structure together with cup cores; and

(17) FIG. 12 an exemplary embodiment of a separated guiding of a coaxial line and a compensation line through spatially separated ferrites.

(18) FIG. 1A shows an exemplary embodiment of an equivalent circuit diagram of the measuring bridge 1. The measuring bridge 1 provides a first matching pad 2, a second matching pad 3 and a third matching pad 4. The first matching pad 2 comprises at least three resistors 2.sub.1, 2.sub.2, 2.sub.3, which are arranged in a T-structure. The second matching pad 3 also comprises at least three resistors 3.sub.1, 3.sub.2, 3.sub.3, which are also arranged in a T-structure. The same applies for the third matching pad 4 which also comprises at least three resistors 4.sub.1, 4.sub.2, 4.sub.3 which are arranged in a T-structure. In this context, the second resistor 2.sub.2 of the first matching pad 2 is connected to the second resistor 3.sub.2 of the second matching pad 3. The third resistor 2.sub.3 of the first matching pad 2 is connected to the third resistor 4.sub.3 of the third matching pad 4.

(19) The third resistor 3.sub.3 of the second matching pad 3 is connected to a calibration standard 5. This calibration standard 5 may be embodied either directly within the measuring bridge 1 or it can be connected to the measuring bridge 1. In the latter case, the measuring bridge 1 provides a first connector 6 which can be, for example, a connector socket, to which the separate calibration standard 5 is connected, especially screw-connected.

(20) At the other side, the second resistor 4.sub.2 of the third matching pad 4 is connected to the device under test 7 to be measured. In this context, the device under test 7 is preferably screw-connected to a second connector 8, wherein the second connector 8 is preferably embodied as a connector socket and connected to the second resistor 4.sub.2 of the third matching pad 4.

(21) The first resistor 2.sub.1 of the first matching pad 2 is connected to a signal generator 9. This signal generator 9 can be embodied within a network analyser and connected via a third connector 10 to the first resistor 2.sub.1 of the first matching pad 2. The third connector 10 is also preferably a connector socket. However, it is also possible for the measuring bridge 1 to be integrated directly as a plug-in card within a network analyser, so that the third connector 10 is only a type of plug contact or inter-substrate connector.

(22) The resistors of the first matching pad 2, the second matching pad 3 and the third matching pad 4 may be selected in such a manner that the signal generator 9 sees a matched load. Furthermore, in the direction of the calibration standard 5 (Match), the second matching pad 3 must provide an impedance which corresponds to that of the calibration standard 5, so that no further reflections occur. The same also applies for the third matching pad 4, of which the resistors are selected in such a manner that, with a device under test 7 connected which provides the same impedance as the calibration standard 5, no additional reflections occur. For a 50 ohm system, it is the case, for example, that all resistors in the first matching pad 2, the second matching pad 3 and the third matching pad 4, provide a resistance value of approximately 16.78 ohms.

(23) Furthermore, the first resistor 3.sub.1 of the second matching pad 3 and the first resistor 4.sub.1 of the third matching pad 4 is connected to an element 11 which suppresses a common-mode component on its two signal inputs. The element 11 suppressing the common-mode component in the exemplary embodiment from FIG. 1A is a balun 11.sub.1. It is clearly evident that the first signal line 12.sub.1 connects the inner conductor of the balun 11.sub.1 to the first resistor 3.sub.1 of the second matching pad 3. At the other side, a second signal line 12.sub.2 connects the first resistor 4.sub.1 of the third matching pad 4 to an outer conductor of the balun 11.sub.1. The balun 11.sub.1 can be constructed, for example, very simply from a coaxial cable which is surrounded by ferrites. The end of the balun 11.sub.1 facing away from the measuring bridge 1 is connected to a fourth connector 13 to which the electronic measuring device of the network analyser, especially various mixer units and/or analog-digital converters, are connected. The outer conductor of this end of the balun 11.sub.1 is therefore connected to the reference ground. The balun 11.sub.1 converts a differential signal into a ground-referenced signal.

(24) For the case that the device under test 7 terminates the connection between the second resistor 4.sub.2 and the device under test 7 in the same manner as the calibration standard 5, the signals on the first signal line 12.sub.1 and the second signal line 12.sub.2 are of equal magnitude in their amplitude and in phase. The balun 11.sub.1 causes this common-mode component to be suppressed, so that no voltage relative to the reference ground can be measured at the fourth connector 13. Something different occurs if the second connector 8 is not terminated in the same manner by the device under test 7 as by the calibration standard 5. In this case, the signals on the first signal line 12.sub.1 and the second signal line 12.sub.2 are not exactly identical, so that the odd-mode component can be measured relative to the reference ground at the fourth connector 13.

(25) This construction allows the reflection behaviour of the device under test 7 to be measured without difficulty. As already explained, the calibration standard 5 can also be arranged within the measuring bridge 1, so that the first connector 6 is not required. In this case, however, the measuring bridge 1 provides a slight imbalance, which must be determined by means of a calibration process, because the device under test must continue to be connected via the second connector 8 to the measuring bridge 1. In order to realise the maximum possible bandwidth, the third resistor 3.sub.3 of the second matching pad 3 may therefore preferably be connected to a first connector 6. In this context, the first connector 6 should provide the same electrical properties as the second connector 8. The attainable insulation between the generator path (signal generator 9) and measurement path (connector 13) is accordingly determined by how well the connection for the device under test 7 can be balanced via the calibration standard 5 in the case of “match”.

(26) As will be explained later, the resistors of the first, second and third matching pad 2, 3, 4 are preferably thin-film resistors which are embodied on a substrate 25, for example, a ceramic or a quartz substrate or another suitable carrier material. If it is directly integrated in the measuring bridge 1, the calibration standard 5 can also be realised in the form of one or more thin-film resistors.

(27) FIG. 1B shows a further exemplary embodiment of an equivalent circuit diagram of a measuring bridge 1, wherein a bias voltage can be supplied to the device under test 7. This is achieved via a so-called bias unit 20. The bias unit 20 provides an adjustable direct current decoupled from the high-frequency at the second connector 8 to which the device under test 7 is connected. The directivity attainable with the measuring bridge 1 is generally impaired by the insertion of such a bias unit 20. To compensate the influences of the bias unit 20 on the directivity, a dummy-bias unit 21 may be added to the measuring bridge at the side of the calibration standard 5. In terms of circuit technology, the construction of this dummy-bias unit 21 is the same as the construction of the bias unit 20. However, the difference is that the dummy-bias unit 21 does not feed any direct current into the calibration standard 5. The measuring bridge 1 is accordingly once again converted into a symmetrical condition, because changes in the phase position caused by the bias unit 20 are also caused in the calibration standard 5 by the dummy-bias unit 21.

(28) A further difference between the exemplary embodiment from FIG. 1A and the exemplary embodiment from FIG. 1B is that the second resistor 2.sub.2 of the first matching pad 2 and the second resistor 3.sub.2 of the second matching pad 3 have been combined to form a first equivalent resistor 22.sub.1. For the 50 ohm system, this equivalent resistor provides a value of approximately 33.56 ohms. Furthermore, the third resistor 2.sub.3 of the first matching pad 2 and the third resistor 4.sub.3 of the third matching pad 4 have been combined to form a second equivalent resistor 22.sub.2. This second equivalent resistor 22.sub.2 also provides a value of approximately 33.56 ohms for the 50 ohm system. The remainder of the functioning of the measuring bridge 1 from FIG. 1B corresponds to the functioning of the measuring bridge 1 shown in FIG. 1A, so that with regard to the other components, reference is made to the deliberations relating to FIG. 1A.

(29) FIG. 1C shows a further exemplary embodiment of an equivalent circuit diagram of the measuring bridge 1, wherein a differential amplifier 11.sub.2 has been used for the common-mode-suppressing element 11.sub.1. However, this differential amplifier 11.sub.2 must provide a sufficiently high common-mode suppression. In this manner, the spatial arrangement of the measuring bridge 1 can be reduced. However, by contrast with the use of a balun 11.sub.1, the upper threshold frequency is significantly lower when using a differential amplifier 11.sub.2. Accordingly, differential amplifiers 11.sub.2 which provide a sufficiently high common-mode suppression, so far only up to an upper frequency range of approximately 5 GHz, are suitable. The remaining circuit structure from FIG. 1C corresponds to the circuit structure from FIG. 1B and FIG. 1A, for which reason reference is made in the following to the previous part of the description.

(30) FIG. 2A shows an exemplary embodiment for the realisation of the three matching pads 2, 3, 4 of the measuring bridge 1. In this context, the measuring bridge 1 is realised using thin-film technology based on microstrip lines. The first resistor 3.sub.1 and the third resistor 3.sub.3 of the second matching pad 3 and the first resistor 4.sub.1 and the second resistor 4.sub.2 of the third matching pad 4 and the first resistor 2.sub.1 of the first matching pad 2 are clearly recognisable.

(31) As already explained with reference to FIG. 1B, the second resistor 3.sub.2 of the second matching pad 3 and the second resistor 2.sub.2 of the first matching pad 2 may be combined to form a first equivalent resistor 22.sub.1. The same also applies for the third resistor 2.sub.3 of the first matching pad 3 and the third resistor 4.sub.3 of the third matching pad 4, which may be combined to form the second equivalent resistor 22.sub.2. These resistors are thin-film resistors which can be trimmed to their nominal value by means of a laser.

(32) Accordingly, the measuring bridge 1 may be embodied on a substrate 25, for example, on a ceramic 25 or on a quartz substrate 25 or another suitable carrier medium 25. The parasitic losses of a ceramic carrier are significantly lower at high frequencies than those of the known synthetic materials which are used in printed circuit boards.

(33) FIG. 2B shows an exemplary embodiment for the realisation of a calibration standard 5 which is embodied directly on the measuring bridge 1. It is clearly evident that a first region 26 enclosed by a dashed-line in FIG. 2B contains the excerpt from FIG. 2A. With reference to the first region 26 enclosed by a dashed-line, reference is made to the parts of the description relating to FIG. 2A. Especially at low frequencies, when the phase rotation of the supply lines and the influences of the transitions for the coupling of the device under test 7 are negligible, the calibration standard 5 can be balanced by a simple 50 ohm resistor. In FIG. 2B, this calibration standard 5 is balanced by two 100-ohm thin-film resistors 27.sub.1, 27.sub.2, which are each connected through a via 28.sub.1, 28.sub.2 to the reference ground on the rear side of the substrate 25, that is, on the rear side of the carrier material 25 and are accordingly connected in parallel. These two 100-ohm thin-film resistors 27.sub.1, 27.sub.2 can also be trimmed to their exact nominal value by means of a laser. Furthermore, a conductor track 29 is shown which connects the third resistor 3.sub.3 of the second matching pad 3 to the two 100-ohm thin-film resistors 27.sub.1, 27.sub.2. This conductor track 29 provides a V-shaped recess 30 at its end facing towards the calibration standard 5. Accordingly, the current flow is not guided to the end, thereby reducing the capacitive scattering field.

(34) FIG. 2C shows an exemplary embodiment of the measuring bridge 1 which is suitable for connection to a separate calibration standard 5. The region 26 from FIG. 2B which contains the three matching pads 2, 3, 4, as was described with reference to FIG. 2A, is clearly recognisable in the centre of FIG. 2C. Furthermore, the measuring bridge 1 from FIG. 2C provides a first connector 6, a second connector 8 and a third connector 10. The connectors 6, 8, 10 can be connector sockets. The separate calibration standard 5 can be screwed to the first connector 6, whereas the second connector 8 is connected to the device under test 7. As already explained, the first connector 10 can be connected to the signal generator 9. It is important that the first connector 6 provides the same electrical properties as the second connector 8. This means that, like the attenuation of the signal in the two connectors 6, 8, the phase rotations which are caused by the first connector 6 and the second connector 8 may be be approximately identical in magnitude. By preference, the third connector 10 also provides the same electrical properties. It is clearly recognisable that the measuring bridge 1 also provides a housing 31 in addition to the substrate 25. This housing 31 is preferably made from a conducting metal, such as aluminium, and prevents interference radiation from being coupled into the measuring bridge 1.

(35) FIG. 3 shows an exemplary embodiment of a chocked balun 11.sub.1. As already explained with reference to FIG. 1A, the first signal line 12.sub.1 is connected to the inner conductor 35 of the balun 11.sub.1, whereas the second signal line 12.sub.2 is connected to the outer conductor 36 of the balun 11.sub.1. The balun 11.sub.1 from FIG. 3 has a coaxial construction. In this context, the inner conductor 35 is radially surrounded by an outer conductor 36. In the exemplary embodiment from FIG. 3, two ferrites 37.sub.1 and 37.sub.2 are also present, which radially surround the outer conductor 36. The interfering sheath currents I.sub.M on the outer conductor 36 are suppressed with the assistance of the ferrites 37.sub.1, 37.sub.2. As already explained, this interfering sheath current I.sub.M which would lead to an unbalanced signal component at the output of the balun 11.sub.1 is suppressed by the ferrites 37.sub.1, 37.sub.2 on the sheath of the coaxial line, so that the same current I.sub.B flows in the two output lines 35, 36 in opposite directions.

(36) The ferrites 37.sub.1, 37.sub.2 are preferably embodied as ferrite rings which can have a different length and a different size and can be made from different ferrite materials. These ferrites 37.sub.1, 37.sub.2 can be arranged at different distances on the coaxial conductor. The sequence and arrangement of the ferrites 37.sub.1, 37.sub.2 is guided by the frequency range to be covered and must be matched from case to case. The matching can be implemented in such a manner that the first connector 6 is connected to the calibration standard 5 and also the second connector 8 is connected to a known calibration standard 5. Via the signal generator 9, a signal is fed to the measuring bridge 1 via the third connector 10. The ferrites are then displaced on the outer conductor 36 until the smallest possible voltage minimum is measured at the fourth connector 13. In the case of ideal components, this voltage would be equal to 0.

(37) However, with increasing frequencies, problems occur which are caused by resonance phenomena on the outer conductor 36 of the coaxial line 38. Furthermore, the behaviour of the ferrite materials at relatively high frequencies is not ideal and in part undefined. For this reason, a different balun concept is preferably used for the balancing of frequencies above, for example, 4 GHz.

(38) FIG. 4 shows an exemplary embodiment of a resistive balun 11.sub.1, which can be used for relatively high-frequencies. The first signal line 12.sub.1 and the second signal line 12.sub.2 are clearly recognisable. These two signal lines 12.sub.1, 12.sub.2 are balanced lines which are arranged on the upper side of the substrate 25. A ground surface is arranged on the rear side of the substrate 25, wherein a resistor layer 32, which is disposed beneath the first signal line 12.sub.1 and the second signal line 12.sub.2 is embedded within the ground surface. A housing 31 also ensures that this resistive balun 11.sub.1 is shielded.

(39) FIG. 5 shows an exemplary embodiment which describes in greater detail the field distribution for common-mode signals and odd-mode signals in the resistive balun 11.sub.1 from FIG. 4. The left-hand drawing from FIG. 5 describes the field distribution for a common-mode signal in a resistive balun 11.sub.1. It is clearly evident that the current flows in the same direction in both signal lines 12.sub.1, 12.sub.2. As already explained with reference to FIG. 4, the signal lines 12.sub.1, 12.sub.2 are separated by the substrate 25 from a resistor layer 32 disposed beneath them. The substrate 25 is preferably a ceramic. The field lines of the electrical field are clearly evident. As already explained, in the case of the common-mode signal, the current flows in the same direction in both signal lines 12.sub.1, 12.sub.2. The current circuit is closed by the current in the ground surface on the rear side of the substrate. This current can be attenuated with the assistance of the resistor layer 32 on the rear side of the substrate.

(40) By contrast, a different situation arises with an odd-mode signal, as shown in the right-hand drawing of FIG. 5. In this case, the currents in the first signal line 12.sub.1 and in the second signal line 12.sub.2 flow in a different direction. Such an odd-mode signal is only slightly influenced by the resistor layer, as shown by the field distribution of the electrical field.

(41) FIG. 6 shows an exemplary embodiment of a resistive balun 11.sub.1 with a DGS structure 39 (Defective Ground Structure). The parts of the resistor layer 32 which are disposed directly beneath the first signal line 12.sub.1 and the second signal line 12.sub.2 have been removed so that the resistor layer 32 influences the odd-mode signal even less. The DGS structure 39 splits the resistor layer 32, which is a thin-film resistor layer 32, into two mutually separate thin-film resistor layers 32, wherein the width of the DGS structure 39 is matched to the substrate material used and the frequency range to be covered, and wherein it is embodied directly between the two signal lines 12.sub.2, 12.sub.2 on the lower side of the substrate 25. In the example from FIG. 6, the width of the DGS structure 39 is significantly greater than the distance between the first signal line 12.sub.1 and the second signal line 12.sub.2.

(42) The attenuation of the odd-mode signal can also be further reduced in that the slot width between the first signal line 12.sub.2 and the second signal line 12.sub.2 is reduced, because this concentrates the electromagnetic field in the region of the slot, and the field components in the region of the resistor layer 32 are reduced, so that the first signal line 12.sub.1 and the second signal line 12.sub.2 are disposed close together. In this context, the slot width can be selected in such a manner that it encloses a region of, for example, a few μm up to, for example, a few mm.

(43) The use of such a DGS structure 39 means that, with an odd-mode signal, the field concentration on the resistor layer 32 is reduced, whereas, with a common-mode signal, the electromagnetic fields are concentrated at the edges of the resistor layer 32, so that the common-mode signal is attenuated significantly more strongly than the odd-mode signal.

(44) However, for the balancing of frequencies below 4 GHz, a very long resistor layer 32, which would lead to unacceptably large losses of the odd-mode signal at high frequencies, would be required.

(45) For this reason, the balun 11.sub.1 preferably combines the concepts from FIG. 3, FIG. 4 and FIG. 6. FIG. 7 thus shows an exemplary embodiment of a balun 11.sub.1, which comprises ferrites 37.sub.1, 37.sub.2, 37.sub.3, 37.sub.4, 37.sub.5, and also a resistive part. The balun 11.sub.1 comprises a first portion 50 and a second portion 51. The first portion 50 shows the construction of the resistive part as was explained with reference to FIGS. 4, 5 and 6. The second portion 51 shows the construction as explained with reference to FIG. 3. It is clearly evident that the first portion 50 comprises a substrate 25 on the upper side of which the first signal line 12.sub.1 and at least the one second signal line 12.sub.2 are guided. On the lower side of the substrate 25 in the exemplary embodiment from FIG. 7, a thin-film resistor layer 32 is embodied beneath the signal lines 12.sub.1, 12.sub.2.

(46) As already explained, this thin-film resistor layer 32 attenuates the common-mode signal on the first and the second signal line 12.sub.1, 12.sub.2 significantly more strongly than the odd-mode signal on the two lines. The first portion 50 of the balun 11.sub.1 ensures that the common-mode signal is effectively suppressed for frequencies above approximately 4 GHz. As already explained, this thin-film resistor layer 32 would have to be very long in order to attenuate a low-frequency common-mode signal equally well. For this reason, the second portion 51 of the balun 11.sub.1 comprises a balun 11.sub.1 constructed in a coaxial manner, wherein the coaxial conductor 38 is surrounded by several ferrites 37.sub.1 to 37.sub.5. The inner conductor 35 of the coaxial line 38 of the second portion 51 of the balun 11.sub.1 is connected to the first signal line 12.sub.1, wherein the outer conductor 36 is connected at a first end of the coaxial line 38 to the second signal line 12.sub.2.

(47) Furthermore, the second portion 51 of the balun 11.sub.1 optionally provides a compensation line 52 which is preferably a wire. This compensation line 52 is also connected at its first end to the first signal line 12.sub.1 and accordingly to the inner conductor 35 of the coaxial line 38. Because of the ferrites 37.sub.1 to 37.sub.5, the inductance of the outer conductor 36 of the coaxial line 38 is increased, whereas, in view of the coaxial construction, the inductance of the inner conductor 35 is not increased. This factual situation would lead to an unbalanced construction of the balun 11.sub.1 dependent upon the respective operating frequency. For this reason, the inner conductor 35 is connected to a separate compensation line 52, which is also preferably guided through the same ferrites 37.sub.1 to 37.sub.5 as the coaxial line 38. In this case, the inductance of the inner conductor 35 is increased relative to the housing ground to the same extent as the inductance of the outer conductor 36 is increased relative to the housing ground. The balun 11.sub.1 provides a symmetrical behaviour independently of the operating frequency.

(48) FIG. 8 shows an exemplary embodiment of a combination of a balun 11.sub.1 with ferrites 37.sub.1 to 37.sub.5 and a resistive balun 11.sub.1. By contrast with FIG. 7, which shows a plan view of an exemplary embodiment of the balun 11.sub.1, FIG. 8 visualises a three-dimensional view of an exemplary embodiment of the balun 11.sub.1. The first portion 50 and the second portion 51 are also clearly evident. The first portion 50 forms the resistive part of the balun 11.sub.1, whereas the second portion 51 shows the so-called “choked balun”. FIG. 8 also shows the compensation line 52, which is guided through the same ferrites 37.sub.1 to 37.sub.5 as the coaxial line 38, so that the balun 11.sub.1 is constructed in a symmetrical manner. A housing 31 shields the balun 11.sub.1 from the environment.

(49) FIG. 9 shows a further exemplary embodiment of a combination of a balun 11.sub.1 with ferrites 37.sub.1 to 37.sub.5 and a resistive balun 11.sub.1 with a DGS structure 39. The DGS structure 39 which ensures that the common-mode component is attenuated significantly more strongly than the odd-mode component is clearly evident. The remaining construction from FIG. 9 corresponds to that from FIG. 7, to which reference is made in this context.

(50) FIG. 10 shows an exemplary embodiment of a combination of a balun 11.sub.1 with ferrites 37.sub.1 to 37.sub.5 and a compensation line 52 and a resistive balun 11.sub.1 with DGS structure 39. FIG. 10 shows a three-dimensional view of the balun 11.sub.1, whereas FIG. 9 shows a plan view of the same. The housing 31 which shields the balun 11.sub.1 is also clearly recognisable.

(51) FIG. 11 shows an exemplary embodiment of a combination of a balun 11.sub.1 with ferrites 37.sub.1 to 37.sub.5 and a compensation line 52 together with cup cores 60.sub.1, 60.sub.2 and a resistive balun 11.sub.1 with DGS structure 39. The compensation line 52 is insulated from the coaxial line 38. This can be implemented in that either the compensation line 52 and/or the coaxial line 38 is provided with a shrinkable sleeve.

(52) Two cup cores 60.sub.1, 60.sub.2, which serve to expand the useful frequency range up to frequencies in the kHz range are clearly evident. In this context, the end of the coaxial line 38 which is distant from the first portion 50 of the balun 11.sub.1 is connected to a first cup core 60.sub.1, wherein the first cup core 60.sub.1 comprises a coaxial line 38 coiled within a ferrite.

(53) Furthermore, the end of the compensation line 52, which is distant from the first portion 50 of the balun 11.sub.1 is connected to a second cup core 60.sub.2, wherein the second cup core 60.sub.2 comprises a line 52 coiled within a ferrite, wherein this second cup core 60.sub.2 provides approximately the same electrical properties as the first cup core 60.sub.1. By preference, the cup cores 60.sub.1, 60.sub.2 are not connected to the coaxial line 38 and the compensation line 52, but the coaxial line 38 is coiled in the first cup core 60.sub.1, whereas the compensation line 52 is coiled in the second cup core 60.sub.2.

(54) The compensation line 52, which preferably provides an electrical insulation, is connected to the reference ground at the output of the second cup core 60.sub.2. The outer conductor 38 is also connected to the reference ground at the output of the first cup core 60.sub.1. An electronic measuring device which is not illustrated determines magnitude and phase of the voltage of the inner conductor 35 relative to the reference ground.

(55) FIG. 12 shows an exemplary embodiment of a separate guiding the coaxial line 38 and a compensation line 52 through different ferrites 70.sub.1 to 70.sub.10 or respectively 71.sub.1 to 71.sub.10. In this case, the coaxial line 38 is guided through the ferrites 70.sub.1 to 70.sub.10, whereas the compensation line 52 is guided through the ferrites 71.sub.1 to 71.sub.10. The level by which the ferrites 70.sub.1 to 70.sub.10 increase the inductance of the outer conductor 38 should be exactly as large in this context as the level by which the ferrites 71.sub.1 to 71.sub.10 increase the inductance of the compensation line 52.

(56) All the features described and/or illustrated herein can be combined arbitrarily with one another. The embodiments are not restricted to the exemplary embodiments described. For example, more than 10 ferrites or less than 10 ferrites can, of course, also be used.