ADAPTATION CABLE AND ASSOCIATED STAGE AND AMPLIFIER AND / OR DIVISION DEVICE

Description

[0001] The present invention relates to a radio frequency amplification or division apparatus comprising a cavity combiner, a plurality of first channels and a second channel, the combiner being configured to recombine the first channels into the second channel, each first channel comprising an amplifier and an adaptation line.

[0002] Large scientific instruments such as particle colliders need to be supplied with radio frequency waves for various applications.

[0003] For example, in particle accelerators, acceleration is achieved using cavities powered by high power at frequencies in the range between 10 MHz (megahertz) and 10 GHz (gigahertz). In this context, "high power" is understood to mean power between 100 kilowatts (kW) and several megawatts (MW).

[0004] Thus, for certain applications, it is desirable to have radio frequency waves with high power.

[0005] High power output is achieved using amplifiers. An electronic amplifier is an electronic system that increases the power of an electrical signal. There are two main types of electronic amplifiers: tube amplifiers and solid-state amplifiers.

[0006] US documents 2015 / 123734 A1 and US 2012 / 0293274 A1 present a signal amplification device.

[0007] However, since the unit power of high-frequency amplifiers is limited, it is often useful to couple (or combine) several elementary amplifiers to achieve the desired power.

[0008] Cavity combiners are a well-known method for achieving this. Cavity combiners are a specific example in the field of radio frequency wave combination. They allow, in particular, the summing of a large number of inputs in a single combining stage, thus transforming low-power technology (range between 1 watt and 1 kilowatt) into high-power technology (range between 100 watts and 500 kilowatts). This type of combiner is reciprocal, meaning it can also be used as a power divider.

[0009] Cavity combiners offer low losses, a small footprint, and high power.

[0010] A cavity combiner amplifier device usually includes channels with amplifiers with high-power transistors mounted in push-pull configuration.

[0011] However, this configuration has the disadvantage of requiring a power balun. In the most compact arrangements, the energy dissipation in the balun is generally significant, leading to considerable complexity and additional cost.

[0012] Furthermore, the cavity combiner amplifier requires the use of a large number of parallel inputs.

[0013] However, this requires ensuring reliable operation regardless of the state of each input. For example, in the event of an amplifier failure, the device's output power must be maintained as much as possible.

[0014] In fact, the device's efficiency decreases when an amplifier fails.

[0015] Therefore, there is a need for a cavity combiner amplification device to remedy at least partially the aforementioned drawbacks.

[0016] For this purpose, an amplification device conforming to claim 1 is described.

[0017] According to particular embodiments, the amplification device comprises one or more of the features of claims 2 to 6, taken individually or in all technically possible combinations.

[0018] Other features and advantages of the invention will become apparent from the following description of embodiments of the invention, given by way of example only and with reference to the drawings which are: figure 1 a schematic view of an example of a radio frequency amplification device, figure 2 , a schematic view of an example of a first stage forming part of the apparatus of the figure 1 , figure 3 , a view of an equivalent electrical diagram of the first floor according to the figure 2 , figure 4 , a schematic view of an example amplifier that is part of the device of the figure 1 , figure 5 , a schematic view of an example of a first-way section and a cavity combiner forming part of the apparatus of the figure 1 , figure 6 , a graph showing the temporal evolution of the output power of the amplification device in accordance with the figure 5 , figure 7 , a schematic view of another example of a first-way section and a cavity combiner forming part of the apparatus of the figure 1 , figure 8 , a view of an equivalent electrical diagram of the first-track section and the cavity combiner according to the figure 7 , figure 9 , a perspective mechanical diagram of a differential probe in the form of an insertion loop of the cavity combiner according to the figure 7 , figure 10 , a mechanical diagram in perspective of another example of an insertion loop, figure 11 , a mechanical diagram in perspective of yet another example of an insertion loop, figure 12 , a schematic view of an electrical diagram of another example of a radio frequency amplification device, figure 13 , a schematic view of yet another example of a radio frequency amplification device, and figure 14 , a schematic view of an example of a cavity combiner.

[0019] A radio frequency amplification device 10 is illustrated in the figure 1 .

[0020] The device 10 includes first channels 12 which are recombined by a cavity combiner 14 into a second channel 16.

[0021] Device 10 is designed to operate at an operating frequency which is the operating frequency of the combiner 14.

[0022] More specifically, the device 10 is suitable for operation within a set of operating frequency bands, the frequency band being determined by the frequency band of the combiner 14.

[0023] The operating frequency is an operating frequency between 10 kHz and 300 GHz.

[0024] This means that all components of device 10 are sized to operate specifically at the operating frequency.

[0025] In particular, the impedances of the components are calculated at the operating frequency.

[0026] Furthermore, depending on the power level envisaged for the radio frequency wave, the components are more adapted to support a significant power.

[0027] Following the example of the figure 1 , three first 12 paths are represented, the dotted lines indicating that the number of first 12 paths is arbitrary.

[0028] The number of first lanes 12 is greater than two.

[0029] For example, the number of first 12 lanes is greater than 50 or even greater than 100.

[0030] In the example of the figure 1 , each first lane 12 is identical.

[0031] Only one first path 12 is specifically described in the following, the same remarks applying to the other first paths 12

[0032] The first channel 12 includes a source 18, a first stage 20, an amplifier 22 and a second stage 24.

[0033] Source 18 is suitable for supplying the first channel 12 with radio frequency waves.

[0034] For example, source 18 is a current source.

[0035] The first stage 20 is a low-power stage.

[0036] The term "low power" indicates that the first stage 20 is adapted to interact with radio frequency waves having less energy than the waves with which the second stage 24 is suitable to interact and with which the combiner 14 is suitable to interact.

[0037] In this case, a wave is considered to have a low power when the power of the wave is between 0.1 W and 100 W.

[0038] An example of a first floor 20 is illustrated in figures 2 et 3 .

[0039] The first stage 20 includes a balun 26, a decoupling module 28 and an impedance adapter 30.

[0040] A balun is an electrical circuit used to perform the link between a symmetrical transmission line (two-wire line or parallel printed lines) and an asymmetrical transmission line (coaxial cable or printed line above a ground plane).

[0041] A balun is, for example, made using coiled coaxial cable or a small section of twin-wire line wound on a ferrite toroid.

[0042] Such an embodiment is schematically illustrated by the figure 3 .

[0043] The decoupling module 28 is suitable for performing DC type decoupling (acronym for "direct current" in English which means "direct current").

[0044] According to the illustrated example, the decoupling module 28 includes capacitors 32 and 34.

[0045] The impedance adapter 30 ensures an impedance matching between the impedance of the balun 26 and the decoupling module 28 with respect to the impedance of the amplifier 22.

[0046] For example, as illustrated, the 30 impedance adapter is a coaxial cable.

[0047] The equivalent electrical circuit is a line with discrete elements, that is to say a dipole comprising inductances 36, 38, 40, 42 and a capacitor 44. The inductances 36, 38, 40, 42 are arranged in the form of a Wheatstone bridge connected by a capacitor 44.

[0048] Amplifier 22 is suitable for amplifying an incident wave into an emergent wave with a greater energy than the incident wave.

[0049] Amplifier 22 is, for example, a circuit in a so-called "push-pull" configuration. Such a circuit uses a pair of active components that alternately supply or absorb current to or from a load.

[0050] In this case, as illustrated in the figure 4 The amplifier 22 is an assembly of two transistors 46 and 48 in a so-called "push-pull" configuration. This configuration allows for compact and high-power chips.

[0051] According to the illustrated embodiment, each transistor 46 and 48 is a hardened transistor. The term "hardened" in this context means that each transistor 46 and 48 is protected from reflected waves. As a result, each transistor 46 and 48 exhibits a certain robustness to withstand short circuits for several milliseconds.

[0052] The second stage 24 is a medium power stage.

[0053] The term "medium power" indicates that the second stage 24 is adapted to interact with radio frequency waves with higher energy than the waves with which the first stage 20 but with lower energy than the waves with which the combiner 14 is designed to interact.

[0054] In this case, a wave is considered to have medium power when the power of the wave is between 100 W and 1 kW.

[0055] Each component of the second stage 24 is adapted to interact with a wave exhibiting medium power when the wave power is between 100 W and 1 kW.

[0056] The second stage 24 is designed to provide the interface between the amplifier 22 and the cavity 14.

[0057] According to the specific example of the figure 5 , the second stage 24 includes an adaptation line 50 connecting the amplifier 22 to the combiner 14.

[0058] The adaptation line 50 has a first end connected to the amplifier 22 and a second end connected to the combiner 14.

[0059] The first end is impedance matched to the impedance of amplifier 22.

[0060] The second end is impedance matched to the impedance of the combiner 14.

[0061] The impedance of the matching line 50 is the difference in matching between the impedance of the first end and the impedance of the second end, taking into account skin losses.

[0062] The impedance of the matching line 50 is greater than the ratio between the wavelength associated with the natural frequency and the number 8.

[0063] According to one embodiment, the impedance of the matching line is equal to the ratio between the wavelength associated with the natural frequency and the figure to within 8 to 10%.

[0064] Alternatively, the matching line 50 has an impedance equal to the ratio between the wavelength associated with the natural frequency and the figure 8 to within 5%.

[0065] According to one embodiment, the matching line 50 has an impedance equal to the ratio between the wavelength associated with the operating frequency and the number 8.

[0066] For example, according to the illustrated embodiment, the matching line 50 is a coax having a length equal to the ratio between the wavelength associated with the operating frequency and the number 8.

[0067] According to another embodiment, the adaptation line 50 is a conductor line also called by the English name "stripline".

[0068] According to another embodiment, the matching line 50 is a waveguide.

[0069] According to yet another embodiment, the adaptation line 50 includes components such as inductors and capacitors.

[0070] The second floor 24 is without a circulator and isolator.

[0071] The combiner 14 has a plurality of first ports P1, P2 ... PN and a second port O1.

[0072] Combiner 14 is designed to sum the medium power radio frequency waves introduced into each of the first ports P1, P2 ... PN to generate a high power wave (between 10 kW and 200 kW) on the second port O1.

[0073] Depending on the application, the combiner 14 is adapted to operate with resonant modes of a cavity. The combiner 14 operates by performing voltage or current summation.

[0074] When the combiner 14 operates in TM010 mode, the magnetic field is at its maximum at the walls of the combiner 14 and the electric field is at its maximum at the center of the combiner 14. Power is injected through insertion loops on the wall and taken up by a voltage probe at the center.

[0075] For the following, by way of illustration, it is assumed that combiner 14 is a combiner operating with a resonance in TM010 mode.

[0076] This means that the combiner 14 is suitable for interacting with a medium-power wave when the wave power is between 10 kW and 200 kW. This suitability applies to all components of the combiner 14.

[0077] The operation of device 10 in the event of a malfunction is now described with reference to the figure 6 .

[0078] The results for a combiner 14 with an operating frequency of 200 MHz, 80 first ports, and a 50 adaptation line are visible in the simulation of the figure 6 .

[0079] This confirms the value of the 50 adaptation line, which allows us to largely limit the power loss and therefore the mismatch of the cavity and all the amplifiers 22.

[0080] At an initial instant t0 corresponding to the start of the simulation, each of the first 80 ports transmits their power to the second port O1.

[0081] It is observed at a first instant t1 (at 300 µs) that the power at the combined port amounts to Pout = 72.7 kW for 80*0.93=74.4 kW at the input, which corresponds to a loss of only 0.1 dB.

[0082] At a second instant t2, 500 ns after the initial instant t0, the first 16 ports are deactivated by opening the circuit.

[0083] It is noted that the power at the second port O1 amounts to Pout = 55.5 kW for 59.7 kW input.

[0084] At a third instant t3, 1 µs after the initial instant t0, the impedance of transistors 46 and 48 is brought back to the first 16 ports at the same time.

[0085] It is observed that the power at the second port O1 decreases slightly (to 51.47 kW) because the 16 undamaged amplifiers 22 absorb radio frequency power (approximately 225 W per transistor 46 and 48). This corresponds to a loss of 0.63 dB (see fourth instant t4 at approximately 1.5 µs).

[0086] At a fifth instant t5, 2 µs after the initial instant t0, the first 16 ports are short-circuited at the same time.

[0087] It is observed that the power at the second port O1 returns to a nominal level of 55.6 kW for an input of 64*0.93=59.7 kW (see sixth instant t6 at approximately 2.5 µs). This corresponds to a loss of 0.31 dB.

[0088] It thus appears that the adaptation line 50 allows the combiner 14 to see an almost identical impedance on all the first ports P1 to PN.

[0089] The correct sizing of the adaptation line 50 allows the combiner 14 to be isolated from each amplifier 22.

[0090] The matching line 50 is sized so that the combiner 14 always sees approximately the same impedance regardless of the state of each amplifier 22.

[0091] Put another way, when an amplifier 22 fails, the operation of the combiner 14 is not affected.

[0092] Such a configuration makes it possible to maintain the efficiency of the combiner 14 almost constant without taking into account the number of active amplifiers 22.

[0093] By definition, the efficiency of the combiner 14 is defined by the ratio between the power on the second port O1 and the sum of the powers on each of the first ports P1 to PN.

[0094] This allows the replacement of a state-of-the-art isolator (i.e., a circulator and matching resistor assembly) with the 50 matching line.

[0095] This results in a reduction of costs.

[0096] Furthermore, the yield is optimized.

[0097] There figure 7 illustrates another example of a second floor 24. The elements are identical to the embodiment of the figure 5 are not described again. Only the differences are highlighted.

[0098] The impedance line 50 is replaced by a switching circuit 52.

[0099] An example of a switching circuit 52 is detailed in the figure 8 .

[0100] The switching circuit 52 includes a first switch 54 and a second switch 56.

[0101] The first switch 54 exhibits an initial switching time.

[0102] The first switch 54 is a fast switch, meaning that the first switching time is between 10 nanoseconds (ns) and 1 millisecond (ms).

[0103] The first switch 54 is designed to control the second switch 52 by emitting an output signal.

[0104] The first switch 54 is an electronic circuit containing PIN diodes. A PIN diode (English acronym for "Positive Intrinsic Negative diode") is a diode consisting of an undoped region, called intrinsic I, interposed between two doped regions P and N.

[0105] More specifically, the first switch 54 includes transistors and PIN diodes.

[0106] In the particular case shown, the first switch 54 has two assemblies connected together and identical.

[0107] The first setup allows switching in the presence of a positive half-cycle, while the second setup allows switching in the presence of a negative half-cycle.

[0108] The first circuit is a transistor 60, one terminal of which is connected to ground, another terminal is connected to the cathode of a diode 62 and to a set of an inductor 64 and a resistor 66, and another terminal is connected to a diode 64 and to the resistor 66.

[0109] According to the example, transistor 60 is an insulated gate field effect transistor more commonly known as a MOSFET (English acronym for "Metal Oxide Semiconductor Field Effect Transistor" - which translates to "metal-oxide-semiconductor field effect transistor").

[0110] Similarly, the second circuit is a transistor 70, one terminal of which is connected to ground, another terminal is connected to the anode of a diode 72 and to a set of an inductor 74 and a resistor 76, and another terminal is connected to a diode 78 and the resistor 76.

[0111] The second switch 56 is an electromechanical relay.

[0112] The electromechanical relay is designed to operate in two positions, a closed position in which the amplifier 22 is connected to the cavity 14 and an open position in which the amplifier 22 is not connected to the cavity 14.

[0113] The switching time of the second switch 56 is less than half the switching time of the first switch. The switching time of the second switch 56 is called the second switching time.

[0114] The second switching time is greater than 1 ms.

[0115] The second switch 56 is a safety switch.

[0116] The operation of device 10 according to the figure 7 is now described.

[0117] The second switch 56 is normally closed.

[0118] More specifically, the second switch 56 is in a short-circuited state when the second switch 56 is powered by a zero output signal.

[0119] As soon as the output signal is no longer zero, the second switch 56 is opened to allow the operation of the device 10.

[0120] The switching circuit 52 allows the safety and de-energization of an amplifier 22 coupled to a combiner 14 by creating a short circuit.

[0121] The short circuit is achieved by switching to ground.

[0122] In the case of a power injection in a magnetic field into a combiner 14, if a loop is grounded, the loop impedance becomes highly reactive.

[0123] Therefore, the loop consumes no power. Furthermore, the loop does not alter the cavity's operation, as the cavity impedance remains unchanged. Only the short-circuit resistance of a few milliohms introduces losses in the combiner 14.

[0124] In the case of electric field injection, the switching circuit opens the loop, resulting in a high impedance. It should also be noted that, for a cavity with voltage injection, a circuit opening occurs.

[0125] As a result, the switching circuit 52 has a transition time shorter than the reaction time of the combiner 14.

[0126] Therefore, the switching circuit 52 replaces a circulator.

[0127] The presence of a circulator is therefore eliminated.

[0128] In addition, the switching circuit allows the amplifier to be switched off without cutting off the entire device 10.

[0129] On the one hand, this prevents further damage to the faulty amplifier 22. Indeed, with a short circuit, no radio frequency power is fed back to a transistor 46, 48 of the amplifier 22.

[0130] On the other hand, this facilitates maintenance since it allows the faulty amplifier 22 to be disconnected and replaced without interrupting the operation of device 10.

[0131] This results in reduced costs and optimized performance.

[0132] There figure 9 presents a perspective mechanical diagram of a differential probe in the form of an insertion loop 51 of the combiner 14 according to the figure 7 .

[0133] In some terminology, the insertion loop is also called a coupler.

[0134] In this representation, the insertion loop 51 has a connecting rod 80 connected to two conduction rods 82 and 84. In this case, the insertion loop 51 also has a voltage input 86. The first conduction rod 82 is connected to the voltage input 86 while the second conduction rod 84 is connected to ground.

[0135] The equivalent electrical circuit of the insertion loop 51 is an inductance connected on one side to the voltage input and on the other side to ground.

[0136] It is proposed, with reference to figures 10 et 11 , an insertion loop 51 with differential inputs.

[0137] Differential inputs are defined as inputs that are 180° out of phase with another input. This phase shift is implemented before the insertion loop 51

[0138] Such an insertion loop 51 comprises at least two conduction rods, each connected to a separate input.

[0139] Depending on the method of implementation of the figure 10 The insertion loop 51 contains the same elements as the insertion loop 51 of the figure 9 In this particular case, for convenience in naming, all elements are referred to as primary elements in contrast to the elements that are added, which are referred to as secondary elements.

[0140] The insertion loop 51 further includes a secondary connecting rod 88, connected to two secondary conduction rods 90 and 92. In this case, the insertion loop 51 further includes a secondary voltage input 86 separate from the primary voltage input 86. The first secondary conduction rod 82 is connected to ground while the second secondary conduction rod 84 is connected to the second secondary voltage input 86.

[0141] Depending on the method of implementation of the figure 10 The insertion loop 51 contains the same elements as the insertion loop 51 of the figure 9 .

[0142] The insertion loop 51 further includes a second voltage input 96 distinct from the first voltage input 86.

[0143] The second conduction rod 84 is connected to the second voltage input 96.

[0144] The operation of the insertion loop 51 is illustrated with reference to the figure 12 which illustrates a device 10 according to the embodiment of the figure 5 in which an insertion loop 51 conforms to the embodiment of the figure 11 is used.

[0145] In this case, the output voltage of each channel 12 is accessible to the insertion loop 51 by difference between the two inputs 86 and 96, so that the operation of the device 10 is comparable to the operation of a device 10 according to the prior art.

[0146] Similarly, in the case of the insertion loop 51 of the figure 10 , the output voltage of each channel 12 is accessible to the insertion loop 51 by the difference between the two inputs 86 and 94.

[0147] The insertion loop 51 does indeed act as an input coupler for the combiner 14.

[0148] Such an insertion loop 51 eliminates the need for a balun in the second stage 24.

[0149] The implementation method of insertion loop 51 according to the figure 11 also eliminates the need for a decoupling module in the second stage 24.

[0150] In all cases, the balun function is located inside cavity 14.

[0151] The result is that the first channel 12 can be made on a printed circuit board with a standard substrate instead of a substrate capable of withstanding high-power radio frequency waves.

[0152] Furthermore, such an insertion loop 51 allows the number of insertion loops 51 in the cavity 14 to be doubled. This allows the quality coefficient to be adjusted and therefore the device 10 to be reduced in height time.

[0153] Therefore, the reliability and performance of device 10 are increased.

[0154] In the embodiments presented, the first stage 20, the amplifier 22 and the second stage 22 are part of the same board usually called the "amplifier board".

[0155] Such a card is suitable for being plugged into a 14-port combiner.

[0156] The methods of implementation of figures 13 And 14 illustrate the case where the amplifier board includes only the first stage 20, the amplifier 22 and not the second stage 22.

[0157] In these embodiments, the second stage 22 is integrated into each first port P1 to PN.

[0158] The amplifier cards plug directly into each port and do not have a second stage 22.

[0159] This makes the manufacture of amplifier boards easier.

[0160] Depending on the specific case of the figure 14 It should be noted that each first port P1 to PN has an insertion loop 51 conforming to the figure 11 , an impedance line 50 and a switching circuit 52.

[0161] Such first P1 to PN ports therefore combine the advantages of each of the aforementioned components.

[0162] All the embodiments presented make it possible to achieve high-power amplification capable of competing with other amplification technologies.

[0163] In fact, device 10 allows for a greater limited volumetric power and limits the cost.

[0164] Furthermore, it is proposed to combine as many functions as possible in the same equipment, which improves the compactness of the device 10.

[0165] Maintenance of device 10 is also facilitated.

[0166] In addition, device 10 offers improved reliability.

[0167] In addition, the performance of device 10, namely efficiency, bandwidth, linearity, and rise time, is better than in the prior art.

[0168] As a result, device 10 is better suited for continuous and uninterrupted operation.

[0169] In general, the proposed implementation methods make it possible to obtain high-power amplification devices at lower costs.

[0170] The frequency range envisaged for device 10 covers the range from a few tens of megahertz up to the current technological limits, i.e. more than a hundred gigahertz.

[0171] The device 10 is usable for numerous applications. In particular, the device 10 is advantageously used in large scientific instruments, in an accelerating cavity, and in the field of telecommunications.

[0172] Furthermore, device 10 is also reversible, as it can be used as a radio frequency divider. In this case, the second port O serves as the input and the first ports P1 to PN serve as the output. Additionally, the differential probe 51 interacts with voltage outputs.

[0173] It should be noted that the term "differential probe" refers to a probe comprising an insertion loop 51 and at least two inputs / outputs 86, 96. In particular, the term "differential probe" refers to a probe in which two inputs / outputs are 180° out of phase with each other.

[0174] Depending on the method of implementation of the figure 10 The differential probe 51 comprises two pairs of conduction rods 82, 84, 90, 92. One conduction rod 82, 92 of each pair is connected to a corresponding voltage input 86, 96.

[0175] Depending on the method of implementation of the figure 11, the differential probe 51 comprises a single pair of conduction rods 82, 84, each conduction rod 82, 84 being connected to a corresponding voltage input 86, 96.

[0176] The phase shift is performed before the insertion loop 51, for example at a voltage input 86, 96. It should be noted that the 180° phase shift is not necessarily performed by a single voltage input 86, 96. For example, each voltage input 86, 96 performs a phase shift, the sum of the two phase shifts being equal to 180°.

Claims

1. Radio frequency amplification or division apparatus (10) comprising a cavity combiner (14), a plurality of first channels (12) and a second channel (16), the combiner being configured to combine the first channels (12) into the second channel, each first channel (12) comprising an amplifier and a match line (50) for a radio frequency cavity combiner, the match line (50) being adapted for an operating frequency of between 10 kHz and 300 GHz, the match line (50) connecting the amplifier to the cavity combiner, the match line having an impedance equal to or greater than the impedance of a coax with a length equal to the ratio between the wavelength associated with the operating frequency and the number 8, each first channel (12) comprising a medium-power stage (24) for a radio frequency cavity combiner, the medium-power stage connecting the amplifier to the cavity combiner, the stage (24) comprising the match line (50), the cavity combiner comprises a plurality of first ports (P1, P2, PN) and a second port (O1), characterised in that the medium-power stage (24) also comprises a switching circuit (52), the switching circuit (52) being capable of producing a short-circuit by ground switching, the stage (24) having no circulator, each first port (P1, P2, PN) comprises an insertion loop (51).

2. Radio frequency amplification or division apparatus (10) according to claim 1, in which the match line (50) has an impedance equal to the impedance of a coax with a length equal to the ratio between the wavelength associated with the operating frequency and the number 8, to within 10%.

3. Radio frequency amplification or division apparatus (10) according to claim 1 or 2, in which the match line (50) has an impedance equal to the impedance of a coax with a length equal to the ratio between the wavelength associated with the operating frequency and the number 8.

4. Radio frequency amplification or division apparatus (10) according to any one of claims 1 to 3, wherein the match line (50) is a coaxial cable.

5. Radio frequency amplification or division apparatus (10) according to any one of claims 1 to 4, in which the stage (24) has no isolator.

6. Radio frequency amplification or division apparatus (10) according to any one of claims 1 to 5, in which the stage (24) is adapted for radio frequency waves with a power of between 100 W and 1 kW.