JP3999944B2 - Heterojunction bipolar transferred electron transistor, and voltage controlled frequency variable oscillator, injection locking oscillator, self-oscillating mixer, and receiver using the same - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to a heterojunction bipolar transferred electron transistor comprising III-V compound semiconductor elements, and a voltage-controlled variable frequency oscillator, injection-locked oscillator, self-oscillating mixer, and receiver using the same.
[0002]
[Prior art]
  The heterojunction bipolar transferred electron transistor operates as a high frequency oscillator or a self-oscillation mixer as described in the following prior art.
Prior literature
“Demonstration of a 77-GHz Heterojunction Bipolar Transferred Electron Device”, JKTwynam, M. Yagura, N. Takahashi, E. Suematsu and H. Sato, IEEE Electron Device Lett. 21 (01), pp. 2-4, 2000 .
[0003]
  Conventional heterojunction bipolar transferred electrons have a single cathode region adjacent to the base region. Therefore, in order to obtain a high oscillation output power, a high operating current is required and a wide cathode region is required. However, when the area of the cathode region is increased, there is a problem that the performance of the device is deteriorated. FIG. 26 shows an example of a conventional heterojunction bipolar transferred electron transistor having a rectangular cathode layer 3 having a length L and a width W. In FIG. 26, the metal base ohmic contact electrode 5 is formed around the cathode layer 3 so as to provide a low base resistance important for device operation.
[0004]
  FIG. 27 shows a heterojunction bipolar transferred electron transistor in which the cathode area is increased by increasing the width W of the rectangular cathode layer 3 in FIG. 26 to W ′. FIG. 28 shows a heterojunction bipolar transfer electron transistor in which the cathode area is increased by increasing the length L of the rectangular cathode layer 3 in FIG. 26 to L ′. 26 to 28, 1 is an n-type semiconductor anode layer, 2 is a p-type semiconductor base layer, 3 is an n-type semiconductor cathode layer, 4 is a metal anode ohmic contact electrode, 5 is a metal base ohmic contact electrode, 6 Is a metal cathode ohmic contact electrode, 7 is a metal second anode electrode, 8 is a metal second base electrode, and 9 is a metal second cathode electrode. In FIG. 26 to FIG. 28, the active layer is hidden under the semiconductor base layer 2.
[0005]
[Problems to be solved by the invention]
  However, the conventional heterojunction bipolar transfer electron transistor having a single cathode region has the following problems. That is, first, in the case of the heterojunction bipolar transfer electron transistor shown in FIG. 27, since the width W of the n-type semiconductor cathode layer 3 is increased, the n-type semiconductor cathode layer 3 is provided so as to provide a low base resistance. The area of the metal-based ohmic contact electrode 5 formed by surrounding the substrate is also increased. Therefore, there is a problem in that the base resistance important for the efficient operation of the device increases as the thermal resistance increases. Further, in the case of the heterojunction bipolar transferred electron transistor shown in FIG. 28, since the electrode length L of the n-type semiconductor cathode electrode 3 is increased, the n-type semiconductor cathode electrode 3 is provided so as to give a low base resistance. The length of the metal base ohmic contact electrode 5 formed by surrounding is also increased. Therefore, there is a problem that the length of the path through which the current flows through the metal base ohmic contact electrode 5 is increased, thereby increasing the base resistance of the element.
[0006]
  Accordingly, an object of the present invention is to provide a heterojunction bipolar transferred electron transistor capable of obtaining a high oscillation output power with a low base resistance and thermal resistance, and a voltage controlled frequency variable oscillator, injection locking oscillator, and It is to provide a self-oscillating mixer and a receiving device.
[0007]
[Means for Solving the Problems]
  In order to achieve the above object, the first invention provides:
  In a heterojunction bipolar transfer electron transistor in which a semiconductor anode layer, a semiconductor active layer, a semiconductor base layer, and a semiconductor cathode layer are sequentially laminated,
  The semiconductor cathode layer is divided into a plurality of cathode regions.And
  The interval between the narrowest portions in the cathode region adjacent to each other in the extending direction of the interface between the semiconductor base layer and the semiconductor active layer is not more than twice the thickness of the semiconductor active layer in the direction perpendicular to the interface.Ru
It is characterized by that.
[0008]
  According to the above configuration, the semiconductor cathode layer formed on the semiconductor base layer is divided into a plurality of cathode regions. Therefore, the interface between the cathode region and the semiconductor base layer is reduced, and the thermal resistance is reduced. In addition, the active base region immediately below the cathode region and the inactive base region in the vicinity thereof become smaller and the base resistance becomes lower..
[0009]
  Furthermore, the interval between the narrowest portions in the cathode region adjacent to each other in the extending direction of the interface between the semiconductor base layer and the semiconductor active layer is set to be not more than twice the thickness of the semiconductor active layer in the direction perpendicular to the interface. ing. Therefore, the gun domains generated in the semiconductor active layer immediately below each cathode region overlap, the electric fields are coupled to each other, and synchronous propagation and oscillation occur, thereby obtaining a high oscillation output power.
[0010]
  In addition, the first embodiment
  In the heterojunction bipolar transfer electron transistor of the first invention,
  The plurality of cathode regions are formed adjacent to each other on a single semiconductor base layer, and are electrically connected to each other through metal electrodes.
[0011]
  According to this embodiment, the plurality of cathode regions are electrically connected to each other through the metal electrodes and function as one semiconductor cathode layer. Thus, when the area of the semiconductor cathode layer is substantially increased and the interval between the narrowest portions in each cathode region is not more than twice the thickness of the semiconductor active layer, the semiconductor active layer immediately below each cathode region is formed. The generated gun domains are combined to cause synchronous oscillation, and high oscillation output power can be obtained. In addition, when the interval between the cathode regions is larger than twice the thickness of the semiconductor active layer, each gun domain is synchronously propagated by connecting a resonance circuit having a high external Q value, and high oscillation is achieved. Output power can be obtained.
[0012]
  MaTheSecondExamples of
  In the heterojunction bipolar transfer electron transistor of the first invention,
  The interface between each of the cathode regions and the semiconductor base layer is rectangular.
[0013]
  According to this embodiment, since the interface between each of the cathode regions and the semiconductor base layer is rectangular, the narrowest region of the adjacent cathode regions can be obtained long over one side of the rectangle. Therefore, by electrically connecting the plurality of cathode regions to each other with metal electrodes, when the interval between the narrowest portions is not more than twice the thickness in the direction perpendicular to the interface in the semiconductor active layer, The coupling of the cancer domain generated in the semiconductor active layer immediately below each cathode region becomes stronger, and a higher oscillation output power can be obtained. Furthermore, it becomes easy to create a device.
[0014]
  Also,ThirdExamples of
  the aboveSecondIn the heterojunction bipolar transferred electron transistor of the embodiment of
  Each of the cathode regions is formed in a rectangular shape having a long side and a short side, and the long side is arranged adjacent to the long side of another cathode region.
[0015]
  According to this embodiment, the rectangular cathode regions are arranged with their long sides adjacent to the long sides of the other cathode regions. Therefore, compared with the case where the short sides are arranged adjacent to each other, the coupling of the cancer domain generated in the semiconductor active layer immediately below each cathode region becomes stronger, and synchronization transmission is likely to occur.
[0016]
  Also,4thExamples of
  the aboveThirdIn the heterojunction bipolar transferred electron transistor of the embodiment of
  The cathode regions are arranged in two rows in the longitudinal direction.
[0017]
  According to this embodiment, since the cathode regions are arranged in two rows in the longitudinal direction, the thermal resistance is lowered.
[0018]
  Also,5thExamples of
  In the heterojunction bipolar transfer electron transistor of the first invention,
  The impurity doping concentration in the semiconductor active layer decreases from the semiconductor base layer side toward the semiconductor anode layer side.
[0019]
  According to this embodiment, since the impurity doping concentration in the semiconductor active layer is provided with a gradient, the current density distribution in the electron traveling direction (depth direction) is offset, and the time average electric field strength in the device is increased. It becomes constant to some extent. Thus, efficiency and output power are increased.
[0020]
  Also,6thExamples of
  In the heterojunction bipolar transfer electron transistor of the first invention,
  The semiconductor active layer is composed of at least two layers, and the impurity doping concentration in one of the layers is reduced from the semiconductor base layer side toward the semiconductor anode layer side.
[0021]
  According to this embodiment, since the impurity doping concentration in one layer of the semiconductor active layer composed of at least two layers is provided with a gradient, the efficiency and the output power are improved.
[0022]
  Also,7thExamples of
  the above5thExamples of or6thIn the heterojunction bipolar transferred electron transistor of the embodiment of
  The decrease in the impurity doping concentration in the semiconductor active layer is proportional to the distance from the semiconductor base layer side.
[0023]
  According to this embodiment, the efficiency and the output power are further increased.
[0024]
  Also,8thExamples of
  In the heterojunction bipolar transfer electron transistor of the first invention,
  A ballast layer is laminated on each of the cathode regions.
[0025]
  In the above heterojunction bipolar transferred electrons, unlike the case of Gunn diodes and monopolar transferred electrons disclosed in U.S. Pat. No. 3,918,009, positive feedback due to heat (current increases as temperature rises and temperature increases). Rises) and current concentration occurs (current variations occur due to temperature variations in the element, which are amplified by positive feedback, resulting in a large distribution of current density in the element).
[0026]
  According to this embodiment, ballast resistors are stacked on each of the cathode regions. Therefore, current concentration due to a planar temperature distribution within each cathode region and between each cathode region as described above is prevented, and the time-average electric field strength of the entire device becomes constant. Thus, a decrease in oscillation intensity is prevented and a stable high output oscillation output can be obtained.
[0027]
  The voltage controlled frequency variable oscillator of the second invention is
  It is characterized in that the heterojunction bipolar transferred electron transistor of the first invention is used.
[0028]
  In the case of a gun oscillation element, the oscillation frequency is mainly determined by the length of the active layer and the electron saturation speed. However, the oscillation frequency can be changed by changing the saturation voltage by changing the bias voltage. For example, in the case of a Gunn diode, it functions as a voltage-controlled frequency variable oscillator by changing the voltage between the cathode and the anode, but the current also greatly fluctuates as the voltage changes. As a result, the current consumption and the output power are fluctuated simultaneously, and the characteristics as a voltage controlled frequency variable oscillator are deteriorated. In the case of the monopolar transferred electron transistor disclosed in US Pat. No. 3,918,009, a bias voltage is applied between the cathode and the anode, so that the voltage between the base and the anode changes, the anode current changes, and the consumption Current and output power will fluctuate simultaneously.
[0029]
  According to the above configuration, the voltage controlled frequency variable oscillator is formed using the heterojunction bipolar transferred electron transistor of the first invention. Here, in the heterojunction bipolar transferred electron transistor, the anode current is controlled by the voltage between the base and the cathode, and has a property that hardly depends on the voltage between the base and the anode. . Therefore, by changing the voltage between the base and the anode, the oscillation frequency can be changed without changing the anode current, and a voltage controlled frequency variable oscillator having a favorable characteristic with less fluctuation of the oscillation output with respect to the oscillation frequency. can get.
[0030]
  The injection locked oscillator of the third invention is
  It is characterized in that the heterojunction bipolar transferred electron transistor of the first invention is used.
[0031]
  The gun oscillation element varies in oscillation frequency and oscillation output depending on the voltage condition and current condition of the element. When the synchronization signal is injected, the above operating conditions fluctuate, so that the oscillation frequency is drawn to an integral multiple of the synchronization signal frequency. In the case of a Gunn diode, a sync signal is injected between the cathode and the anode. However, as described above, since the operating condition of the Gunn diode needs to be affected by the power of the sync signal, the injection signal requires a large amount of power. It is.
[0032]
  According to the above configuration, the injection locked oscillator is formed using the heterojunction bipolar transferred electron transistor of the first invention. Here, in the case of the heterojunction bipolar transferred electron transistor, since the input of the synchronization signal between the base and the cathode leads to amplification of the anode current, the base-anode is compared with the Gunn diode with the same power injection signal. The voltage and current between them can be changed greatly. Therefore, if the same oscillation output is obtained, the power of the injection signal can be reduced as compared with the Gunn diode. Alternatively, it is possible to take a wide frequency pullable frequency that can be injection-synchronized as compared with the Gunn diode, depending on the intensity of a constant injection signal.
[0033]
  Further, in the case of the above Gunn diode, the oscillation output may flow backward to the synchronization signal source to increase the phase noise of the synchronization signal source. In the case of the monopolar transferred electron transistor disclosed in U.S. Pat.No. 3,918,009, since the synchronization signal is input from the gate terminal, the isolation from the anode to the gate is better than in the case of the Gunn diode, Less backflow of oscillation output to the synchronization signal source. However, it is still not sufficient, and it is necessary to prevent backflow by a filter or buffer amplifier.
[0034]
  According to the above configuration, since the heterojunction bipolar transferred electron transistor has a very large value as the early voltage in the bipolar transistor, the isolation from the anode to the cathode is very good. Therefore, the backflow of the oscillation signal to the cathode side is effectively prevented, and when the synchronization signal is input from the cathode with the base grounded, the oscillation output does not flow back to the synchronization signal source and the phase noise does not increase. .
[0035]
  The self-oscillation mixer of the fourth invention is
  It is characterized in that the heterojunction bipolar transferred electron transistor of the first invention is used.
[0036]
  The Gunn oscillation element can form a self-oscillation mixer having both an oscillator and a mixer function by utilizing the fact that the oscillation output varies depending on the voltage condition and current condition of the element. In the case of a Gunn diode, since input / output is generated between the same cathode and anode, the isolation between signals is poor, and there is no conversion gain because it is a two-terminal element. In the case of the monopolar transferred electron transistor disclosed in U.S. Pat. No. 3,918,009, the isolation from the anode to the gate is better than that in the case of the Gunn diode, but it is not sufficient. Without backflow prevention due to, the amplifier on the input side may be saturated with a large local signal.
[0037]
  According to the above configuration, since the heterojunction bipolar transferred electron transistor has a very large value as the early voltage in the bipolar transistor, the isolation from the anode to the cathode is very good. Therefore, the leakage of the oscillation signal to the cathode side can be eliminated, and when the input signal is input from the cathode with the base grounded, the oscillation output can be easily isolated.
[0038]
  Furthermore, since the input of the synchronizing signal between the base and the cathode of the heterojunction bipolar transferred electron transistor has an amplifying action on the anode current, a conversion gain may be obtained. In addition, in the case of a mixer using a normal transistor, the high frequency characteristics of the transistor affect the conversion efficiency of the mixer. ing. However, in the case of the above heterojunction bipolar transferred electron transistor, the delay time (very small) between the cathode and base determines the limit of the high frequency signal, so the response time is fast and it can be used at higher frequencies. Become.
[0039]
  The receiving device of the fifth invention is
  The self-oscillation mixer of the fourth invention is used.
[0040]
  In the case of communication in a frequency band of 30 GHz or higher, broadband communication can be easily performed, and radio waves do not reach far. Therefore, broadband communication such as data and images in the premises and home is relatively free. Can be used. However, since the gain of the amplifying element is small and the component price is high in the frequency band of 30 GHz or higher, amplification is performed with a circuit at a low intermediate frequency where components can be obtained as easily as possible, preferably at an intermediate frequency of about 1 GHz to 2 GHz. In many cases. Therefore, in the receiver, the intermediate frequency includes a frequency that is 5% or less of the transmission frequency, and the local oscillation frequency and the high-frequency signal frequency are close to each other. Therefore, it is difficult for the receiver to prevent local signal leakage from the mixer unit with a filter or an antenna.
[0041]
  According to the above configuration, since the self-oscillation mixer according to the fourth aspect of the present invention is used for the receiver, there is no leakage to the high frequency signal input at the local oscillation frequency. Therefore, an excellent receiver can be obtained.
[0042]
DETAILED DESCRIPTION OF THE INVENTION
  Hereinafter, the present invention will be described in detail with reference to the illustrated embodiments. In a heterojunction bipolar transferred electron oscillator oscillator or mixer, gun oscillation occurs in the active layer and an oscillation signal appears between the anode electrode and the base electrode (usually ground).
[0043]
  The cathode electrode serves to apply a DC bias to the device, but is almost completely separated from the high frequency signal. FIG. 4 shows a schematic cross-sectional view of a heterojunction bipolar transferred electron device. In FIG. 4, 11 is an n-type semiconductor anode layer, 12 is an n-type active layer, 13 is a p-type semiconductor base layer, and 14 is an n-type semiconductor cathode layer. Then, a cancer domain is generated in the n-type active layer 12 directly adjacent to the p-type semiconductor base layer 13 immediately below the n-type semiconductor cathode layer 14, and moves across the n-type active layer 12 while spreading at an angle of 40 degrees. To do. The arrows in FIG. 4 are the flow of electrons.
[0044]
  (Reference example)
  Figure 1 shows the bookReference exampleFIG. 2 is a cross-sectional view during operation of a two-finger heterojunction bipolar transferred electron transistor having two cathode layers in FIG. In FIG. 1, 15 is n.⁺Is a type GaAs anode layer, 16 is an n type GaAs active layer, and 17 is p⁺The reference numeral 18 denotes a type GaAs base layer, 18 denotes a first n-type AlGaAs cathode layer, and 19 denotes a second n-type AlGaAs cathode layer.
[0045]
  The first n-type AlGaAs cathode layer 18 and the second n-type AlGaAs cathode layer 19 are spatially separated by a distance larger than twice the thickness of the n-type GaAs active layer 16, and each n-type AlGaAs cathode layer 18, There is no interaction between the cancer domains generated in the n-type GaAs active layer 16 immediately below 19. Here, the curve in the n-type GaAs active layer 16 is a contour line obtained by connecting iso-electron concentration points during device operation, and is obtained by a two-dimensional device simulator. In addition, 20 is the 1st cancer domain which has propagated, 21 is the 2nd cancer domain which has propagated.
[0046]
  Since the first and second gun domains 20 and 21 based on the first and second n-type AlGaAs cathode layers 18 and 19 propagate independently, the oscillation signals generated in the cathode layers 18 and 19 are essentially independent. As long as the output is not connected to a resonant circuit having a high external Q value, the frequency and phase of each are different. This reduces the output power of the multi-finger element. This problem is unique to heterojunction bipolar transferred electrons and is not found in bipolar transistors.
[0047]
  2 shows a two-finger heterojunction bipolar transferred electron in which the first and second n-type AlGaAs cathode layers 18 and 19 shown in FIG. 1 are spatially separated by a distance greater than twice the thickness of the n-type GaAs active layer 16. The simulation result of the output current density of a triode is shown. As described above, there is no interaction between the first and second cancer domains 20, 21 under the two n-type AlGaAs cathode layers 18, 19. Oscillation occurs in two separate fundamental waves and their harmonics, and the maximum current density of the fundamental wave having a high current density is 9000 A / cm.²It is. The two separate fundamental waves and harmonics are generated by two independent first and second gun domains 20 and 21, and a fundamental wave F1 having a frequency of 106 GHz and a harmonic wave 2F1 of this fundamental wave F1 (frequency 212 GHz), 3F1 (frequency 316 GHz), a fundamental wave F2 having a frequency 140 GHz, and a harmonic 2F2 (frequency 280 GHz) of the fundamental wave F2. As described above, it is obvious that the device performance is deteriorated when the second n-type AlGaAs cathode layer 19 is provided at a non-optimal position where the two cancer domains are not coupled and synchronized.
[0048]
  In order to make the simulation results more realistic, random conditions that cause asymmetry of the initial electric field distribution were adopted as the initial conditions. However, this asymmetry did not disappear even when the simulation was advanced. The device shown in FIG. 1 has two cathode regions, and each gun domain 20, 21 appears to be uncoupled and in a different propagation state. This is predicted that the output power is greatly reduced without synchronizing the signals generated in the gun domains 20 and 21. Therefore, a resonant circuit having a high external Q value is connected to this element. By doing so, synchronous propagation of the two cancer domains 20, 21 can be achieved, and the output current can be increased.
[0049]
  FIG. 3 is a simulation result of the output current density of a single finger heterojunction bipolar transferred electron transistor having the same cathode finger width as the two finger heterojunction bipolar transferred electron transistor shown in FIG. In this simulation result, the magnitude of the output current density of the fundamental wave F0 is 12500 A / cm in a short circuit.²It is.
[0050]
  As mentioned above, the bookReference example, The first and second n-type AlGaAs cathode layers 18 and 19 are spatially separated by a distance larger than twice the thickness of the n-type GaAs active layer 16 to provide a two-finger heterojunction bipolar transferred electron Configure the Aether. In this case, the two first and second gun domains 20 and 21 generated in the n-type GaAs active layer 16 immediately below the two n-type AlGaAs cathode layers 18 and 19 propagate independently of each other and pass through the cathode layers 18 and 19. The generated oscillation signals are also independent of each other. Therefore, the maximum current density of the fundamental wave F2 having a high current density is 9000 A / cm.²Output current density of a single finger heterojunction bipolar transferred electron transistor with the same cathode finger width of 12500 A / cm²Smaller than.
[0051]
  So bookReference exampleIn this case, a resonance circuit having a high external Q value is connected to the triode element. By doing so, the two cancer domains 20 and 21 can be propagated synchronously, and the output current can be increased.
[0052]
  Ie bookReference exampleIn the heterojunction bipolar transferred electron device of⁺The n-type AlGaAs cathode layer formed on the type GaAs base layer 17 is divided into two cathode layers, a first n-type AlGaAs cathode layer 18 and a second n-type AlGaAs cathode layer 19. Then, the metal cathode ohmic contact electrodes formed on the individual cathode layers are electrically connected to each other, thereby substantially increasing the area of the cathode layer and obtaining a high oscillation output power. In that case, the individual cathode layers 18 and 19 can have the same size and shape as the cathode layer in the case of the conventional single finger element, so that the interface between the cathode region and the base region is a small rectangle. Thus, the thermal resistance can be suppressed. Further, the base resistance can be reduced by reducing the active base region directly under the cathode region and the inactive base region in the vicinity thereof.
[0053]
  However, thisReference exampleIf the two cathode layers 18 and 19 are separated from each other by a distance larger than twice the thickness of the active layer 16, the cathode layers 18 and 19 function individually and the effect of increasing the area of the cathode layer is obtained. Absent. Therefore, in this case, a resonance circuit having a high external Q value is connected to propagate the two gun domains 20 and 21 synchronously to obtain a high oscillation output power.
[0054]
  BookReference exampleIn the above description, a two-finger heterojunction bipolar transfer electron transistor having two cathode layers has been described as an example. However, three or more cathode layers are spatially separated from each other by a distance greater than twice the thickness of the active layer. The same characteristics can be obtained even with a multi-finger heterojunction bipolar transfer electron transistor provided remotely. Therefore, the same effect can be obtained by connecting a resonant circuit having a high external Q value.
[0055]
  <FirstEmbodiment>
  FIG. 5 is a cross-sectional view of a two-finger heterojunction bipolar transferred electron transistor in the present embodiment. In FIG. 5, 25 is n.⁺Type GaAs anode layer, 26 is n-type GaAs active layer, 27 is p⁺A type GaAs base layer, 28 is a first n-type AlGaAs cathode layer, 29 is a second n-type AlGaAs cathode layer, and 30 is a common propagating cancer domain.
[0056]
  Here, the first and second n-type AlGaAs cathode layers 28 and 29 are spatially separated by a distance not more than twice the thickness of the n-type GaAs active layer 26. In this case, the cancer domains generated and spread in the n-type GaAs active layer 26 immediately below the respective n-type AlGaAs cathode layers 28 and 29 are overlapped, the electric fields are coupled to each other, and the propagation and synchronization are synchronized even without an external circuit having a high Q value. Oscillation occurs. the aboveReference exampleAs in the case of, random conditions were introduced as initial conditions, but as the simulation progressed, asymmetry disappeared and synchronous domain propagation occurred immediately.
[0057]
  FIG. 6 is a simulation result of the output current density of the two-finger heterojunction bipolar transferred electron transistor in the present embodiment. Here, as described above, the distance between the two n-type AlGaAs cathode layers 28 and 29 is less than twice the thickness of the n-type GaAs active layer 26, the two generated cancer domains overlap, and the electric field is coupled to each other. Thus, synchronous propagation and oscillation occur. As a result, in this element, one fundamental wave F0 having a frequency of 100 GHz and harmonics 2F0 (frequency 200 GHz) and 3F0 (frequency 300 GHz) of the fundamental wave F0 are generated, and the output current density at the fundamental wave F0 is large. About 14,000 A / cm²It is.
[0058]
  This value is due to the reduced capacitance between the two cathode layers, resulting in an output current density of 12500 A / cm for the single finger device shown in FIG.²Slightly larger. This indicates that the output power increases as the element area increases. In addition, output signals of fundamental waves and harmonics other than the fundamental wave F0 and harmonics 2F0 and 3F0 are not seen.
[0059]
  As described above, in the present embodiment, the first and second n-type AlGaAs cathode layers 28 and 29 are spatially separated by a distance equal to or less than twice the thickness of the n-type GaAs active layer 26, Constructs a two-finger heterojunction bipolar transferred electron transistor. In this case, the two gun domains generated and spread in the n-type GaAs active layer 26 immediately below the two n-type AlGaAs cathode layers 28 and 29 overlap, the electric fields are coupled to each other, and are synchronized even without an external circuit having a high Q value. Propagation and oscillation occur. Therefore, the output current density of the fundamental wave F0 is about 14000 A / cm.²Output current density of a single finger heterojunction bipolar transferred electron transistor with the same cathode finger width of 12500 A / cm²Higher oscillation output power can be obtained without connecting a resonance circuit having a high external Q value.
[0060]
  In the present embodiment, a two-finger heterojunction bipolar transferred electron transistor having two cathode layers has been described as an example. However, three or more cathode layers are spatially formed with an active layer thickness of 2 The same effect can be achieved even with multi-finger heterojunction bipolar transferred electrons provided apart from each other by a distance of twice or less.
[0061]
  <SecondEmbodiment>
  FIG. 7 shows the above shown in FIG.FirstIt is a figure which shows typically an example of the element planar structure of the multi-finger heterojunction bipolar transfer electron triode whose mutual space | interval in the some cathode layer in embodiment is 2 times or less of the thickness of an active layer.
[0062]
  In FIG. 7, 31 is an n-type semiconductor anode layer, 32 is a p-type semiconductor base layer, 33 is an n-type semiconductor cathode layer, 34 is a metal anode ohmic contact electrode, 35 is a metal base ohmic contact electrode, and 36 is a metal cathode ohmic contact. Electrode. In FIG. 7, for easy understanding of each region, a metal anode electrode connected to the metal anode ohmic contact electrode 34, a metal base electrode connected to the metal base ohmic contact electrode 35, and a metal cathode ohmic contact. A metal cathode electrode connected to the electrode 36 is omitted. The active layer is hidden under the p-type semiconductor base layer 32.
[0063]
  In the present embodiment, a plurality of rectangular cathode layers 33 are arranged in a straight line with their short sides adjacent to each other. The metal base ohmic contact electrode 35 completely surrounds each cathode layer 33. In this case, the interval between the narrowest portions (hereinafter referred to as the minimum separation distance) Xmin between the short sides of each cathode layer 33 is 2.4 μm or less (ie, the active layer) with respect to a typical active layer thickness of 1.2 μm. Less than twice the thickness). Therefore, aboveFirstAs in the case of the embodiment, since a plurality of gun domains generated and spread in the active layer immediately below each cathode layer 33 overlap and the electric field is coupled to each other, synchronized propagation and oscillation occur, resulting in high oscillation output power. It is obtained.
[0064]
  However, since the plurality of cathode layers 33 have their short sides adjacent to each other, the coupling between the cancer domains immediately below the adjacent cathode regions is relatively weak. However, there is an advantage that the thermal resistance is low.
[0065]
  <ThirdEmbodiment>
  FIG. 8 is a diagram showing an element plan structure different from that in FIG. 7 in a multi-finger heterojunction bipolar transferred electron transistor in which the distance between the cathode layers is not more than twice the thickness of the active layer.
[0066]
  In FIG. 8, 41 is an n-type semiconductor anode layer, 42 is a p-type semiconductor base layer, 43 is an n-type semiconductor cathode layer, 44 is a metal anode ohmic contact electrode, 45 is a metal base ohmic contact electrode, and 46 is a metal cathode ohmic contact. Electrode. In FIG. 8, the metal anode electrode, the metal base electrode, and the metal cathode electrode are omitted. The active layer is hidden under the p-type semiconductor base layer 42.
[0067]
  In the present embodiment, a plurality of rectangular cathode layers 43 are arranged in a straight line with their long sides adjacent to each other. The metal base ohmic contact electrode 45 completely surrounds each cathode layer 43. In this case, the minimum separation distance Xmin between the long sides of each cathode layer 43 is not more than twice (4 μm) the typical active layer thickness of 2 μm. Therefore, a plurality of gun domains that are generated and spread in the active layer immediately below each cathode layer 43 are overlapped, and the electric fields are coupled to each other, and synchronous propagation and oscillation occur to obtain high oscillation output power.
[0068]
  The plurality of cathode layers 43 are adjacent to each other on their long sides, soSecondCompared to the case where the short sides are adjacent to each other as in the embodiment, the coupling between the cancer domains immediately below the adjacent cathode region is relatively strong. However, it has a relatively high thermal resistance.
[0069]
  <4thEmbodiment>
  FIG. 9 is a diagram showing an element plan structure different from those in FIGS. 7 and 8 in a multi-finger heterojunction bipolar transferred electron transistor in which the distance between the cathode layers is not more than twice the thickness of the active layer.
[0070]
  In FIG. 9, 51 is an n-type semiconductor anode layer, 52 is a p-type semiconductor base layer, 53 is an n-type semiconductor cathode layer, 54 is a metal anode ohmic contact electrode, 55 is a metal base ohmic contact electrode, and 56 is a metal cathode ohmic contact. Electrode. In FIG. 9, the metal anode electrode, the metal base electrode, and the metal cathode electrode are omitted. The active layer is hidden under the p-type semiconductor base layer 52.
[0071]
  In the present embodiment, a plurality of rectangular cathode layers 53 are arranged adjacent to each other with their long sides shifted from each other. The metal base ohmic contact electrode 55 completely surrounds each cathode layer 53. In this case, the minimum separation distance Xmin between the long sides of each cathode layer 53 is not more than twice (3 μm) the typical active layer thickness of 1.5 μm. Therefore, a plurality of gun domains that are generated and spread in the active layer immediately below each cathode layer 53 are overlapped, and the electric fields are coupled to each other, and synchronous propagation and oscillation occur to obtain high oscillation output power.
[0072]
  According to the present embodiment, since the plurality of cathode layers 53 are adjacent to each other with their long sides approximately half the distance of the thickness of the active layer, the guns immediately below the adjacent cathode region are disposed. It is possible to achieve both a strong bond between domains and a low thermal resistance.
[0073]
  <5thEmbodiment>
  FIG. 10 is a diagram showing an element plan structure different from those of FIGS. 7 to 9 in the multi-finger heterojunction bipolar transferred electron transistor in which the distance between the cathode layers is not more than twice the thickness of the active layer. In FIG. 10, 61 is an n-type semiconductor anode layer, 62 is a p-type semiconductor base layer, 63 is an n-type semiconductor cathode layer, 64 is a metal anode ohmic contact electrode, 65 is a metal base ohmic contact electrode, and 66 is a metal cathode ohmic contact. Electrode. In FIG. 10, the metal anode electrode, the metal base electrode, and the metal cathode electrode are omitted. The active layer is hidden under the p-type semiconductor base layer 62.
[0074]
  In the present embodiment, a plurality of rectangular cathode layers 63 are arranged in an oblique straight line and in two rows with their long sides being shifted and adjacent to each other. The metal base ohmic contact electrode 65 completely surrounds each cathode layer 63. In this case, the minimum separation distance Xmin between the long sides of each cathode layer 63 is twice (2 μm) or less of a typical active layer thickness of 1 μm. Therefore, a plurality of gun domains generated and spread in all the active layers immediately below the cathode layer 63 overlap, the electric fields are coupled to each other, and synchronous propagation and oscillation occur to obtain a high oscillation output power.
[0075]
  According to the present embodiment, since the plurality of cathode layers 63 are adjacent to each other at approximately half their mutual long sides at intervals of twice or less the thickness of the active layer, A strong bond between the domains can be provided, and a low thermal resistance can be realized.
[0076]
  In this case, if the distance between the short sides of each cathode layer 63 is set to be twice or less the thickness of the active layer, the coupling between the respective gun domains can be further enhanced. Will also increase.
[0077]
  <6thEmbodiment>
  FIG. 11 is a diagram showing an element plan structure different from those in FIGS. 7 to 10 in the multi-finger heterojunction bipolar transferred electron transistor in which the interval between the plurality of cathode layers is not more than twice the thickness of the active layer. In FIG. 11, 71 is an n-type semiconductor anode layer, 72 is a p-type semiconductor base layer, 73 is an n-type semiconductor cathode layer, 74 is a metal anode ohmic contact electrode, 75 is a metal base ohmic contact electrode, and 76 is a metal cathode ohmic contact. Electrode. In FIG. 11, the metal anode electrode, the metal base electrode, and the metal cathode electrode are omitted. The active layer is hidden under the p-type semiconductor base layer 72.
[0078]
  In the present embodiment, the plurality of rectangular cathode layers 73 are arranged in two rows in the longitudinal direction, with the long sides and the short sides adjacent to each other, and the adjacent long sides shifted by approximately half each other. It is attached. The metal base ohmic contact electrode 75 completely surrounds each cathode layer 73. In this case, the minimum separation distance Xmin between the long sides of each cathode layer 73 is not more than twice (3 μm) the typical active layer thickness of 1.5 μm. Therefore, a plurality of gun domains generated and spread in the active layer immediately below the cathode layer 73 are overlapped, and the electric field is coupled to each other, and synchronous propagation and oscillation occur, resulting in high oscillation output power.
[0079]
  According to the present embodiment, since the plurality of cathode layers 73 are adjacent to each other with their long sides approximately half the distance of the thickness of the active layer, the guns directly below the adjacent cathode region are disposed. A strong bond between the domains can be provided, and a low thermal resistance can be realized. Furthermore, since all the cathode layers 73 are arranged in a zigzag pattern, the above-mentioned5thCompared to the embodiment, the uniformity of the thermal resistance throughout the device can be improved.
[0080]
  In this case, if the distance between the short sides of each cathode layer 73 is set to be twice or less the thickness of the active layer, the coupling between the respective gun domains can be further enhanced, but the thermal resistance Will also increase.
[0081]
  <7thEmbodiment>
  the aboveSecond~6thIn the embodiment, detailed description of the semiconductor structure is omitted for the sake of simplicity. It is also possible to apply conventional techniques to the basic essential structure. For example, between the semiconductor cathode layers 33, 43, 53, 63, 73 and the metal cathode ohmic contact electrodes 36, 46, 56, 66, 76, one cap layer doped at a higher concentration than the semiconductor cathode layer is provided. Addition of the above can reduce the cathode resistance. Alternatively, the manufacturing method can be simplified by thinning the periphery of the active layer. The present embodiment relates to a multi-finger heterojunction bipolar transferred electron transistor in which a cap layer is provided between the semiconductor cathode layer and the metal cathode ohmic contact electrode, and the periphery of the active layer is formed thin.
[0082]
  FIG. 12 shows a cross section of a two-finger heterojunction bipolar transferred electron transistor in the present embodiment. In FIG. 12, reference numeral 81 denotes a GaAs anode layer (800 nm, N_D(Nitrogen dose) = 5 × 10¹⁸cm^-3), 82 is a GaAs active layer (1700 nm, N_D= 2 × 10¹⁶cm^-3→ 3 × 10¹⁶cm^-3), 83 is a GaAs base layer (150 nm, N_A= 4 × 10¹⁹cm^-3).
[0083]
  84 is Al._xGa_1-xAs cathode gradient layer (20 nm, x = 0.2 → 0.35, N_D= 5 × 10¹⁷cm^-3), 85 is Al_0.35Ga_0.65As cathode layer (40 nm, N_D= 5 × 10¹⁷cm^-3), 86 is Al_0.35Ga_0.65As cathode ballast layer (100 nm, N_D= 1 x 10¹⁷cm^-3), 87 is Al_xGa_1-xAs cathode cap gradient layer (50 nm, x = 0.35 → 0.0, N_D= 5 × 10¹⁷cm^-3), 88 is a GaAs cathode cap layer (50 nm, N_D= 5 × 10¹⁸cm^-3).
[0084]
  89 is a metal anode ohmic contact electrode (AuGe / Ni / Au), 90 is a metal base ohmic contact electrode (Ti / Pt / Au), and 91 is a metal cathode ohmic contact electrode (AuGe / Ni / Au).
[0085]
  In the figure, the distance between the narrowest portions (hereinafter referred to as the minimum inter-cathode distance) Xmin between the cathode layers 85 in the figure is 2.5 μm, and the active layer thickness Y is 1.7 μm. Therefore, a plurality of gun domains generated and spread in the active layer 82 immediately below the cathode layers 84 and 85 are overlapped, the electric fields are coupled to each other, and synchronous propagation and oscillation occur. Further, as described above, in the present embodiment, the cathode layers 84 and 85 and the metal cathode ohmic contact electrode 91 are more N than the cathode layers 84 and 85._DA high cathode cap layer 88 is provided. Therefore, the cathode resistance can be reduced. Further, the outer wall of the GaAs active layer 82 is inclined so that the periphery is thin. Therefore, the completeness of the element can be improved.
[0086]
  <8thEmbodiment>
  In the present embodiment, a cap layer is provided between the semiconductor cathode layer and the metal cathode ohmic contact electrode, and N is formed on the active layer._DThe present invention relates to a multi-finger heterojunction bipolar transferred electron gradient with a slope.
[0087]
  FIG. 13 shows a cross section of a two-finger heterojunction bipolar transferred electron transistor in the present embodiment. In FIG. 13, reference numeral 101 denotes a GaAs anode layer (500 nm, N_D= 5 × 10¹⁸cm^-3), 102 is In_0.49Ga_0.51P etch stop layer (20nm, N_D= 1 x 10¹⁸cm^-3), 103 is a GaAs active layer (800 nm, N_D= 1.5 × 10¹⁶cm^-3), 104 is a GaAs active layer (1000 nm, N_D= 1.5 × 10¹⁶cm^-3→ 3 × 10¹⁶cm^-3), 105 is a GaAs base layer (160 nm, N_A= 2 × 10¹⁹cm^-3), 106 is In_0.49Ga_0.51P base protective layer (30 nm, N_D= 5 × 10¹⁷cm^-3).
[0088]
  Reference numeral 107 denotes a GaAs cathode layer (40 nm, N_D= 1 x 10¹⁸cm^-3), 108 is In_xGa_1-xAs cathode cap gradient layer (50 nm, x = 0.0 → 0.5, N_D= 1 x 10¹⁹cm^-3), 109 is In_0.5Ga_0.5As cathode cap layer (50 nm, N_D= 2 × 10¹⁹cm^-3).
[0089]
  110 is a metal anode ohmic contact electrode (AuGe / Ni / Au), 111 is a metal base ohmic contact electrode (Ti / Pt / Au), and 112 is a metal cathode ohmic contact electrode (Ti / Pt / Au).
[0090]
  In the figure, the minimum inter-cathode distance Xmin is 2 μm, and the active layer thickness Y is 1.8 μm. Therefore, a plurality of gun domains that are generated and spread in the active layer 104 immediately below the cathode layer 107 are overlapped, and the electric fields are coupled to each other, and synchronous propagation and oscillation occur. Further, as described above, in the present embodiment, N is more than the cathode layer 107 between the cathode layer 107 and the metal cathode ohmic contact electrode 112._DHigh cathode cap layers 108 and 109 are provided. Therefore, the cathode resistance can be reduced. Further, the GaAs active layer 104 has N_DThe slope is provided. Therefore, the completeness of the element can be improved. That is, N in the GaAs active layer 104_DThis inclination cancels out the current density distribution in the traveling direction (depth direction) of electrons, makes the time-average electric field strength in the device constant to some extent, and increases the efficiency and output power.
[0091]
  <9thEmbodiment>
  In the present embodiment, a ballast layer is provided between the semiconductor cathode layer and the metal cathode ohmic contact electrode to prevent current concentration in the cathode finger, and the time-average electric field strength of the entire device is made constant so that the oscillation intensity is constant. In particular, the present invention relates to a multi-finger heterojunction bipolar transferred electron device that prevents a decrease in the frequency and achieves a stable high oscillation output.
[0092]
  FIG. 14 shows a cross section of a two-finger heterojunction bipolar transferred electron transistor in the present embodiment. In FIG. 14, 121 is a GaAs anode layer, 122 is a GaAs active layer (impurity concentration gradient), 123 is a GaAs base layer, and 124 is Al._0.3Ga_0.7As cathode layer, 125 is Al_0.35Ga_0.65As ballast layer, 126 is Al_xGa_1-xAs cathode gradient layer, 127 is n⁺GaAs cathode cap layer.
[0093]
  Also, 128 is a metal anode ohmic contact electrode, 129 is a metal base ohmic contact electrode, and 130 is a metal cathode ohmic contact electrode.
[0094]
  FIG. 15 schematically shows a state of current concentration in the heterojunction bipolar transferred electron transistor. FIG. 15 shows a longitudinal section of the anode layer 131, the active layer 132, and the base layer 133 in order to display the state of current concentration. Although the cathode layer and each electrode are omitted, a cathode layer forming region 134 is shown on the base layer 133.
[0095]
  Originally, the current must flow uniformly within the range of the cathode layer forming region 134. However, since the temperature rises due to the current 137 flowing in the active layer 132, the current 137 concentrates in the central portion 135 where the temperature rise is large as shown in FIG. 15, and the current density in the peripheral portion 136 decreases. is there. In the gun oscillation element, a constant time-average electric field intensity in the active region is one condition of oscillation, and the oscillation intensity is particularly greatly affected. For this reason, the planar current density variation due to the current concentration described above results in variations in time-average electric field strength in the device, resulting in a region that does not oscillate partially or a region in which the oscillation strength is reduced. Alternatively, when the current concentration distribution is large, the oscillation of the element may be completely stopped.
[0096]
  Therefore, in the present embodiment, as shown in FIG. 14, a ballast layer 125 is provided between the cathode layer 124 and the metal cathode ohmic contact electrode 130 to prevent current concentration in the cathode finger, and the active layer The time-average electric field strength in the 122 region is made constant.
[0097]
  FIG. 16 is provided between a cathode layer and a metal cathode ohmic contact electrode in a single finger heterojunction bipolar transfer electron device having the same basic structure as the multi-finger heterojunction bipolar transfer electron device shown in FIG. The effect of the ballast layer. In FIG. 16, the upper and lower limits of the oscillating current of the anode current are compared between those in which the thickness of the ballast layer is changed to 400 nm, 300 nm, 200 nm, and 100 nm and that in which the ballast layer is eliminated.
[0098]
  As shown in FIG. 16, when the thickness of the ballast layer is 400 nm to 200 nm, the heterojunction bipolar transferred electron device is oscillated at an anode current of 70 mA to 82 mA and has a fundamental frequency of about 140 GHz. A second harmonic of about 280 GHz is obtained. However, when the thickness of the ballast layer is smaller than 200 nm, the output decreases, and the range of the anodic current that can be oscillated becomes extremely narrow and substantially constant.
[0099]
  In the heterojunction bipolar transferred electron transistor, unlike the Gunn diode, the current density distribution in the electron traveling direction (depth direction) due to current spreading in the active layer region is inevitable. Therefore, in the present embodiment, by providing an impurity concentration gradient in the active layer 122, the current density distribution in the electron traveling direction (depth direction) is offset, and the time-average electric field strength in the active layer region is kept constant. Like to do. However, when a planar current density distribution of the element (center part 135 and peripheral part 136 in FIG. 15) is generated due to current concentration, the current spread in the horizontal direction in the cross section of FIG. 15 also changes as shown in FIG. It will end up. This is considered to be because the design of the above-mentioned blunt concentration gradient is distorted by current concentration when there is no ballast layer.
[0100]
  However, as shown in FIG. 14, when the ballast layer 125 is formed between the cathode layer 124 and the metal cathode ohmic contact electrode 130, a planar current density distribution (the center portion 135 in FIG. The peripheral portion 136) is eliminated, and the current density distribution in the depth direction is effectively offset by the impurity concentration gradient in the depth direction in the active layer 122. As a result, the time average electric field strength of the entire active layer 122 can be made constant, and the oscillation efficiency can be optimized. In particular, the efficiency can be improved when using a high current with a large temperature rise.
[0101]
  Here, the resistance value of the ballast layer 125 determined by the thickness of the ballast layer 125 needs to be equal to or higher than the cathode layer 124. A larger resistance value of the ballast layer 125 is preferable because current concentration can be prevented. When the resistance value of the ballast layer 125 is large, an extra voltage corresponding to a voltage drop due to the ballast resistance is required between the cathode layer 124 and the base layer 123, so that the resistance value has a practical upper limit. However, since oscillation occurs between the base layer 123 and the anode layer 121, the ballast resistance value itself does not adversely affect the oscillation output and the oscillation frequency width. This point is completely different from other bipolar devices.
[0102]
  By the way, as the material of the ballast layer 125, in addition to a semiconductor material such as AlGaAs, a resistive material that can be stacked on the emitter can be used. In the case of the semiconductor material, it is desirable to use a material having a lower electron mobility than the material of the active layer 122. If a material having a high electron mobility is used, not only the thickness of the ballast layer 125 is increased and the manufacturing difficulty is increased, but also when the blunt substance concentration is decreased to reduce the thickness, the ballast layer 125 is reduced. In some cases, the device operation becomes unstable due to the effect of traveling electrons.
[0103]
  In consideration of the above, the material for the ballast layer 125 is Al._xGa_{1- x}If As (0.15≤x≤0.35) is used, the resistance at low temperature is low and the resistance rises only at the part where the temperature has risen. It is done. Therefore, in the case of an application in which a signal is introduced from the cathode layer 124 (an injection-locked oscillator or a self-oscillation mixer as described in the following embodiment), the loss of the signal is small, which is more preferable.
[0104]
  <10thEmbodiment>
  As described above, a multi-finger heterojunction bipolar transfer electron device comprising a cathode layer divided into a plurality of cathode regions on a single base layer and electrically connecting each cathode region Oscillation output higher than that of a single finger heterojunction bipolar transferred electron transistor by connecting a resonance circuit with a high Q value or making the minimum separation distance Xmin of each cathode region less than twice the thickness of the active layer You can gain power. Therefore, a large oscillation output can be obtained by using the multi-finger heterojunction bipolar transferred electron transistor having the above configuration for a voltage-controlled frequency variable oscillator, an injection locking oscillator, a self-oscillating mixer, a receiving device, and the like. . Hereinafter, the description will be made sequentially.
[0105]
  First, the present embodiment relates to a voltage controlled frequency variable oscillator (VCO) using the multi-finger heterojunction bipolar transferred electron transistor. FIG. 17 is a circuit diagram of the present VCO.
[0106]
  As shown in FIG. 17A, the base electrode 142 of the multi-finger heterojunction bipolar transferred electron device 141 is grounded, and the anode electrode 145 is connected to the first bias circuit (the first inductance 144 in FIG. 17A). The cathode electrode 148 is connected to the cathode power supply terminal 146 via a second bias circuit (second inductance 147 in FIG. 17A). Further, an output terminal 150 is connected to the anode electrode 145 through a DC bias cut capacitor 149. On the other hand, the cathode electrode 148 is grounded via the capacitor 151.
[0107]
  In the above configuration, a positive voltage is applied to the anode electrode 145 from the anode power supply terminal 143 through the first bias circuit 144. On the other hand, a negative voltage is applied from the cathode power supply terminal 146 to the cathode electrode 148 through the second bias circuit 147. The oscillation output is taken out from the anode electrode 145 to the output terminal 150 through the DC bias cut capacitor 149.
[0108]
  FIG. 18 shows a VCO formed by changing the impurity concentration gradient of the GaAs active layer in the single finger heterojunction bipolar transfer electron transistor having the same basic structure as the multifinger heterojunction bipolar transfer electron transistor shown in FIG. Anode-base voltage-frequency characteristics are shown. Here, the first impurity concentration gradient has a Si doping concentration of 1.2 × 10 toward the base side from the anode side.¹⁶cm^-3→ 3.0 × 10¹⁶cm^-3This is a linear change. On the other hand, the second impurity concentration gradient has a Si doping concentration of 1.2 × 10 from the anode region to 900 nm.¹⁶cm^-3To 2.0 × 10¹⁶cm^-3Is continuously changed to 2.0 × 10 at 400 nm up to the base region.¹⁶cm^-3To 3.0 × 10¹⁶cm^-3This is a linear change.
[0109]
  As shown in FIG. 18, a change in the oscillation frequency was observed by changing the voltage between the anode and the base. At that time, the anode current hardly changes. Note that the variation width of the oscillation frequency is wider in the element having the second impurity concentration gradient, in which the blunt concentration gradient of the active layer is closer to the ideal. This is because the time-average electric field strength at which gun oscillation can occur is limited (width), but when the electric field is changed by changing the bias for VCO operation, the time-average electric field strength in the active layer is originally distributed. If it exists, it is considered that a portion exceeding the limit of the time-average electric field strength occurs earlier and oscillation stops (or the output decreases).
[0110]
  Here, the oscillation output is generated between the base and the anode and is not output to the cathode side. However, since the bias between the cathode and the base greatly affects the oscillation, in order to avoid unnecessary noise from being superimposed on the bias, FIG. As shown in (a), it is desirable to ground the cathode electrode 148 at a high frequency.
[0111]
  The base electrode 142 may be grounded at a high frequency, and a stub may be used. FIG. 17B is a circuit diagram of a VCO having a configuration in which a base electrode is connected to a stub. A stub (open stub 163 having a wavelength of λ / 4 with respect to the fundamental oscillation frequency in FIG. 17B) is connected to the base electrode 162 of the multi-finger heterojunction bipolar transferred electron device 161. The anode electrode 164 is connected to the anode power supply terminal 166 via a bias circuit (inductance 165 in FIG. 17B), while the cathode electrode 167 is grounded. Further, an output terminal 169 is connected to the anode electrode 164 through a DC bias cut capacitor 168. Reference numeral 170 denotes a base power supply terminal.
[0112]
  In the above configuration, the circuit is composed of only a positive power supply that applies a positive voltage to the base by grounding the cathode. The base electrodes 142 and 162 are preferably grounded not only with respect to the fundamental oscillation frequency but also with respect to harmonics.
[0113]
  Examples of gun oscillation elements that can change the oscillation frequency by changing the bias voltage include a Gunn diode and a monopolar transferred electron transistor disclosed in US Pat. No. 3,918,009. However, in either case, the current greatly fluctuates with the change of the bias voltage, so that the current consumption and the output power fluctuate at the same time, which deteriorates the characteristics as a VCO.
[0114]
  As described above, according to the present embodiment, the VCO is formed using the heterojunction bipolar transferred electron transistor in which the minimum interval between the adjacent cathode regions is not more than twice the thickness of the semiconductor active layer. Yes. Therefore, a larger oscillation output can be obtained as compared with the case of a single finger heterojunction bipolar transferred electron transistor. In that case, the heterojunction bipolar transferred electron transistor has a property that the anode current hardly depends on the voltage between the base and the anode. Therefore, by changing the voltage between the base and the anode, it is possible to change the oscillation frequency without changing the anode current, and to obtain a VCO having good characteristics with less fluctuation of the oscillation output with respect to the oscillation frequency. It can be done.
[0115]
  <11thEmbodiment>
  The present embodiment relates to an injection locked oscillator (ILO) using the multi-finger heterojunction bipolar transferred electron transistor. FIG. 19 is a circuit diagram of the present ILO.
[0116]
  As shown in FIG. 19A, the base electrode 172 of the multi-finger heterojunction bipolar transferred electron transistor 171 is grounded, and the anode electrode 173 is connected to the first bias circuit (the first inductance 174 in FIG. 19A). The cathode electrode 176 is connected to the cathode power supply terminal 178 via a second bias circuit (second inductance 177 in FIG. 19A). Further, the anode electrode 173 is connected to the output terminal 180 via the first DC bias cut capacitor 179. On the other hand, the cathode electrode 176 is connected to the synchronization signal input terminal 182 via the second DC bias cut capacitor 181.
[0117]
  In the above configuration, a positive voltage is applied from the anode power supply terminal 175 to the anode electrode 173 through the first bias circuit 174. On the other hand, a negative voltage is applied to the cathode electrode 176 from the cathode power supply terminal 178 through the second bias circuit 177. Then, a synchronization signal (−12.5 dBm) having a frequency half the desired oscillation frequency is input from the synchronization signal input terminal 182 to the cathode electrode 176 through the second DC bias cut capacitor 181, and the first DC bias is applied from the anode electrode 173. The oscillation output of the desired oscillation frequency is taken out to the output terminal 180 through the cutting capacitor 179.
[0118]
  FIG. 20 shows the impurity concentration of the GaAs active layer in the single finger heterojunction bipolar transferred electron transistor having the same basic structure as the multifinger heterojunction bipolar transferred electron transistor shown in FIG. The synchronous signal frequency-oscillation frequency characteristic in the ILO produced using the element whose inclination is the first impurity concentration inclination and the element whose inclination is the second impurity concentration inclination is shown. As shown in FIG. 20, the oscillation frequency changes by changing the frequency of the synchronization signal input from the synchronization signal input terminal. It should be noted that the range in which the oscillation frequency matches the integral multiple of the injection locking signal frequency (the pullable frequency range) is such that the element having the second impurity concentration gradient is closer to the ideal non-blunt concentration gradient of the active layer. wide. This is because the synchronization signal between the cathode and the base flows as fluctuations in the anode current, and the time-average electric field strength changes, so that the oscillation is synchronized with the synchronization signal and the oscillation frequency becomes an integral multiple of the synchronization signal. However, if the time-average electric field strength in the active layer originally has a distribution, the influence of the change in the time-average electric field strength due to the synchronization signal becomes relatively small, and the frequency that can be synchronized becomes narrow. Conceivable. Since the oscillation signal is not output to the cathode side, there is no backflow of the oscillation output to the synchronization signal source, and the phase noise does not increase.
[0119]
  The base electrode 172 may be grounded at a high frequency, and a stub may be used. FIG. 19B is a circuit diagram of an ILO having a configuration in which a base electrode is connected to a stub. The base electrode 192 of the multi-finger heterojunction bipolar transferred electron 191 is a stub (in FIG. 19B, a λ / 2 wavelength line 193 and a capacitor 194 with respect to the fundamental oscillation frequency are used as a short stub in high frequency). Is grounded. The anode electrode 195 is connected to the anode power supply terminal 197 via a bias circuit (inductance 196 in FIG. 19B), while the cathode electrode 198 is grounded. Further, an output terminal 200 is connected to the anode electrode 195 via a DC bias cut capacitor 199, and an open stub 201 having a λ / 4 wavelength with respect to the synchronization signal frequency is connected. In addition, a synchronization signal input terminal 203 is connected to the base electrode 192 via a DC bias cut capacitor 202. Reference numeral 204 denotes a base power supply terminal.
[0120]
  In the above configuration, the cathode is grounded, and a circuit with only a positive power supply for applying a positive voltage from the base power supply terminal 204 to the base is obtained. In this ILO circuit, since the short stubs 193 and 194 are connected to the base electrode 192, not only the ground for the fundamental oscillation frequency but also the harmonics are grounded. Since the short stubs 193 and 194 appear to be open from a synchronization signal having a frequency that is half the fundamental oscillation frequency, the synchronization signal is input from the synchronization signal input terminal 203 to the base electrode 192. In this case, since the synchronization signal is current amplified by the parasitic bipolar transistor of the multi-finger heterojunction bipolar transferred electron transistor 191, the same pulling range can be achieved with a small amount of power. Further, by inserting an open stub 201 of λ / 4 wavelength with respect to the synchronization signal frequency on the anode side so that the synchronization signal voltage does not appear on the anode electrode 195, the isolation of each signal can be easily improved. Although a synchronization signal can be input between the base and the anode, a synchronization signal with a large power is required.
[0121]
  In the present embodiment, a synchronization signal having a frequency that is half the desired oscillation frequency is used as the injection locking signal. However, a synchronization signal having a frequency that is 1 / integer of the desired oscillation frequency can also be used. However, the half-frequency case is preferable because the frequency range in which the synchronization signal can be drawn is widest.
[0122]
  Here, the ballast resistance can be minimized by using AlGaAs as the ballast resistance of the multi-finger heterojunction bipolar transferred electron transistors 171 and 191. The smaller the ballast resistance, the smaller the loss due to the ballast resistance of the injection locking signal, and the injection signal power can be further reduced.
[0123]
  There is the Gunn diode as a Gunn oscillation element in which when the synchronizing signal is injected, the oscillation frequency is pulled to an integral multiple of the synchronizing signal frequency. However, in the case of the Gunn diode, since a synchronization signal is injected between the cathode and the anode, a large power synchronization signal is required in order that the operation condition of the Gunn diode is affected by the power of the synchronization signal.
[0124]
  As described above, in the present embodiment, the ILO is formed using a heterojunction bipolar transferred electron transistor in which the minimum distance between the adjacent cathode regions is not more than twice the thickness of the semiconductor active layer. . Therefore, a larger oscillation output can be obtained as compared with the case of a single finger heterojunction bipolar transferred electron transistor. In that case, since the input of the synchronization signal between the base and the cathode in the heterojunction bipolar transfer electron transistor leads to the amplification of the anode current and the voltage, the base- The voltage and current between the anodes can be changed greatly. Therefore, if the same oscillation output is obtained, the power of the injection signal can be reduced as compared with the Gunn diode. Alternatively, the frequency pullable frequency that can be injection-synchronized as compared with the Gunn diode can be widened depending on the intensity of a constant injection signal.
[0125]
  <12thEmbodiment>
  The present embodiment relates to a self-oscillation mixer using the multi-finger heterojunction bipolar transferred electron transistor. FIG. 21 is a circuit diagram of the self-oscillation mixer.
[0126]
  As shown in FIG. 21A, the base electrode 212 of the multi-finger heterojunction bipolar transferred electron transistor 211 is grounded, and the anode electrode 213 is connected to the first bias circuit (first inductance 214 in FIG. 19A). The cathode electrode 216 is connected to the cathode power supply terminal 218 via a second bias circuit (second inductance 217 in FIG. 21A). Further, the anode electrode 213 is connected to the output terminal 220 via the first DC bias cut capacitor 219. On the other hand, the cathode electrode 216 is connected to the signal input terminal 222 via the second DC bias cut capacitor capacitor 221.
[0127]
  In the above configuration, a positive voltage is applied to the anode electrode 213 from the anode power supply terminal 215 through the first bias circuit 214. On the other hand, a negative voltage is applied to the cathode electrode 216 from the cathode power supply terminal 218 through the second bias circuit 217. Then, a synchronizing signal (78 GHz, −12.5 dBm) having a frequency half the desired oscillation frequency and the frequency-converted signal are input from the signal input terminal 222 to the cathode electrode 216 through the second DC bias cut capacitor 221, and the anode The oscillation output of the desired oscillation frequency is taken out from the electrode 213 to the output terminal 220 through the first DC bias cut capacitor 219.
[0128]
  In that case, a 79 GHz frequency conversion signal is observed with respect to an input signal (frequency converted signal) having a frequency of 1 GHz, and a 2 GHz frequency conversion signal is observed with respect to an input signal having a frequency of 80 GHz, for up-conversion and down-conversion. It was found that it can be used for both.
[0129]
  In addition, as shown in FIG. 22, the element (225) having the second impurity concentration gradient that is closer to the ideal impurity concentration gradient of the active layer and the element having the first impurity concentration gradient (the gradient is monotonic) ( 226), in the range 227,228 where the output signal linearly follows the increase in the frequency converted signal (input signal) intensity, the element (225) having the second impurity concentration gradient is more input. It exists up to a large signal power region and has good linearity. This is because there is a limit (width) in the time-average electric field strength at which gun oscillation can occur, but when the signal input between the cathode and base flows as a fluctuation in anode current for mixing, a change occurs in the time-average electric field strength. In addition, if the time-average electric field intensity in the active layer originally has a distribution, it is considered that a portion exceeding the electric field limit is generated earlier and the oscillation output is lowered.
[0130]
  The base electrode 212 may be grounded at a high frequency, and a stub may be used. FIG. 21B is a circuit diagram of a self-oscillating mixer having a configuration in which a base electrode is connected to a stub. The base electrode 232 of the multi-finger heterojunction bipolar transferred electron 231 is a stub (in FIG. 21B, a λ / 2 wavelength line 233 and a capacitor 234 with respect to the fundamental oscillation frequency are used as a short stub in high frequency). Is grounded. The anode electrode 235 is connected to the anode power supply terminal 237 via a bias circuit (inductance 236 in FIG. 21B), while the cathode electrode 238 is grounded via an inductance 239. Further, an output terminal 241 is connected to the anode electrode 235 via a DC bias cut capacitor 240. Further, a signal input terminal 243 is connected to the cathode electrode 238 via a DC bias cut capacitor 242, and an open stub 244 having a wavelength of λ / 2 with respect to the fundamental oscillation frequency is connected. Further, a synchronization signal input terminal 246 is connected to the base electrode 232 via a DC bias cut capacitor 245. Reference numeral 247 denotes a base power supply terminal.
[0131]
  In the above configuration, the base is grounded to the fundamental oscillation frequency and its harmonics by the short stubs 233 and 234, the cathode is grounded in a direct current manner by the inductance 239, and a positive voltage is applied from the base power supply terminal 247 to the base. It becomes a circuit of only the positive power supply which adds. Then, the frequency converted signal is input from the signal input terminal 243, and the short stubs 233 and 234 appear to be open from the synchronization signal having a frequency that is half the fundamental oscillation frequency. Therefore, the synchronization signal from the synchronization signal input terminal 246 to the base electrode 232 And the cathode is grounded with respect to the synchronization signal by an open stub 244 of λ / 4 wavelength with respect to the synchronization signal frequency. In this case, since the synchronization signal is current-amplified by the parasitic bipolar transistor of the multi-finger heterojunction bipolar transferred electron transistor 231, the same pulling range can be achieved with a small amount of power.
[0132]
  In the circuit shown in FIG. 21 (b), the signal input terminal 243 can be simply grounded, and the frequency converted signal can be input from the synchronization signal terminal 246 or the base power supply terminal 247.
[0133]
  In the present embodiment, in particular, on the anode side, an open stub 223 having a λ / 4 wavelength with respect to the fundamental oscillation frequency (in the case of the circuit shown in FIG. 21A), 248 (the circuit shown in FIG. 21B). When the anode is grounded at the local oscillation frequency, the local frequency is generated between the base and anode (both terminals are grounded) and is not output to the cathode side. A mixer that does not output from the multi-finger heterojunction bipolar transferred electron 211, 231 to the outside is configured.
[0134]
  Further, in the present embodiment, a synchronization signal having a frequency half the desired oscillation frequency is used as the injection signal, but a synchronization signal having a frequency that is a fraction of the desired oscillation frequency can also be used. However, the half-frequency case is preferable because the frequency range in which the synchronization signal can be drawn is widest. Further, although the injection locking method is used to stabilize the oscillation frequency, a PLL (Phase Locked Loop Oscillator) can be configured as the VCO.
[0135]
  The Gunn diode and the monopolar transferred electron transistor are examples of Gunn oscillation elements that can constitute a self-oscillating mixer having both an oscillator and a mixer function. However, in the case of the Gunn diode, since input / output is generated between the same cathode and anode, the isolation between signals is poor and there is no conversion gain because it is a two-terminal element. Also, in the case of the above monopolar transferred electron transistor, the isolation from the anode to the gate is not sufficient, and if the backflow prevention by the filter or buffer amplifier is not performed, the amplifier on the input side will be saturated with a large local signal There is.
[0136]
  As described above, in the present embodiment, the self-oscillation mixer is formed using a heterojunction bipolar transferred electron transistor in which the minimum interval between the adjacent cathode regions is not more than twice the thickness of the semiconductor active layer. Yes. Therefore, a larger oscillation output can be obtained as compared with the case of a single finger heterojunction bipolar transferred electron transistor. In that case, since the heterojunction bipolar transferred electron transistor has a very large value as the early voltage in the bipolar transistor, the isolation from the anode to the cathode is very good. Therefore, the leakage of the oscillation signal to the cathode side can be eliminated, and by isolating the oscillation output from the base by grounding the base and inputting the input signal from the cathode. .
[0137]
  <13thEmbodiment>
  The present embodiment relates to a receiving apparatus using the multi-finger heterojunction bipolar transferred electron transistor. The receiving apparatus in the present embodiment uses a self-oscillating mixer shown in FIG. 21 (a) in which the anode is grounded at the local oscillation frequency by an open stub. Hereinafter, transmission / reception between the receiving apparatus and the transmitter will be described.
[0138]
  As shown in FIG. 23, the first transmitter 251 and the second transmitter 252 are fixed stations for each area (adjacent houses in this embodiment), and a broadband information signal to be transmitted is a terrestrial TV ( UHF440MHz ~ 800MHz), BS satellite broadcast intermediate frequency (1GHz ~ 1.35GHz), CS satellite broadcast intermediate frequency (1.29GHz ~ 1.55GHz) are arranged at 440MHz ~ 1960MHz by block converter (Fig. 24 (a)) 256). In CS satellite broadcasting, it is assumed that the right-handed wave and the left-handed-handed wave are frequency-converted from 1.4 GHz to 1.66 GHz and 1.70 GHz to 1.96 GHz, respectively.
[0139]
  The adjacent first and second transmitting stations 251 and 252 shift the frequency of the wideband information signal 256 at the respective transmitting stations so as not to interfere with each other, and the frequency bands 257 and 258 near 60 GHz in FIG. Upconvert and send.
[0140]
  The first receiver 253 receives a radio signal 257 (see FIG. 24B) from the first transmitting station 251 in the area with an antenna, removes unnecessary waves through a filter,SecondAs shown in FIG. 25, the self-oscillation mixer in the embodiment is mixed with the local frequency 259 of 60 GHz and down-converted to the original frequency (440 MHz to 1960 MHz) 260. And it demodulates with a commercially available television receiver or a satellite broadcast television receiver.
[0141]
  In that case, the second receiver 254 (see FIG. 23) used in the vicinity of the boundary line of the area of the adjacent transmitter transmits a radio signal having a frequency band different from the frequency band of the radio signal 257 of the first transmitting station 251. When receiving (for example, the radio signal 258 from the second transmitting station 252), the frequency of the local signal 261 of the second transmitting station 252 is as shown in FIG. When there is a leakage of the local signal 261, the leakage of the local signal 261 becomes a reception interference 255 of the first receiver 253 that is close.
[0142]
  However, as shown in FIG. 21A, each receiver 253, 254 in this embodiment includes a self-oscillation mixer that is grounded at the local oscillation frequency by an open stub 223 having a λ / 4 wavelength with respect to the fundamental oscillation frequency on the anode side. Is used. Therefore, even when a plurality of receivers 253 and 254 are used at the same time as described above, the local signal from the receivers 253 and 254 has a low unnecessary radiation level, so that reception without interfering with each other is received. Is possible. This is because the self-oscillation mixer anode using the multi-finger heterojunction bipolar transferred electron ground is grounded at the local frequency, so there is almost no leakage of the local signal through the power supply circuit etc. This is because there is very little leakage.
[0143]
【The invention's effect】
  As is apparent from the above, the heterojunction bipolar transferred electron transistor of the first invention divides the semiconductor cathode layer formed on the semiconductor base layer into a plurality of cathode regions. Resistance can be lowered. Also,Since the interval between the narrowest portions in the adjacent cathode regions is less than twice the thickness of the semiconductor active layer, the cancer domains generated in the semiconductor active layer immediately below each cathode region overlap, the electric fields are coupled to each other, and Propagation and oscillation occur and high oscillation output power can be obtained.
[0144]
  MaIn the heterojunction bipolar transfer electron transistor of the first embodiment, the plurality of cathode regions are formed adjacent to each other on a single semiconductor base layer, and are electrically connected to each other by metal electrodes. Therefore, the substantial area of the semiconductor cathode layer can be increased. Therefore, if the interval between the narrowest portions between the cathode regions is set to be twice or less the thickness of the semiconductor active layer, the gun domains generated in the semiconductor active layer immediately below the cathode regions are coupled to generate synchronous oscillation. To obtain high oscillation output power.
[0145]
  MaTheSecondIn the heterojunction bipolar transferred electron transistor of this embodiment, since the interface between each of the cathode regions and the semiconductor base layer is rectangular, the narrowest portion of the cathode region adjacent to each other is formed on one side of the rectangle. You can get long by crawl. Therefore, when the interval between the narrowest portions is not more than twice the thickness of the semiconductor active layer in the direction perpendicular to the interface, the coupling of the gun domain can be made stronger and higher oscillation output power can be obtained. be able to. Furthermore, creation of a device can be facilitated.
[0146]
  Also,ThirdIn the heterojunction bipolar transferred electron transistor of the embodiment, the long sides of the rectangular cathode regions are arranged adjacent to the long sides of the other cathode regions, so that the short sides are arranged adjacent to each other. Compared to the case, the coupling of the cancer domain generated in the semiconductor active layer immediately below each cathode region is strengthened, and the synchronous transmission can be easily generated.
[0147]
  Also,4thIn the heterojunction bipolar transferred electron transistor of this embodiment, the cathode regions are arranged in two rows in the longitudinal direction, so that the thermal resistance can be further reduced.
[0148]
  Also,5thIn the heterojunction bipolar transferred electron transistor of this embodiment, the impurity doping concentration in the semiconductor active layer decreases from the semiconductor base layer side to the semiconductor anode layer side. The current density distribution in the (depth direction) can be canceled out, and the time-average electric field strength in the element can be made constant to some extent. Therefore, efficiency and output power can be increased.
[0149]
  Also,6thIn the heterojunction bipolar transferred electron transistor of this embodiment, the semiconductor active layer is composed of at least two layers, and the impurity doping concentration in one of the layers decreases from the semiconductor base layer side toward the semiconductor anode layer side. As a result, efficiency and output power can be increased.
[0150]
  Also,7thIn the heterojunction bipolar transferred electron transistor of this embodiment, the decrease in impurity doping concentration in the semiconductor active layer is proportional to the distance from the semiconductor base layer side, so that the efficiency and output power are It can be further increased.
[0151]
  Also,8thIn the heterojunction bipolar transferred electron transistor of this embodiment, since a ballast layer is laminated on each cathode region, current concentration due to a planar temperature distribution in each cathode region and between each cathode region is prevented. The time average electric field strength of the entire device can be made constant. Accordingly, it is possible to prevent a decrease in oscillation intensity and to obtain a stable high output oscillation output.
[0152]
  Further, since the VCO of the second invention is configured using the heterojunction bipolar transferred electron transistor of the first invention, it is compared with the case where the single finger heterojunction bipolar transferred electron transistor is used. A large oscillation output can be obtained. Furthermore, the anode current is almost independent of the base-anode voltage. Therefore, by changing the voltage between the base and the anode, the oscillation frequency can be changed without changing the anode current. That is, according to the present invention, it is possible to realize a VCO having a good characteristic in which the oscillation output is large and the fluctuation of the oscillation output is small with respect to the oscillation frequency.
[0153]
  In addition, since the ILO of the third invention is configured by using the heterojunction bipolar transferred electron transistor of the first invention, compared with the case of using the single finger heterojunction bipolar transferred electron transistor. A larger oscillation output can be obtained. Furthermore, the input of the synchronizing signal between the base and the cathode can be amplified between the base and the anode. Therefore, the voltage and current between the base and the anode can be greatly changed by the synchronization signal having the same power as compared with the case where other gun oscillation elements are used, and the power of the injection signal can be reduced. Alternatively, it is possible to take a wider frequency pullable frequency that can be injection-synchronized than the other gun oscillation elements, depending on the intensity of a fixed injection signal.
[0154]
  Further, since the Early voltage has a very large value, the isolation from the anode to the cathode is very good, and the backflow of the oscillation signal to the cathode side can be effectively prevented. Therefore, it is possible to prevent the oscillation output from flowing backward to the synchronization signal source by grounding the base and inputting the synchronization signal from the cathode, and it is possible to prevent the phase noise from increasing.
[0155]
  In addition, the self-oscillation mixer of the fourth invention is configured by using the heterojunction bipolar transferred electron transistor of the first invention. Thus, a larger oscillation output can be obtained. Furthermore, the Early voltage has a very large value. Therefore, the isolation from the anode to the cathode is good, and the leakage of the oscillation signal to the cathode side can be eliminated. That is, according to the present invention, by isolating the base from the ground and inputting the input signal from the cathode, it is possible to easily isolate the oscillation output.
[0156]
  In addition, since the receiving device of the fifth invention is configured using the self-oscillation mixer of the fourth invention, it is possible to prevent the local oscillation frequency from leaking to the high frequency signal input. Therefore, unnecessary radiation of the local signal can be reduced, and it is possible to prevent the adjacent receiving apparatus from being damaged.
[Brief description of the drawings]
FIG. 1 is a cross-sectional view during operation of a two-finger heterojunction bipolar transferred electron device in which the distance between the cathode layers as an example of the heterojunction bipolar transferred electron device of the present invention is greater than twice the thickness of the active layer It is.
2 is a diagram showing a simulation result of output current density in the two-finger heterojunction bipolar transferred electron transistor shown in FIG. 1; FIG.
FIG. 3 is a diagram showing a simulation result of output current density of a single finger heterojunction bipolar transferred electron transistor having the same cathode finger width as that of the two finger electron transistor shown in FIG. 1;
FIG. 4 is a cross-sectional view of a heterojunction bipolar transferred electron triode.
FIG. 5 is a cross-sectional view during operation in a two-finger heterojunction bipolar transferred electron transistor in which the distance between the cathode layers is not more than twice the thickness of the active layer.
6 is a diagram showing a simulation result of output current density in the two-finger heterojunction bipolar transferred electron transistor shown in FIG.
7 is a diagram showing a planar structure of the electron triode shown in FIG.
FIG. 8 is a diagram showing a planar structure different from FIG.
FIG. 9 is a diagram showing a planar structure different from those in FIGS. 7 and 8;
FIG. 10 is a diagram showing a planar structure different from those shown in FIGS.
FIG. 11 is a diagram showing a planar structure different from those shown in FIGS.
12 is a view showing a cross-sectional structure of the two-finger heterojunction bipolar transfer shown in FIG.
13 is a view showing a cross-sectional structure different from FIG.
FIG. 14 is a cross-sectional view of a two-finger heterojunction bipolar transferred electron transistor having a ballast layer.
FIG. 15 is a diagram schematically showing a current concentration state in a heterojunction bipolar transferred electron transistor.
FIG. 16 is a diagram showing the relationship between the thickness of the ballast layer and the upper and lower limits of the oscillation current of the anode current.
17 is a circuit diagram of a VCO using the electron triode shown in FIG.
FIG. 18 is a graph showing anode-base voltage-frequency characteristics in a VCO using electron-to-device elements having different impurity concentration gradients in an active layer.
19 is a circuit diagram of the ILO using the electron triode shown in FIG.
FIG. 20 is a diagram showing a synchronization signal frequency-oscillation frequency characteristic in an ILO using an electron triode element having a different impurity concentration gradient of an active layer.
FIG. 21 is a circuit diagram of a self-oscillating mixer using the electron triode shown in FIG.
FIG. 22 is a diagram showing input / output power characteristics in a self-oscillating mixer using electron toroid elements having different impurity concentration gradients in an active layer.
23 is a diagram showing a positional relationship between a transmitter and a receiver for explaining functions of the receiver using the self-oscillation mixer shown in FIG.
24 is an explanatory diagram of a frequency band of an information signal from the transmitter shown in FIG. 23 and an up-conversion frequency band by each transmitting station.
25 is an explanatory diagram of down-conversion by each receiver shown in FIG. 23. FIG.
FIG. 26 is a diagram showing a planar structure of a conventional single finger heterojunction bipolar transferred electron triode.
FIG. 27 is a diagram showing a planar structure of an electron triode in which the width of the cathode layer in FIG. 26 is increased.
FIG. 28 is a diagram showing a planar structure of an electron triode in which the length of the cathode layer in FIG. 26 is increased.
[Explanation of symbols]
15, 25, 81, 101, 121 ... GaAs anode layer,
16, 26, 82, 103, 122 ... GaAs active layer,
17, 27, 83, 105, 123 ... GaAs base layer,
18, 28 ... 1st AlGaAs cathode layer,
19, 29 ... the second AlGaAs cathode layer,
20 ... 1st cancer domain,
21 ... The second cancer domain,
30 ... Common cancer domain,
31, 41, 51, 61, 71 ... n-type semiconductor anode layer,
32, 42, 52, 62, 72 ... p-type semiconductor base layer,
33, 43, 53, 63, 73 ... n-type semiconductor cathode layer,
34, 44, 54, 64, 74, 128 ... metal anode ohmic contact electrodes,
35, 45, 55, 65, 75, 129 ... metal-based ohmic contact electrodes,
36, 46, 56, 66, 76, 130 ... metal cathode ohmic contact electrodes,
84 ... Al_xGa_1-xAs cathode gradient layer,
85 ... Al_0.35Ga_0.65As cathode layer,
86 ... Al_0.35Ga_0.65As cathode ballast layer,
87,126 ... Al_xGa_1-xAs cathode gradient layer,
88,127 ... GaAs cathode cap layer,
89,110 ... AuGe / Ni / Au anode ohmic contact electrode,
90,111 ... Ti / Pt / Au base ohmic contact electrode,
91 ... AuGe / Ni / Au cathode ohmic contact electrode,
102 ... In_0.49Ga_0.51P etch stop layer,
104 ... GaAs active layer (N_DSlope),
106 ... In_0.49Ga_0.51P-based protective layer,
107 ... GaAs cathode layer,
108 ... In_xGa_1-xAs cathode gradient layer,
109 ... In_0.5Ga_0.5As cathode cap layer,
112 ... Ti / Pt / Au cathode ohmic contact electrode,
124 ... Al_0.3Ga_0.7As cathode layer,
125 ... Al_0.35Ga_0.65As ballast layer,
141,161,171,191,211,231 ... multi-finger
  Heterojunction Bipolar Transfer Electron Period,
142, 162, 172, 192, 212, 232 ... Base electrode,
143, 166, 175, 197, 215, 237 ... anode power supply terminal,
144,147,165,174,177,196,214,217,236,239
                                        ... inductance,
145,164,173,195,213,235 ... anode electrode,
146, 178, 218 ... cathode power supply terminal,
148,167,176,198,216,238 ... cathode electrode,
149,168,179,181,199,202,219,221,
              240, 242, 245 ... DC bias cut capacitors,
150, 169, 180, 200, 220, 241 ... output terminals,
151, 194, 234 ... capacitors,
163, 223, 248 ... λ / 4 wavelength open stubs for the fundamental oscillation frequency
170, 204, 247 ... Base power supply terminal,
182, 203, 246 ... Sync signal input terminal,
193,233 ... λ / 2 wavelength line with respect to fundamental oscillation frequency,
201: Open stub of λ / 4 wavelength with respect to synchronization signal frequency
222, 243 ... signal input terminals,
244: Open stub of λ / 2 wavelength with respect to the fundamental oscillation frequency,
251 ... 1st transmitter (transmitting station),
252 ... second transmitter (transmitting station),
253 ... first receiver,
254 ... second receiver,
255 ... Reception interference,
256 ... frequency band of transmission information signal,
257 ... Transmission information signal from the first transmitter (transmitting station) after up-conversion,
258 ... Transmission information signal from the second transmitter (transmitting station) after up-conversion,
259 ... Local frequency for the first transmitter (station),
261: Local frequency for the second transmitter (station).

Claims (14)

In a heterojunction bipolar transfer electron transistor in which a semiconductor anode layer, a semiconductor active layer, a semiconductor base layer, and a semiconductor cathode layer are sequentially laminated,
The semiconductor cathode layer is divided into a plurality of cathode regions ,
Spacing of the narrowest portion in the cathode region adjacent to each other in the extending direction of the interface between the semiconductor base layer and the semiconductor active layer, Ru der 2 times the thickness of the perpendicular direction to the interface in the semiconductor active layer Heterojunction Bipolar Transfer Electron Period In the heterojunction bipolar transfer electron transistor according to claim 1,
The plurality of cathode regions are formed adjacent to each other on a single semiconductor base layer, and are electrically connected to each other by metal electrodes. In the heterojunction bipolar transferred electron transistor according to claim 1 or 2,
A heterojunction bipolar transferred electron transistor characterized in that an interface between each cathode region and the semiconductor base layer is rectangular . In the heterojunction bipolar transfer electron transistor according to claim 3 ,
Each cathode region, as well as a rectangular shape having long and short sides, heterojunction bipolar transfer, characterized that you have been arranged to be adjacent to the long side to the long side of the other cathode region Door electron trio. In the heterojunction bipolar transferred electron according to claim 4,
Each cathode region is a heterojunction bipolar transferred de electron triode, characterized in that it is arranged in two rows in the long side direction. In the heterojunction bipolar transferred electron device according to any one of claims 1 to 5,
A heterojunction bipolar transferred electron transistor, wherein an impurity doping concentration in the semiconductor active layer decreases from the semiconductor base layer side toward the semiconductor anode layer side . In the heterojunction bipolar transferred electron transistor according to any one of claims 1 to 5 ,
The semiconductor active layer is composed of at least two layers, and an impurity doping concentration in one of the layers is reduced from the semiconductor base layer side toward the semiconductor anode layer side. Door electron trio. In heterojunction bipolar transferred de electron triode according to 請 Motomeko 6 or claim 7,
The reduction of the impurity doping concentration definitive the semiconductor active layer, heterojunction bipolar transferred de electron triode characterized that you proportional to the distance from the semiconductor base layer side. In the heterojunction bipolar transferred electron transistor according to any one of claims 1 to 8 ,
A heterojunction bipolar transferred electron transistor comprising a ballast layer laminated on each cathode region . In the heterojunction bipolar transferred electron device according to any one of claims 1 to 9,
Said semiconductor anode layer, a semiconductor active layer, a semiconductor base layer and the semiconductor cathode layer, III group element G a, I n, at least one and V elements A s of A l, P, N, at least one S b A heterojunction bipolar transfer electron transistor comprising a group III- V compound semiconductor containing two or more . A voltage-controlled frequency variable oscillator using the heterojunction bipolar transferred electron transistor according to any one of claims 1 to 9 . Claims 1 to injection locking oscillator characterized by using a heterojunction bipolar transferred de electron triode according to any one of claims 9. Self-oscillating mixer characterized by using a heterojunction bipolar transferred de electron triode according to any one of claims 1 to 9. A receiver using the self-oscillating mixer according to claim 13 .

Images