Performance characterization of a microwave transistor subject to the noise and matching requirements

Abstract

In this paper, the gain G T of a microwave transistor is expressed analytically in terms of the mismatchings (V in ≥ 1, V out ≥ 1) at the ports, noise figure F ≥ F min and the [z]-parameter and noise parameters. Firstly, because the input termination Z S determines the noise F ≥ F min , thus the input termination Z S is pre-determined to lie on the tangent constant noise and available gain circles so that the maximum power delivery is ensured for the given noise. Then, a design configuration is constructed in the input impedance Z in -plane covering the gain and the required input and output mismatch circles within the Unconditionally Stable Working Area for the predetermined input termination Z S . Finally, the compatible (F ≥ F min , G T , V in ≥ 1, V out ≥ 1) quadrates for either required or optimum (V in ≥ 1, V out ≥ 1) couples are obtained with their (Z S , Z L ) couples from the analysis of the design configuration. Furthermore, a case study is also presented for the full flexible performance characterization of a selected microwave transistor. It can be concluded that the near future microwave transistor is expected to be identified by performance data base built by its compatible (F ≥ F min , G T , V in ≥ 1, V out ≥ 1) quadrates and the (Z S , Z L ) terminations within the device operation domain to overview all the possible low-noise amplifier designs using the full device capacity. produce high technology transistor, which is, of course, a matter of the semiconductor technology. However, the second stage is to be able to utilize the full capacity of the high technology device that necessitates to analyze the potential performance of the device based on the linear circuit and noise theories and construct the trade-off relations simultaneously among the noise, gain, mismatch losses, bias condition V DS , I DS , and operation frequency f. Thus, a feasible design target space can be built subject to the potential performance of the active device to overview all possible designs using the full capacity of the device. Otherwise, the design targets will generally either fail to be satisfied or allow working the circuit under the potential performance of the device.However, today's conventional LNA design flow consists of trading-off based on the designer's experience and intuition among the often contrasting goals of low noise, high gain, and input and output matches in either iterative or non-iterative approach; none of which are not based on the full capacity of the device. Furthermore, series or shunt feedback is assessed to ease the noise/gain trade-off at the LNA's input, by bringing closer, in the source reflection coefficient plane, the optimum noise (Г opt ) and the maximum gain terminations at the device's input port [7][8][9][10][11][12].A different design approach that consists of evaluating the noise/gain trade-off at the transistor level is introduced by Haus and Adler [13], with a new term 'noise measure (M)' tying the two key performance components of the amplifier, then this methodology is subs...

Competitive evolutionary algorithms for building performance database of a microwave transistor

“…Performance database can be built in the 2 stages: (1) First stage is modelling the signal and noise parameters of the transistor as functions of the device operation parameters of (V DS , I DS , f). For this purpose, the artificial intelligence tools such as multilayer perceptron, support vector regression machine and generalized regression neural network can be used with the typical works given in previous studies [1][2][3] (Figure 1); (2) The final stage is to solve the highly nonlinear performance measure equations of the transistor for a predetermined design strategy for the (Source Z S , Load Z L ) terminations using either analytical [4][5][6][7][8] or numerical [9][10][11][12] methods with the Scattering (S-) and Noise (N-) parameters at the chosen operating conditions.…”

“…In this work, to build the FDTS using the numerical optimization, the objective can be considered mainly as finding out the transducer gain G T and the corresponding source Z S and load Z L terminations providing the required noise F req and mismatchings at input V inreq and output V outreq ports of the transistor subject to the physical realizability conditions given by Equations (5) to (7). This objective is expressed in the 2 different ways in the optimization process: In the first way, the same logic is followed as the analytical method in Güneş and Demirel,8 thus the optimization problem in Case I consists of the 2 steps: Since the noise figure depends solely on the source termination Z S as can be seen from the (1); the first step of Case I is to obtain the source termination Z S providing the available gain G avmax constrained by the required noise figure F (8.a). In the second step of Case I, the load termination Z L is used as an instrument in the cost function (8.b) to realize mismatching requirements at input V inreq and output V outreq ports Equations (3) to (4) provided that the Z S termination from the step 1 is used as a source impedance.…”

Full flexible performance characterization of a feedback applied transistor with LNA applications

“…In the case where the feedback applied transistor's noise figure takes place completely within the USWA, the maximum gain circle subject to the Freq ≥ Fmin will be the tangential gain circle, the equations of which are given in detail in Demirel and Güneş . The output impedance subject to the Freq ≥ Fmin and V out = 1 can be found at the tangential point of the gain and the noise circle.…”

“…To the best of our knowledge, other studies present the first instance where the compatible ( Freq ≥ Fmin, Vinreq ≥ 1, Voutreq ≥ 1, G Tmin ≤ G T ≤ G Tmax ) quadruplets are determined with their ( Z S , Z L ) terminations to solve the highly nonlinear performance equations of a microwave transistor, analytically and numerically, within the device's operation ( V DS , I DS , f ) domain, respectively, and where the trade‐off relations are obtained among the gain G T ( f ), noise F ( f ), and mismatch losses at the input V inreq ≥ 1 and output V outreq ≥ 1 . In the existing literature, a number of works have been published on LNA designs with feedback.…”

“…Since the z‐parameter domain facilitates combining the six different conditional stability configurations in the S‐parameter domain into the single configuration within either port impedance plane , here, it will also be beneficial to work out the full performance characterization in the port impedance planes (Figure ). With this, the work is presented as a flow diagram (Figure ), which can be considered in terms of the following two main stages.…”