Loading...

Design and Implementation of a X Band Front-end Receiver in GaAs Technology

Sadeghabadi, Elham | 2018

1118 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 50668 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Medi, Ali
  7. Abstract:
  8. This thesis begins in an attempt to design an X band front-end receiver in GaAs pHEMT technology, suitable for microwave circuits which should be high frequency, low noise, high gain, as well as high power handling and should have stable characteristics at large number of productions. The front-end receiver consists of three blocks, namely limiter, low noise amplifier, and mixer, which are designed in ISDL group previously. The measurement result of the fabricated low noise amplifier was satisfactory, since it was close to its simulation result. However, due to the drawbacks of modeling the switching behavior of devices, the results of the two other blocks, i.e., limiter and mixer – which function based on switching diodes and transistors – contradicted their simulation results. Therefore, this thesis concentrates on these two blocks which are elaborated upon hereunder. Limiters, the first block in receiver chain, protect receiver against high power signals. In fact, when an overdriven signal enters a receiver, its limiter turns on. As result of that, the overdriven signal is returned back and cannot reach the LNA. After ceasing the overdriven signal, the limiter should return to its low loss state immediately so that the receiver can resume receiving messages. The duration taking limiter to return to its low loss state after stopping the overdriven signal is named recovery time, which heavily depends on the switching speed of limiter devices. The desirable value for recovery time is less than 100ns. Besides the low recovery time, the other most important characteristics of limiters include high power handling capability and low transmission loss at limiter's off-state. ISDL’s limiter is designed based on anti-parallel Shottky diodes. At each stage of the limiters based on anti-parallel diodes, two series of diodes inversely connected in parallel together. When the limiter turns on, its diodes turn on too. The measurement results of ISDL’s limiter show that it can handle overdriven signals with power less than 41.5dBm. Also, its recovery time is more than 1500ns at maximum input power. This limiter’s power handling is so acceptable; however, its recovery time is extremely higher than the desirable value, i.e., 100ns. The challenge against reducing recovery time is that simulating the limiter’s recovery time by factory models shows that it is less than 1ns. In fact, the simulation cannot show this phenomenon. Therefore, the model of diodes should be corrected. For doing so, the physics of Shottky diodes is studied in this thesis thoroughly. It is often said that Shottky diodes are “majority carrier” semiconductor devices. Hence, the current of diode should be stopped as soon as stopping the input voltage. However, according to the study of Shottky diode physics, the amount of minority carriers no longer remains ignorable. Indeed, minority carriers increase when the diode current increases to such an extent that the amount of them becomes comparable with that of the majority carriers. Now, the minority carriers should be depleted for turning diodes off. The duration taking diodes to deplete minority carriers is called storage time. According to the physics of Shottky diodes, the relation between the amount of minority carries and the total current of diode is obtained, which results in finding the relation between storage time and the total current of diode. Fortunately, the diagram of storage time against the current of diode was similar to the measurement diagram of the recovery time of ISDL’s limiter against the input power. Therefore, we modified the factory model of Shottky diodes based on the equation of minority carriers and its storage time against the diode current. There are some unknown constant values in these equations, which are extracted by comparing the measurement diagram with a simulation diagram of ISDL’s limiter based on the new diodes model. As a result, we completed the factory model of Shottky diodes, which can explain both the results of measurement and those of the diode physics. In the next step, two methods – that is – using anti parallel diodes and unilateral diodes were examined for designing limiters based on our diode model in an attempt to reducing the recovery time. In the first strategy, anti-parallel diodes were utilized. By optimizing the forward and reverse currents of diodes, the recovery time of this limiter was reduced to less than one-third of that of the ISDL’s limiter at 40dBm input power. In the second strategy, unilateral diodes were used; also, some grounded lines were connected in parallel to diodes in order to provide discharge paths for diodes. In this way unlike the previous one, all stages of limiter do not turn on simultaneously. Rather, the stages of limiter turn on one by one, by doing so, before the forward current of on stages reach the critical value, in which the recovery time of limiter exceeds the maximum value, i.e., 100ns, next stage turns on and reduces the current of previous stages. Therefore, the recovery time of limiter was reduced to 100ns at all input powers less than 40dBm. The superiority of this method rather the first one is due to its higher reverse currents, which result in precipitating the process of discharging diodes. Mixers are the other block in receiver chain. As previously mentioned, this block is also implemented in ISDL group. Unfortunately, its specifications were not satisfactory and compatible with the result of simulation. As the mixer and some other blocks were fabricated in a singular chip, testing the mixer was not possible separately. It should be mentioned that the drawbacks of this chain was related to the mixer, since the accuracy of the other blocks of this chain were previously proved by fabricating and testing them separately. In this thesis, two familiar types of microwave mixers, that is, single balanced active mixer and diode passive mixer were designed. These blocks are not optimized. Rather, they are simple and suitable for examining their features on measurement
  9. Keywords:
  10. Receivers ; Low Noise Amplifier (LNA) ; X Band ; Mixer ; Schottky Diodes ; Complementary Metal Oxide Semiconductor Technology (CMOS) ; X Band Front-end Receiver

 Digital Object List

 Bookmark

...see more