Low Noise Amplifier Design Research Paper
Excerpt from Research Paper :
A matching device needs to be chosen taking the frequency range into consideration.
There are a few ways in which the low noise amplifier performance designed can be improved. Since GaAs Field effect transistors are used in these amplifiers, biasing point selection needs to be done carefully. The positive voltage should only be applied while the negative voltage is already on. While switching it off, the positive should always precede the negative. The arrangement should always additional inches added to allow for lumped components. They should not be intruding upon the main length. The design should be encased properly in aluminum such that no air holes creep in. Measurement of gain, VSWR and noise factor using components such as network analyzers, preamplifiers and spectrum analyzers should only be done once the LNA has reached a stable point. It is said to be stable when application of the DC voltage supply does not display any fluctuations in the Spectrum analyzer. If it does then the circuit can be regulated with varying capacitor values. The output power of the LNA can be checked using signal generators as well, which transfer RF signals into them.
Another approach of designing a low noise amplifier is by using a MESFET within the frequency range of 5-6 GHz and applying the Advance Design System (ADS). The main factors to be considered here are that the amplifier needs to have a low noise level, moderate gain and stable design. In order to set the noise to a minimum, we need to decide on the most favorable condition for a certain transistor. Information about the MESFET circuit aids in the analyzing the performance. One of the successful experiments used the GaAs MESFET NE 76000 which has a low range of noise figure and displays gain via the K-Band. Low noise requires transistors to be DC biased at an optimal operating point. It varies based on applications ( which at be of different levels of noise, gain and power ) and transistor type.
A microwave amplifier with a single stage can be designed by placing matching networks beside the transistor on each side. This alters the input and output impedance. They change over to the load and source impedance. The best way to study the gain here is to find the transducer power gain which takes both load and source mismatch into consideration. The extent to which an amplifier maintains stability or resists fluctuation is an essential guiding factor when it comes to the design. This can be evaluated using the S-parameters, terminations and matching networks. Amplifiers have the requirement of being ended at both the input and output ports for them to function in a certain way or convey the highest power to a certain load. This makes an impedance matching network to be a necessity. They can be created using mathematical calculations or in a graphical way using Smith charts. The most efficient gain can be achieved when the noise figure value is the least. This shows a gradual rise with the increase in frequency.
One particular low noise amplifier which is being widely used in radio-astronomy is the Wide Band LNA. Adding a shunt or a feedback resistor are some ways to lower the S11 (input reflection coefficient) within a broad bandwidth. This on the other hand can add on noise and alter the signal coming through. Hooking up an inductor at the source, along with an input series inductor, causes the S11 value and noise temperature to stay low. Hence most wide band low noise amplifiers are executed using balanced circuit arrangement. It has a power dissipation which is twice the regular level. An 8-20 GHz wide band LNA utilizes the intrinsic feedback method of transistors using the Cgd (gate-drain capacitor). The loading impedance of the transistor at the initial stage can be adjusted to bring down the S11 over a broad bandwidth. Transistors which have an outside inductor connected to it and a RC loading impedance can help in wide band input matching.
ADS or Agilent's advanced design system is one main tools used to design LNAs. The combination used in the 8-20 GHz wide band design mentioned in a paper by Robert Hu, has 3 transistors working in the same bias condition. The gate -- bias circuits which constitutes the transistors are designed using resistors. This is opposed to quarter wave stubs which are narrow band. It can lead to more problems with different frequency ranges. The thickness of the 1st gate --
bias resistor is lowered down to reduce its noise. Spiral inductors attached to the drain -- bias sections of the 1st two transistors tend to enhance the transistor gain due to their impedance being high. As a way to decrease the circuit's output impedance, one can use a meander line rather than spiral inductors. The noise of the transistor is assessed by allocating a temperature Td to the intrinsic drain capacitor. The other parts of the system are maintained at the original temperature. Liquid Helium cryostat can be used on low noise amplifiers designed using this procedure. It cools it down enough to be used as a sensitive cryogenic amplifier. This in turn finds applications in radio-telescope receivers.
Wireless communication techniques these days, has a significant usage of CMOS. It forms a part of the radio transceiver and works pretty well due to its ease of integration, technological scalability and low expenses. The level of responsiveness of most radio receivers these days is governed by the low noise amplifier functioning as a part of it. The LNA serves to fulfill the primary aim of noise and input match (SNIM) for a certain stage of power dispersion. LNA design can vary based on the goals. Some ways in which low noise amplifiers can be designed based on CMOS are as follows:
1. Simultaneous noise and input matching (SNIM)
2. Classical noise matching (CNM)
3. Power constrained noise optimization method (PCNO)
4. Power constrained simultaneous noise and input matching (PCSNIM)
The Classical noise matching technique (CNM) directs the LNA design towards a minimum value of noise factor. This is done using the optimal noise impedance on the amplifier. A matching circuit is placed between the amplifier source and input port. This produces LNAs with a very low NF. However a disparity between the value of the optimal impedance and the complex conjugate of the impedance at the input can result in a major difference in the gain value. The casecode arrangement in LNA design is among the well accepted ones due to its bandwidth, gain and reverse isolation having high values. One limitation of this method is that it not possible to get the input match and minimum NF at the same time.
The Simultaneous noise and input matching technique (SNIM) utilizes series feedback to achieve an efficient transfer of the optimum noise impedance to the point needed. It uses inductive source degeneration as a part of the casecode system for usage in narrow band systems. The design usually goes smoothly in cases where the transistor, power dissipation and frequency are large in magnitude. Errors tend to show up when the transistor size is small, which lowers the power dispersion. SNIM does not work well in situations where the optimal noise impedance gets higher then the input impedance of the amplifier. At a low power dissipation level, we can identify a particular size of transistor which provides the lowest value of NF while still keeping track of input matching.
The Power constrained noise optimization method (PCNO) focuses on the matching of the noise and gain due to restricted levels of power dissipation. The most efficient transistor size ( NF = minimum) can be defined using a mathematical equation. This method tends to drift towards the SNIM technique as the power dissipation augments. However both of them fail to work properly with applications requiring low levels of power which forms a part of recent technical advancements. Here is where the Power constrained simultaneous noise and input matching (PCSNIM) technique comes into play which includes another capacitor to allow SNIM at the required stage. The increased noise resistance and lowered overall cutoff frequency happen to be two of its low points.
RF CMOS has been facing certain problems in their circuit usage since they came into existence. Some of these are low values of gain frequency along with inefficient noise performance. There has been lack of a well defined process for RF CMOS design. Ultra wide band CMOS low noise amplifier design techniques have shown some advances in recent years. It utilizes a Chebyshev filter arrangement for the input matching stage and a source follower controlling the output matching. A second approach which allows matching for Ultra wide band uses a distributed amplifier. Even though it shows fairly decent performance in gain and input matching, it takes up more power in future phases and requires more space as well.
The main problem which shows up in LNA design is that the most favorable…
Sources Used in Documents:
Sahoolizadeh, Hossein., Kordalivand, Alishir M,, and Heidari, Zargham. "Design and Simulation of Low Noise Amplifier Circuit for 5 GHz to 6 GHz" World Academy of Science, Engineering and Technology (2009): 99-102 Web. 17 April. 2010.
Nguyen, Trung-Kien., Kim, Chung-Hwan., Ihm, Gook-Ju., Yang, Moon-Su., and Lee, Sang-Gug. " CMOS Low-Noise Amplifier Design Optimization Techniques" IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES 52.5(May 2004): 1433-1442 Web. 17 April. 2010.
Thai, Trang., "Low Noise Amplifier Design for Operating frequency of 4.2 GHz" Web. 18 April. 2010.
"LNA Design Using SpectreRF" cadence.com September 2004 Web. 17 April. 2010
Cite This Research Paper: