Off-campus UMass Amherst users: To download campus access theses, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this thesis through interlibrary loan.

Theses that have an embargo placed on them will not be available to anyone until the embargo expires.

Access Type

Open Access

Document Type


Degree Program

Electrical & Computer Engineering

Degree Type

Master of Science in Electrical and Computer Engineering (M.S.E.C.E.)

Year Degree Awarded

January 2008

Month Degree Awarded



Cascode, LNA, SiGe, Radiometer


Radiometers measures background radiation noise power of a target. The dominant quality factor of the radiometer is determined by how sensitive it is, so the lower the noise figure and the higher the gain, the more sensitive it is. It must also calibrate out any interfering noise such as sky background and system noise. Any change in gain of the radiometer receiver must also be taken into account. A Dicke radiometer compensates system gain and noise variation by switching between the target and a known noise source. To accomplish this, a single pole, double throw (SPDT) switch, switches between the receiving antenna and the noise source. The common terminal of the switch goes to the input of the low noise amplifier (LNA). This switch has a noise figure approximately equivalent to its loss and its noise is amplified by the LNA. To eliminate the loss of the switch, this paper studies a new approach of combining the switch and the LNA to become a “switchable” LNA by designing a two-stage gain block with the first stage capable of switching between the two inputs. Because the first stage is amplifying, there is no signal loss. This thesis investigates the new switching LNA and the design approach used in choosing the technology, the transistor size, biasing, extractions and matching. Two variations of the design were built using IBM’s SiGe 8HP 120nm process. The expected and measured results are compared. Results show a measured gain of 10dB and noise figure of 5dB at 19GHz. These results fall short of expectations for reasons explained in the thesis. The overall performance of this switching LNA is compared to the traditional methods. Performance criteria include gain, noise figure, isolation, matching and linearity vs. frequency and their stability vs. power and temperature variation. Power consumption, physical size and cost are also considered. The degree to which the two inputs track one another is discussed.


First Advisor

Robert W. Jackson