Amplifiers – With semiconductor amplifying device – Including atomic particle or radiant energy impinging on a...
Reexamination Certificate
2001-06-05
2003-05-06
Mottola, Steven J. (Department: 2817)
Amplifiers
With semiconductor amplifying device
Including atomic particle or radiant energy impinging on a...
C330S260000
Reexamination Certificate
active
06559724
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to quasi-optic arrays, such as grid arrays, and, in particular to techniques for enhancing the gain and bandwidth of active unit cells that comprise such arrays.
2. Description of Related Art
Broadband communications, radar and other imaging systems require the transmission of radio frequency (“RF”) signals in the microwave and millimeter wave bands. In order to efficiently achieve the levels of output transmission power needed for many applications at these high frequencies, a technique called “power combining” has been employed, whereby the output power of individual components are coupled, or combined, thereby creating a single power output that is greater than an individual component can supply. Conventionally, power combining has used resonant waveguide cavities or transmission-line feed networks. These approaches, however, have a number of shortcomings that become especially apparent at higher frequencies. First, conductor losses in the waveguide walls or transmission lines tend to increase with frequency, eventually limiting the combining efficiency. Second, these resonant waveguide cavities or transmission-line combiners become increasingly difficult to machine as the wavelength gets smaller. Third, in waveguide systems, each device often must be inserted and tuned manually. This is labor-intensive and only practical for a relatively small number of devices.
Several years ago, spatial power combining using “quasi-optics” was proposed as a potential solution to these problems. The theory was that an array of microwave or millimeter-wave solid state sources placed in a resonator could synchronize to the same frequency and phase, and their outputs would combine in free space, minimizing conductor losses. Furthermore, a planar array could be fabricated monolithically and at shorter wavelengths, thereby enabling potentially thousands of devices to be incorporated on a single wafer.
Since then, numerous quasi-optical devices have been developed, including detectors, multipliers, mixers, and phase shifters. These passive devices continue to be the subject of ongoing research. Over the past few years, however, active quasi-optical devices, namely oscillators and amplifiers, have evolved. One benefit of spatial power combining (over other methods) using quasi-optics is that the output power scales linearly with chip area. Thus, the field of active quasi-optics has attracted considerable attention in a short time, and the growth of the field has been explosive.
It is believed that the first quasi-optical grid array amplifier was a grid developed by M. Kim et al at the California Institute of Technology. This grid used 25 MESFET differential pairs, demonstrating a gain of 11 dB at 3 GHz. As shown in
FIG. 1
, a typical grid amplifier
10
is an array of closely-spaced differential pairs of transistors
14
on an active grid
12
sandwiched between an input and output polarizer,
18
,
24
. An input signal
16
passes through the horizontally polarized input polarizer
18
and creates an input beam incident from the left that excites rf currents on the horizontally polarized input antennas
20
of the grid
12
. These currents drive the inputs of the transistor pair
14
in the differential mode. The output currents are redirected along the grid's vertically polarized antennas
22
, producing a vertically polarized output beam
30
via an output polarizer
24
to the right.
The cross-polarized input and output affords two important advantages. First, it provides good input-output isolation, reducing the potential for spurious feedback oscillations. Second, the amplifier's input and output circuits can be independently tuned using metal-strip polarizers, which also confine the beam to the forward direction. Numerous grid amplifiers have since been developed and have proven thus far to have great promise for both military and commercial RF applications and particularly for high frequency, broadband systems that require significant output power levels (e.g.>5 watts) in a small, preferably monolithic, package. Moreover, a resonator can be used to provide feedback to couple the active devices to form a high power oscillator.
Unfortunately, conventional active grids arrays, such as amplifiers and oscillators have not been as efficient as is desirable. In particular, reported grid array amplifiers using simple differential pair unit cells exhibit only relatively limited gain, on the order of 10 dB or less. The limited gain limits the applications to which conventional grid arrays may be employed. Moreover, in addition to gain, frequency response and impedance matching are all critical criteria for the design of microwave and millimeter wave devices. The current state of quasi-optic amplifier design does not adequately address these issues.
There is thus a definite need for active quasi-optic grid arrays, and particularly the unit cells that comprise the arrays, that yield higher gains, at higher frequencies. It would be further desirable to have such components that offer greater flexibility in impedance matching, thereby improving the bandwidth and manufacturability of such designs.
SUMMARY OF THE INVENTION
The present invention, which addresses these needs, resides in an architecture for improving the gain and bandwidth of active quasi-optic grid array unit cells. A method of the invention includes providing a two active networks and applying reinforcing signals to each of the networks. The first active network is driven by an input signal of a given magnitude and polarity and the second active network is driven by an input signal that is equal and opposite to the input signal that drives the first network. The first network includes a signal input port, a signal output port, a feedback tie-in port and a feedback take-off port. Similarly, the second network includes a signal input port, a signal output port, a feedback tie-in port, and a feedback take-off port. The method then applies to the feedback tie-in port of the first active network, via a feedback path, a reinforcing signal derived from the feedback take-off port on one of the active networks of the unit cell, and applies to the feedback tie-in port of the second active network, via a feedback path, a reinforcing signal derived from the feedback take-off port on the other one of the active networks of the unit cell. Each of the feedback paths includes a substantially identical feedback network having a transfer function that causes the reinforcing signal applied to each network to add constructively to the input signal applied to that network within the frequency range of interest.
In one aspect of the invention, the reinforcing signal applied to the feedback tie-in port of the first network is derived from the feedback take-off port of the second network, and the reinforcing signal applied to the feedback tie-in port of the second network is derived from the feedback take-off port of the first network. This may be referred to as a cross-coupled, regenerative feedback topology.
In a specific implementation of this aspect, the feedback tie-in port of each network is internally connected to the signal input port of that network and the feedback take-off port of each network is internally connected to the signal output port of that network. This embodiment includes a simple differential pair of active device connected using a crossed-coupled, regenerative feedback topology.
In an alternative aspect of the invention, the reinforcing signal applied to the feedback tie-in port of the first network is derived from the feedback take-off port of the first network and the reinforcing signal applied to the feedback tie-in port of the second network is derived from the feedback take-off port of the second network (broad shunt—shunt config.) In this embodiment, the feedback path of each network includes a substantially identical feedback network and the reinforcing signal applied to the feedback tie-in port is derived via a combination of a frequency
Cheung Chun-Tung
Deckman Blythe C.
DeLisio, Jr. Michael P.
Rosenberg James J.
Rutledge David B.
California Institute of Technology
Fish & Richardson P.C.
Mottola Steven J.
LandOfFree
Techniques for enhancing gain in a quasi-optic grid array does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Techniques for enhancing gain in a quasi-optic grid array, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Techniques for enhancing gain in a quasi-optic grid array will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3011607