Telecommunications – Receiver or analog modulated signal frequency converter – Frequency modifying or conversion
Reexamination Certificate
2000-04-20
2003-10-07
Le, Thanh Cong (Department: 2684)
Telecommunications
Receiver or analog modulated signal frequency converter
Frequency modifying or conversion
C455S323000, C327S355000, C327S356000, C327S359000
Reexamination Certificate
active
06631257
ABSTRACT:
TECHNICAL FIELD
This invention relates generally to electronic signal processing circuits and more particularly to radio frequency (RF) mixers.
BACKGROUND
In the area of radio frequency (RF) receivers or transmitters, mixing circuits (mixers) generally perform frequency translation by multiplying two input signals, each of which may comprise a non-inverted/inverted signal pair. Typically, one input signal is a received signal containing information to be processed, and the other signal is typically a reference signal generated by the mixer or the receiver/transmitter circuits. Mixers have many varied applications. For example, mixers are commonly embodied in equipment used for sending and receiving AM/FM radio signals, both cable and broadcast television signals, and even in cable modems. Mixers may be used in systems designed to send or receive any type of information in an RF signal, such as voice, video, and data. The quality and performance of the mixer generally depends on achieving a closely linear operation and obtaining a controllable and predictable gain. In the prior art, the most commonly used active device mixer is known as the Gilbert cell. Other passive device mixers exist which use arrays of transformers and diodes. However, these passive device mixers are generally not amenable to integrated circuit design. Because its configuration is generally suitable for integrated circuit fabrication, the Gilbert cell has essentially become the standard “on-chip” mixer configuration.
FIG. 1
details the circuit layout of a typical prior art Gilbert cell mixer.
Mixers perform frequency translation by multiplying two frequencies together. The graphs of
FIG. 2
give a hypothetical illustration of how two signals are multiplied. We begin with the time domain representation of the two input signals f
RF
(t)
200
and f
LO
(t)
201
. The local oscillator (LO) signal is, ideally, a square wave, and the RF signal is a hypothetical modulated sinusoidal waveform. Multiplication of these two signals is simplified by transforming each one into the frequency domain and convolving the two transformed signals. Convolution is well known in the art as the method of obtaining the frequency domain representation of two signals multiplied in the time domain. It is also well known in the art to transform a time domain signal into its frequency spectrum through a Fourier Transformation. F
RF
(&ohgr;)
202
and F
LO
(&ohgr;)
203
represent the RF and LO signals in the frequency spectrum. Convolving F
RF
(&ohgr;)
202
and F
LO
(&ohgr;)
203
produces the intermediate frequency (IF) output frequency spectrum F
IF
(&ohgr;)
204
. As shown by
FIG. 2
, the output F
IF
(&ohgr;)
204
contains a scaled version of the RF spectrum centered at the sum of the RF harmonic frequency and the LO first harmonic frequency and another scaled version at the difference of the RF and LO harmonic frequencies. Thus, the product of two signals produces “sum-and-difference” results. If the mixer is an up-converter, the difference result is filtered out or suppressed while the sum result is processed further. Conversely, if the mixer is a down-converter, the sum result is filtered out or suppressed while the difference result is processed further.
To perform this frequency translation, the Gilbert cell, as shown in
FIG. 1
, typically includes a mixer core
10
, an RF input section
11
, and a biasing circuit (not shown). The mixer core
10
is made up of an LO switching interface
12
with input terminals LOP
100
, which receives the non-inverted LO input signal, and LON
101
, which receives the inverted LO input signal, and an IF output with terminals IFP
102
and IFN
103
for providing the mixer output. The LO switching interface
12
also contains two pairs of transistors, Q
3
104
and Q
6
107
, and Q
4
105
and Q
5
106
, wherein the Q
3
104
and Q
6
107
pair are each connected to input terminal LOP
100
, and the Q
4
105
and Q
5
106
pair are each connected to input terminal LON
101
. The LO switching interface
12
operates in such a way so as to quickly switch between turning Q
3
104
and Q
6
107
on while Q
4
105
and Q
5
106
are turned off, and vice versa. Therefore, at any one time, one of the transistor pairs is on and the other off.
The LO input signal which drives the switching process is ideally represented as a square wave to minimize the switching time between the two pairs of transistors. This process is normally accomplished by amplifying an LO sinusoidal signal into compression. By amplifying the signal into compression, the tops of the sinusoid are flattened out or clipped, thus approximating the attributes of a square wave.
The RF input section
11
contains input terminals RFP
108
, which receives the non-inverted RF input signal, and RFN
109
, which receives the inverted RF input signal; two transistors Q
1
110
and Q
2
111
, which are arranged as an emitter coupled differential pair; a series connection from Q
1
110
to the Q
3
104
and Q
4
105
of the mixer core
10
; another series connection from Q
2
111
to Q
5
106
and Q
6
107
of the mixer core
10
; and a pull-down current source
112
connected to the differential pair Q
1
110
and Q
2
111
. Input terminals RFP
108
and RFN
109
may also be used for receiving a biasing signal from the biasing circuit. With connections to the load resistors RLN
113
and RLP
114
through the mixer core
10
and the resistors RE
1
115
and RE
2
116
, the differential pair Q
1
110
and Q
2
111
operates as a differential amplifier. Further, the active mixer core transistors (Q
3
104
& Q
6
107
or Q
4
105
& Q
5
106
) create a cascode stage in the differential amplifier.
A cascode stage is generally well known in the art as an amplifier comprising a common emitter stage followed by a common base stage. For the cascode connection, either transistor of the differential pair may be considered a common emitter stage. As shown in
FIG. 1
, the RF transistors Q
1
110
and Q
2
111
each form a common emitter stage of the circuit. Which ever of the mixer core transistor pairs is activated at any one time, for example Q
3
104
and Q
6
107
, will typically form the common base stage. Thus, the common emitter-common base connection produces a cascode configuration. In a cascode configuration, the effective load resistance, as seen by the RF transistors, i.e., the common emitter stage, is not the value of load resistors, here RLN
113
and RLP
114
, but rather, is typically the much lower input resistance of the transistors Q
3
104
and Q
6
107
.
The addition of the cascode stage in the Gilbert cell generally helps avoid frequency response attenuation caused by Miller effect in the common emitter stage. Miller effect arises in the general situation where there is an impedance straddling the input and output terminals of an active network (assuming the impedance does not affect the voltage gain of the network). Miller's Theorem shows that this straddling impedance is equivalent to the sum of two “Miller” impedances which do not straddle the input and output terminals (i.e., one in the input circuit and one in the output circuit). The values of these input and output “Miller” impedances depend, in general, on both the value of the original straddling impedance and the network's voltage gain. In effect, a small impedance straddling the input/output terminals may be equivalent to an input impedance many times the original straddling impedance value. In the situation of a capacitance, the influence of the straddling capacitance, based on Miller effect, may have an enormous effect on the transistor's high frequency behavior.
A transistor typically has an inherent capacitance between the base and collector, which, in a common emitter configuration, results in effectively straddling the input/output terminals. Because the value of this inherent capacitance will generally be fixed depending on the particular transistor, the voltage gain of the common emitter stage will usually provide a greater influence on the value of the equivalent “Miller” i
Dawe Geoffrey C.
White William A.
Cong Le Thanh
Fulbright & Jaworski L.L.P.
Microtune ( Texas), L.P.
Nguyen Huy
LandOfFree
System and method for a mixer circuit with anti-series... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for a mixer circuit with anti-series..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for a mixer circuit with anti-series... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3135134