Quadrature mixer circuit including three-input local mixers

Abstract

In a quadrature mixer circuit for receiving a radio frequency signal to generate first and second quadrature output signals, a first three-input mixer receives the radio frequency signal, a first local signal having a first frequency and a second local signal having a second frequency to generate the first quadrature output signal, and a second three-input mixer receives the radio frequency signal, the first local signal and the second local signal to generate the second quadrature output signal. The second local signal received by the first three-input mixer and the second local signal received by the second three-input mixer being out of phase by π/2 from each other.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to a quadrature mixer circuit applied to a wireless receiver, and more particularly, to a quadrature mixer circuit applied to a zero intermediate frequency (IF) type (a so-called direct conversion type) wireless receiver or a low IF type wireless receiver.

2. Description of the Related Art

While super heterodyne type wireless receivers have excellent noise figure (NF) characteristics, the super heterodyne type wireless receivers have a large number of components including local oscillators, image removing filters and an IF band-pass filter, which is an obstacle for incorporating a radio frequency (RF) portion and a baseband portion into one chip.

In order to decrease the number of components, various direct conversion type wireless receivers have been developed. In this case, the improvement of quadrature mixer circuits applied to such direct conversion type wireless receivers is indispensable.

In a first prior art quadrature mixer circuit using two-input mixers (see: FIG. 29 of JP-A-9-205382), since a local oscillator signal has the same frequency as that of a radio frequency (RF) signal, a DC offset cannot be completely removed, which requires DC offset removing circuits. Also, trouble in reception sensitivity may be generated. Further, the reception sensitivity of other wireless receivers may be suppressed. This will be explained later in detail.

Even in a second prior art quadrature mixer circuit using two-input mixers and a local oscillator signal having a half frequency of the RF signal, the same disadvantages as the first prior art quadrature mixer circuit exist. This also will be explained later in detail.

In a third prior art quadrature mixer circuit (see: JP-A-9-205382 & Takafumi Yamaji et al, “An I/Q Active Balanced Harmonic Mixer with IM2 Cancelers and a 45° Phase Shifter”, IEEE Journal of Solid-State Circuits, Vol. 33, No. 12, pp. 2240–2246, December 1998), two-input even-ordered harmonic mixers, a voltage controlled oscillator having a frequency different from the frequency of the RF signal and a π/4 phase shifter are provided. In the third prior art quadrature mixer circuit, however, it is difficult to realize the π/4 phase shifter. This also will be explained later in detail.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide a quadrature mixer circuit capable of decreasing the DC offset, suppressing the reception trouble without a π/4 phase shifter and decreasing the number of components such as IF filters and second filters.

According to the present invention, in a quadrature mixer circuit for receiving an RF signal to generate first and second quadrature output signals, a first three-input mixer receives the RF signal, a first local oscillator signal having a first frequency and a second local oscillator signal having a second frequency to generate the first quadrature output signal, and a second three-input mixer receives the RF signal, the first local oscillator signal and the second local oscillator signal to generate the second quadrature output signal. The second local oscillator signal received by the first three-input mixer and the second local oscillator signal received by the second three-input mixer are out of phase by π/2 from each other.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be more clearly understood from the description set forth below, as compared with the prior art, with reference to the accompanying drawings, wherein;

FIG. 1 is a block circuit diagram illustrating a first prior art quadrature mixer circuit applied to a direct conversion type wireless receiver;

FIG. 2 is a circuit diagram for explaining a down-conversion system of the direct conversion type wireless receiver of FIG. 1;

FIG. 3 is a circuit diagram for explaining a DC offset by the self-mixing phenomenon in FIG. 1;

FIG. 4 is a circuit diagram illustrating a second prior art quadrature mixer circuit applied to a direct conversion type wireless receiver;

FIG. 5 is a circuit diagram illustrating a third prior art quadrature mixer circuit applied to a direct conversion type wireless receiver;

FIG. 6 is a block circuit diagram illustrating a first embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention;

FIG. 7 is a block circuit diagram illustrating a second embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention;

FIG. 8A is a detailed circuit diagram of the post stage of the Johnson counter of FIG. 7;

FIG. 8B is a waveform diagram showing the input and output signals of the post stage of the Johnson counter of FIG. 8A;

FIG. 9 is a circuit diagram of a doubly-polarity switching mixer used as each of the three-input mixers of FIGS. 6 and 7;

FIG. 10A is a detailed circuit diagram of the doubly-polarity switching mixer of FIG. 9;

FIG. 10B is a detailed circuit diagram of the linear differential circuit of FIG. 10A;

FIG. 11A is another detailed circuit diagram of the doubly-polarity switching mixer of FIG. 9;

FIG. 11B is a detailed circuit diagram of the linear differential circuit of FIG. 11A;

FIG. 12 is a circuit diagram illustrating a modification of the doubly-polarity switching mixer of FIG. 10A;

FIG. 13 is a circuit diagram illustrating a modification of the doubly-polarity switching mixer of FIG. 11A;

FIG. 14A is a circuit diagram illustrating a modification of the doubly-polarity switching mixer of FIG. 12;

FIG. 14B is a detailed circuit diagram of the linear differential circuit of FIG. 14A;

FIG. 15A is a circuit diagram illustrating a modification of the doubly-polarity switching mixer of FIG. 13;

FIG. 15B is a detailed circuit diagram of the linear differential circuit of FIG. 15A;

FIG. 16 is a block circuit diagram illustrating a third embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention;

FIG. 17 is a block circuit diagram illustrating a fourth embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention; and

FIGS. 18 and 19 are tables showing the frequency division numbers of the frequency dividers of FIG. 17.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Before the description of the preferred embodiments, prior art quadrature phase shift circuits applied to a direct conversion type wireless receiver will be explained with reference to FIGS. 1, 2, 3, 4 and 5.

In FIG. 1, which illustrates a first prior art quadrature phase shift circuit applied to a direct conversion type wireless receiver (see: FIG. 29 of JP-A-9-205382), an antenna 1 is connected via a low noise amplifier 2 to a quadrature mixer circuit 3 which is formed by two-input mixers 31 and 32, a voltage controlled oscillator 33 and a π/2 phase shifter 34.

Also, the low frequency components of the output signals of the two-input mixers 31 and 32 pass through low-pass filters 4 and 5, respectively, and then, the gains of the output signals of the low-pass filters 4 and 5 are controlled by automatic gain control (AGC) amplifiers 6 and 7, respectively. Thus, baseband components I and Q are obtained.

Further, the baseband components I and Q are subjected to analog-to-digital conversion by analog/digital (A/D) converters 8 and 9, and then are supplied to a digital signal processor (DSP) 10 serving as a demodulator.

The principle of the two-input mixers 31 and 32 is explained below.

In the two-input mixers 31 and 32, a second-order term of the transfer characteristics of a non-linear element is used. In the non-linear element, an input u and an output f(u) are represented byf(u)=a₀+a₁u+a₂u²+ . . . +a_nuⁿ+  (1)

When an RF signal u_RF having a frequency f_RF and a local oscillator signal u_Lo having a frequency f_Lo are mixed at the non-linear element, the second term a_zu^z of the formula (1) is represented bya_zu²=a₂(u_RF+u_LO)²=a₂(u_RF²+u_LO²+2u_RFu_LO)  (2)

In this case,u_RF=U_RF·cos(2πf_RFt)  (3)u_LO=U_LO·cos(2πf_LOt)  (4)

Therefore, the third term of the formula (2) is represented by2U_RF·u_LO=2U_RF·U_LO·cos(2πf_RFt)·cos(2πf_LOt)=U_RF·U_LO·[cos{2π(f_RF−f_LO)t}+cos{2π(f_RF+f_LO)t}]  (5)

In such a down-conversion system as illustrated in FIG. 1 where the two-input mixer 32 is followed by the low-pass filter 5, as shown in FIG. 2, only the low frequency component U_RF·U_LO·cos{2π(f_RF−f_LO)t} in the formula (5) is obtained. Note that a difference frequency |f_RF−f_LO| is defined as an intermediate frequency f_IF.

If an ideal mixer consisting of a multiplier has the same conversion gain in both frequencies f_RF and f_RF−f_LO, the in-band noise power converted from two bands around the both frequencies f_RF and f_RF−f_LO is twice (=1/(1²+1²)) from the formula (5). Therefore, the noise figure (NF) of the ideal mixer becomes damaged by two, i.e., 3.01 dB, as compared with a circuit with a single-input signal and a single-output signal.

On the other hand, since the local oscillator signal u_LO is supplied via the π/2 phase shifter 34 to the two-input mixer 31, a local oscillator signal U_LO′ supplied to the two-input mixer 31 is represented byu_LO′=U_LO·sin(2πf_LOt)  (6)

Therefore, the third term of the formula (2) is represented by2U_RF·u_LO′=2U_RF·U_LO·cos(2πf_RFt)·sin(2πf_LOt)=U_RF·U_LO·[−sin{2π(f_RF−f_LO)t}+sin{2π(f_RF+f_LO)t}]  (7)

Thus, in a down-conversion system as illustrated in FIG. 1 where the two-input mixer 31 is followed by the low-pass filter 4, the component, U_RF·U_LO·sin{2π(f_RF−f_LO)t} is obtained. Note that a difference |f_RF−f_LO| is also defined as an intermediate frequency f_IF.

In the direct conversion type wireless receiver of FIG. 1, f_RF=f_LO, so that there is no intermediate frequency f_IF(=0).

The direct conversion type wireless receiver of FIG. 1 has the following advantages.

{circle around (1)} Since there is no intermediate frequency f_IF(=0), no suppression of image is necessary, that is, no image removing filters are necessary. Also, there are few sources for generating spurious waves. Note that the low-pass filters 4 and 5 are provided in a baseband portion, and therefore, the low-pass filters 4 and 5 are easily incorporated into an integrated circuit.

{circle around (2)} At The baseband portion as well as an RF portion including the low noise amplifier 2 and the quadrature mixer circuit 3 can be easily incorporated into one chip.

Thus, the direct conversion type wireless receiver of FIG. 1 can be easily incorporated into one chip, which would decrease the size and decrease the manufacturing cost.

On the other hand, the direct conversion type wireless receiver of FIG. 1 has the following disadvantages.

{circle around (1)} A DC offset may be generated. That is, a slight difference between the frequency f_RF of the RF signal u_RF and the frequency f_LO of the local oscillator signal u_LO appear as a DC offset at the output signal of each of the two-input mixers 31 and 32. Also, as illustrated in FIG. 3, the local oscillator signal u_LO leaks from one input of the two-input mixer 32 (31) to the other input of the two-input mixer 32 (31). As a result, the leaked local oscillator signal is reflected at the output of the low noise amplifier 2 and at the antenna 1 as indicated by X1 and X2, respectively. The reflected local oscillator signals X1 and X2 are down-converted and result in a DC offset which is called a DC offset caused by the self-mixing phenomenon. Since the above-mentioned DC offset is not stable, the DC offset serves as a low frequency noise. In order to remove the DC offset, DC offset removing circuits are required, which is an obstacle in realizing one chip wireless receiver.

{circle around (2)} Ad Since the frequency f_LO of the local oscillator signal u_LO coincides with the frequency f_RF of the RF signal, the leakage of the local oscillator signal u_LO within the wireless receiver of FIG. 1 generates trouble in reception sensitivity around the frequency f_RF.

{circle around (3)} Antenna radiation of the local oscillator signal u_LO suppresses the reception sensitivity of other wireless receivers receiving RF signals using approximately the same frequency f_LO.

In FIG. 4, which illustrates a second prior art quadrature mixer circuit applied to a direct conversion type wireless receiver, a ¼-frequency divider 35, a two-input mixer 36 and a band-pass filter 37 are added to the elements of FIG. 1. That is, if the local oscillator signal u_LO of the voltage controlled oscillator 33 is f_LO, a local oscillator signal having a frequency (¾) f_LO is generated by the ¼-frequency divider 35, the two-input mixer 36 and the band-pass filter 37. Therefore,f_RF=(¾)f_LO

As a result, the frequency f_LO of the voltage controlled oscillator 33 is different from the frequency f_RF of the RF signal.

Even in the direct conversion type wireless receiver of FIG. 4, since the frequency (¾) f_LO at an input of the two-input mixer 31(32) is the same as that at the frequency f_RF at another input of the two-input mixer 31(32), there are the same disadvantages as in the direct conversion type wireless receiver of FIG. 1.

In FIG. 5, which illustrates a third prior art quadrature mixer circuit applied to a direct conversion type wireless receiver (see: JP-A-9-205382) & Takafumi Yamaji et al, “An I/Q Active Balanced Harmonic Mixer with IM2 Cancelers and a 45° Phase Shifter”, IEEE Journal of Solid-State Circuits, Vol. 33, No. 12, pp. 2240–2246, December 1998), the quadrature mixer circuit 3 is constructed by two-input even-harmonic mixers 31′ and 32′, a voltage controlled oscillator 33′ and a π/4 phase shifter 34′.

The principle of the two-input even-harmonic mixers 31′ and 32′ is explained below.

In the two-input even-harmonic mixers 31′ and 32′, a third-order term of the transfer characteristics of a non-linear element is used.

When an RF signal u_RF having a frequency f_RF and a local oscillator signal u_LO having a frequency f_LO are mixed at the non-linear element, the third-order term a₃u³ of the formula (1) is represented bya₃u³=a₃(u_RF+u_LO)³=a₃(u_RF³+u_LO³+3u_RF²u_LO+3u_RFu_LO²)  (8)

The fourth term of the formula (8) is represented by3u_RFu_LO²=3U_RF·U_LO²·cos(2πf_RFt)cos²(2πf_LOt)=3U_RF·U_LO²·cos(2πf_RFt){1+cos(4πf_LOt)}/2=3U_RF·U_LO²·{cos(2πf_RFt)+cos(2πf_RFt)cos(4πf_LOt)}/2=3U_RF·U_LO²[2 cos(2πf_RFt)+cos{2π(f_RFt−2f_LO)t}+cos{2π(f_RFt+2f_LO)t}]/4  (9)

In such a down-conversion system of FIG. 5 where the two-input even-harmonic mixer 32′ is followed by the low-pass filter 5, only the low frequency component 3U_RF·U_LO² cos{2π(f_RFt−2f_LO)t} in the formula (9) is obtained. Note that a difference |f_RF−2f_LO| is defined as an intermediate frequency f_IF.

If an ideal even-harmonic mixer has the same conversion gain in both frequencies f_RF, f_RF−2f_LO and f_RF+2f_LO, the in-band noise power converted from the three bands around the frequencies f_RF, f_RF−2f_LO and f_RF+2f_LO is six-times (=1/(2²+1²+1²)) from the formula (9) by squaring each amplitude thereof. Therefore, the noise figure (NF) of the ideal even-harmonic mixer becomes damaged by six, i.e., 7.78 dB, as compared with a circuit with a single-input signal and a single-output signal.

On the other hand, since the local oscillator signal u_LO is supplied via the π/4 phase shifter 34′ to the two-input even-harmonic mixer 31′, a local oscillator signal u_LO′ supplied to the two-input even-harmonic mixer 31′ is represented byu_LO′=u_LO·cos(2πf_LOt+π/4)  (10)

Therefore, the fourth-order term of the formula (8) is represented by3u_RFu_LO′²=3U_RF·U_LO²·cos(2πf_RFt)cos²(2πf_LOt+π/4)=3U_RF·U_LO²·cos(2πf_RFt){1+cos(4πf_LOt+π/2)}/2=3U_RF·U_LO²·cos(2πf_RFt){1−sin(4πf_LOt)}/2=3U_RF·U_LO²·{cos(2πf_RFt)−cos(2πf_RFt)sin(4πf_LOt)}/2=3U_RF·U_LO²·[2 cos(2πf_RFt)+sin{2π(f_RFt−2f_LO)t}+sin{2π(f_RFt+2f_LO)t}]/4  (11)

Thus, in a down-conversion system of FIG. 5 where the two-input even-harmonic mixer 31′ is followed by the low-pass filter 4, the component 3U_RF·U_LO²·sin{2π(f_RFt−2f_LO)t} of the formula (14) is obtained. Note that a difference |f_RFt−2f_LO| is also defined as an intermediate frequency f_IF.

In the quadrature mixer circuit 3 of FIG. 5, however, as stated in JP-A-9-205382, it is difficult to realize the π/4 phase shifter 34′.

In FIG. 6, which illustrates a first embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention, the two-input mixers 31 and 32 of FIG. 1 are replaced by three-input mixers 31″ and 32″, respectively, and the voltage controlled oscillator 33 of FIG. 1 is replaced by two voltage controlled oscillators 33″A and 33″B, which generate local oscillator signals u_LO1, and u_LO2 having frequencies f_LO1 and f_LO2, respectively. Note that each of the three-input mixers 31″ and 32″ can be constructed by a three-input multiplier (see: JP-A-10-105632).

In the three-input mixers 31″ and 32″, a third-order term of the transfer characteristics of a non-linear element is also used.

When the RF signal U_RF having the frequency f_RF and the local oscillator signals u_LO1 and u_LO2 having frequencies f_LO1 and f_LO2 are mixed at the non-linear element, the third-order term a₃u³ of the formula (1) is replaced bya₃u³=a₃(u_RF+u_LO1+u_LO2)³=a₃(u_RF³+u_LO1³+u_LO2³+3u_RF²u_LO1+3u_RF²u_LO2+3u_RFu_LO1²+3u_LO1²u_LO2+3u_RFu_LO2²+3u_LO1u_LO2²+6u_RFu_LO1u_LO2)  (13)

In this case, the RF signal u_RF and the local oscillator signals u_LO1 and u_LO2 are represented byu_RF=u_RF·cos(2πf_RFt)  (14)u_LO1=u_LO1·cos(2πf_LO1t)  (15)u_LO2=u_LO2·cos(2πf_LO2t)  (16)

Generally, the following triple product of trigonometric functions is known;cos α cos β cos γ={cos(α+β−γ)+cos(β+γ−α)+cos(γ+α−β)+cos(α+β+γ)}/4  (17)

Therefore, the tenth term of the formula (13) is represented by6u_RFu_LO1u_LO2=6U_RF·U_LO1·U_LO2·cos(2πf_RFt)cos(2πf_LO1t)cos(2πf_LO2t)=3U_RF·U_LO1·U_LO2·[cos{2π(f_RF+f_LO1−f_LO2)t}+cos{2π(−f_RF+f_LO1+f_LO2)t}+cos{2π(f_RF−f_LO1+f_LO2)t}+cos{2π(f_RF+f_LO1+f_LO2)t}]/2  (18)

In such a down-conversion system of FIG. 6 where the three-input mixer 32″ is followed by the low-pass filter 5, only the low frequency component 3U_RF·U_LO1·U_LO2 cos{2π(−f_RF+f_LO1+f_LO2)t} in the equation (18) is obtained. Note that a difference |f_RF−f_LO1−f_LO2| is defined as an intermediate frequency f_RF.

On the other hand, since the local oscillator signal u_LO2 is supplied via the π/2 phase shifter 34 to the three-input mixer 31″, a local oscillator signal u_LO2′ supplied to the three-input mixer 31″ is represented byu_LO2′=U_LO2 sin(2πf_LO2t)  (19)

Generally, the following triple product of trigonometric functions is known:sin α cos β cos γ={sin(α+β−γ)+sin(β+γ−α)+sin(γ+α−β)−sin(α+β+γ)}/4  (20)

Therefore, the tenth term of the formula (13) is represented by6u_RFu_LO1u_LO2=6U_RF·U_LO1·U_LO2·cos(2πf_RFt)cos(2πf_LO1t)sin(2πf_LO2t)=3U_RF·U_LO1·U_LO2·[sin{2π(f_RF+f_LO1−f_LO2)t}+sin{2π(−f_RF+f_LO1+f_LO2)t}+sin{2π(f_RF−f_LO1+f_LO2)t}+sin{2π(f_RF+f_LO1+f_LO2)t}]/2  (21)

Thus, in a down-conversion system of FIG. 6 where the three-input mixer 31″ is followed by the low-pass filter 4, the component 3U_RF·U_LO1·U_LO2·sin{2π(f_RF+f_LO1+f_LO2)t} of the formula (21) is obtained. Note that a difference |f_RF−f_LO1−f_LO2| is also defined as an intermediate frequency f_IF.

In FIG. 7, which illustrates a second embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention, the voltage controlled oscillators 33″A and 33″B and the π/2 phase shifter 34 of FIG. 6 are replaced by a voltage controlled oscillator 33″ and a Johnson counter formed by two ½-frequency dividers 71 and 72. In this case, the Johnson counter generates local oscillator signals u_LO/4 and u_LO/4′ out of phase by π/2.

In the three-input mixers 31″ and 32″, sin{2π(f_RF−f_LO1−f_LO2)t} of the formula (21) is also used.

When the RF signal U_RF having the frequency f_RF and the local oscillator signals u_LO and u_LO/4 having frequencies f_LO and f_LO/4 are mixed at the non-linear element, the third-order term a₃u³ of the formula (1) is replaced bya₃u³=a₃(u_RF+u_LO+u_LO/4)³=a₃(u_RF³+u_LO³+u_LO/4³+3u_RF²u_LO+3u_RF²u_LO/4+3u_RFu_LO²+3u_LO²u_LO/4+3u_RFu_LO/4²+3u_LOu_LO/4²+6u_RFu_LOu_LO/4)  (22)

In this case, the local oscillator signal u_LO/4 is represented byu_LO/4=u_LO/4·cos{2π(f_LO/4)t}  (23)

Therefore, the tenth term of the formula (22) is represented by6u_RFu_LOu_LO/4=6U_RF·U_LO·U_LO/4·cos(2πf_RFt)cos(2πf_LOt)cos{2π(f_LO/4)t}=3U_RF·U_LO·U_LO/4·[cos{2π(f^RF−5f_LO/4)t}+cos{2π(f_RF−3f_LO/4)t}+cos{2π(f_RF+5f_LO/4)t}+cos{2π(f_RF+3f_LO/4)t}]/2  (24)

In such a down-conversion system of FIG. 7 where the three-input mixer 32″ is followed by the low-pass filter 5, only the low frequency component 3U_RF·U_LO·U_LO/4·cos{2π(−f_RF+5f_LO/4)t} in the equation (24) is obtained. Note that a difference |f_RF−5f_LO/4| is defined as an intermediate frequency f_RF.

On the other hand, the local oscillator signal u_LO/4′ supplied to the three-input mixer 31″ is represented byu_LO4′U_LO/4·sin{2π(f_LO/4)t}  (25)

Therefore, the tenth term of the formula (24) is represented by6u_RFu_LOu_LO/4=6U_RF·U_LO·U_LO/4·cos(2πf_RFt)cos(2πf_LOt)sin{2π(f_LO/4)t}=3U_RF·U_LO·U_LO/4·[sin{2π(f_RF−5f_LO/4)t}+sin{2π(f_RF−3f_LO/4)t}+sin{2π(f_RF+5f_LO/4)t}+sin{2π(f_RF+3f_LO/4)t}]/2  (26)

Thus, in a down-conversion system of FIG. 7 where the three-input mixer 31″ is followed by the low-pass filter 4, the component 3U_RF·U_LO1·U_LO2·sin{2π(f_RF−5f_LO/4)t} of the formula (26) is obtained. Note that a difference |f_RF−5f_LO/4| is also defined as an intermediate frequency f_IF.

FIG. 8A is a detailed circuit diagram of the post stage of the Johnson counter of FIG. 7 and FIG. 8B is a waveform diagram showing the input and output signals of the post stage of the Johnson counter of FIG. 8A. That is, the post stage counter 72 is constructed by D-type flip-flops connected in series, so that the local oscillator signals U_LO/4 and U_LO/4′ having the same frequencies f_LO/4(I) and f_LO/4(Q) are obtained from a signal having a frequency f_LO/2.

The three-input mixers 31″ and 32″ can be constructed by doubly-polarity switching mixers instead of the three-input multipliers. Doubly-polarity switching mixers other than the three-input multipliers are easily integrated into one chip.

A typical doubly-polarity switching mixer will be explained next with reference to FIGS. 9, 10A, 10B, 11A, 11B, 12, 13, 14A, 14B, 15A and 15B.

In FIG. 9, which is a circuit diagram of the doubly-polarity switching mixer, an RF signal V_RF(t) is switched by two rectangular local oscillator signals S₁(t) and S₂(t). That is, switches 91, 92, 93 and 94 controlled by the local oscillator signals S₁(t) and S₂(t) receive the RF signal V_RF(t) to generate an intermediate frequency signal V_IF. If the local oscillator signal S₁(t) has an amplitude of ±U_LO1 and a frequency of f_LO1 and the local oscillator signal S₂(t) has an amplitude of ±U_LO2 and a frequency of f_LO2, S₁(t) and S₂(t) are represented byS₁(t)=U_LO1(4/π)[cos(2πf_LO1t)−(⅓)cos(6πf_LO1t)+(⅕)cos(10πf_LO1t)−( 1/7)cos(14πf_LO1t)+ . . . ]  (27)S₂(t)=U_LO2(4/π)[cos(2πf_LO2t)−(⅓)cos(6πf_LO2t)+(⅕)cos(10πf_LO2t)−( 1/7)cos(14πf_LO2t)+ . . . ]  (28)

Also, if S₁(t)=U_LO1sgn(S₁(t)) and S₂(t)=U_LO2sgn(S₂(t)),V_IF(t)=u_RF·U_LO1sgn(S₁(t)U_LO2sgn(S₂(t))  (29)u_RF=U_RF cos(2πf_RFt)  (30)Therefore,V_IF(t)=u_RF·U_LO1sgn(S₁(t))U_LO2sgn(S₂(t))=u_RF cos(2πf_RFt)U_LO1sgn(S₁(t))U_LO2sgn(S₂(t))  (31)

Since sgn(S₁(t)) and sgn(S₂(t)) are −1 and +1, and −1 and +1, respectively, their absolute values are represented by|sgn(S₁(t))|=1  (32)|sgn(S₂(t))|=1  (33)

Therefore, sgn(S₁(t)) and sgn(S₂(t)) are represented bysgn(S₁(t))=(4/π)[cos(2πf_LO1t)−(⅓)cos(6πf_LO1t)+( 5/1)cos(10πf_LO1t)−( 1/7)cos(14πf_LO1t)+ . . . ]  (34)sgn(S₂(t))=(4/π)[cos(2πf_LO2t)−(⅓)cos(6πf_LO2t)+( 5/1)cos(10πf_LO2t)−( 1/7)cos(14πf_LO2t)+ . . . ]  (35)

Thus, the formula (31) is represented by

$\begin{array}{cc}\begin{array}{c}{V}_{\mathrm{IF}}\left(t\right)=u_{\mathrm{RF}}·U_{\mathrm{L01}}\mathrm{sgn}\left(S_1\left(t\right)\right)·U_{\mathrm{L02}}\mathrm{sgn}\left(S_2\left(t\right)\right)\\ =U_{\mathrm{RF}}·U_{\mathrm{L01}}·U_{\mathrm{L02}}·\cos \left(2\pi f_{\mathrm{RF}}t\right)·\mathrm{sgn}\left(S_1\left(t\right)\right)·\mathrm{sgn}\left(S_2\left(t\right)\right)\\ =\left(16/{\pi }^2\right)U_{\mathrm{RF}}·U_{\mathrm{L01}}·U_{\mathrm{L02}}·\cos \left(2\pi f_{\mathrm{RF}}t\right)[\cos \left(2\pi f_{\mathrm{L01}}t\right)-\\ \left(1/3\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)+\left(1/5\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)-\\ \left(1/7\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)+\cdots ]·[\cos \left(2\pi f_{\mathrm{L02}}t\right)-\\ \left(1/3\right)\cos \left(6\pi f_{\mathrm{L02}}t\right)+\left(1/5\right)\cos \left(10\pi f_{\mathrm{L02}}t\right)-\\ \left(1/7\right)\cos \left(14\pi f_{\mathrm{L02}}t\right)+\cdots ]\\ =\left(16/{\pi }^2\right)U_{\mathrm{RF}}·U_{\mathrm{L01}}·U_{\mathrm{L02}}·\\ [\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\cos \left(2\pi f_{\mathrm{L02}}t\right)+\\ \left(1/9\right)\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\cos \left(6\pi f_{\mathrm{L02}}t\right)+\\ \left(1/25\right)\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\cos \left(10\pi f_{\mathrm{L02}}t\right)+\\ \left(1/49\right)\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\cos \left(14\pi f_{\mathrm{L02}}t\right)-\\ \left(1/3\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\cos \left(2\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\cos \left(6\pi f_{\mathrm{L02}}t\right)\}+\\ \left(1/5\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\cos \left(2\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\cos \left(10\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/7\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\cos \left(2\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\cos \left(14\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/15\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\cos \left(10\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\cos \left(6\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/21\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\cos \left(14\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\cos \left(6\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/35\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\cos \left(14\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\cos \left(10\pi f_{\mathrm{L02}}t\right)\}+\cdots ]\end{array}& \left(36\right)\end{array}$

Apparent from the formula (36), when the RF signal U_RF is switched by the polarities of the local oscillator signals S₁(t) and S₂(t), basic waves and harmonic waves of the three signals are obtained.

For example, four frequency components such as cos{2π(f_RF−f_LO1−f_LO2)t}, cos{2π(f_RF−f_LO1+f_LO2)t}, cos{2π(f_RF+f_LO1+f_LO2)t} and cos{2π(f_RF+f_LO1−fLO2)t} are obtained from the triple product cos(2πf_RFt)cos(2πf_LO1t) cos(2πf_LO2t) of basic waves with reference to the formula of triple product of trigonometric functions shown in the formula (17).

In this case, if the doubly-polarity switching mixer of FIG. 9 constitutes a down-conversion system where a low-pass filter follows this doubly-polarity switching mixer, only the low frequency component cos{2π(f_RF−f_LO1−f_LO2)t} is obtained. Note that a difference |f_LO1−f_LO2| is defined as an intermediate frequency f_IF.

Also, four frequency components such as cos[2π{f_RF−(2m+1)f_LO1−(2m′+1)f_LO2}t], cos[2π{f_RF−(2m+1)f_LO1+(2m′+1)f_LO2}t], cos[2π{f_RF+(2m+1)f_LO1+(2m′+1)f_LO2}t] and cos[2π{f_RF+(2m+1)f_LO1−(2m′+1)f_LO2}t] are obtained from triple products cos(2πf_RFt) cos{2π(2m+1)f_LO1t}cos{2π(2m′+1)f_LO2t} (m, m′=1, 2, . . . ) of the RF signal V_RF(t) and the odd-higher harmonic waves of the local oscillator signals S₁(t) and S₂(t) with reference to the formula of triple product of trigonometric functions shown in the formula (17). In this case, these triple products decay with a coefficient of 1/{(2m+1) (2m′+1)}. Also, the frequencies of these triple products are on the odd-higher order. For example, when f_LO1=2nf_LO2 (n=2, 3, . . . ) and f_RF=f_LO1+f_LO2, the closest frequency is 7f_LO1/4. Here, if f_LO2=f_LO1/2 and m=m′=1, 3f_LO1/2=f_RF.

If an ideal doubly-polarity switching mixer has the same conversion gain in frequencies f_RF±if_LO1±jf_LO2, the in-band noise power converted from all bands around the frequencies f_RF±if_LO1±jf_LO2 is about six times (=4( 1/1²+⅓²+⅕²+ . . . )( 1/1²+⅓²+⅕²+ . . . )=π⁴/16=6.088) from the formula (36). Note 1/1²+⅓²+⅕²+ . . . +1/(2i+1)²+ . . . π²/8. Therefore, the noise figure (NF) of the ideal doubly-polarity switching mixer becomes damaged by 6.088, i.e., 7.844 dB, as compared with a circuit with a single-input signal and a single-output signal.

On the other hand, when a local oscillator signal S₂′(t) out of phase by π/2 from the local oscillator signal S₂(t) instead of the local oscillator signal S₂(t) is input to the doubly-polarity switching mixer of FIG. 9, the formulae (27), (28), (34) and (35) are replaced byS₁(t)=U_LO1(4/π)[cos(2πf_LO1t)−(⅓)cos(6πf_LO1t)+(⅕)cos(10πf_LO1t)−( 1/7)cos(14πf_LO1t)+ . . . ]  (37)S₂(t)′U_LO2(4/π)[sin(2πf_LO2t)+(⅓)sin(6πf_LO2t)+(⅕)sin(10πf_LO2t)+( 1/7)sin(14πf_LO2t)+ . . . ]  (38)sgn(S₁(t))=(4/π)[cos(2πf_LO1t)−(⅓)cos(6πf_LO1t)+( 5/1)cos(10πf_LO1t)−( 1/7)cos(14πf_LO1t)+ . . . ]  (39)sgn(S₂(t)′)=(4/π)[−sin(2πf_LO2t)+(⅓)sin(6πf_LO2t)−( 5/1)sin(10πf_LO2t)+( 1/7)sin(14πf_LO2t)+ . . . ]  (40)

Therefore, the formula (31) is represented by

$\begin{array}{cc}\begin{array}{c}{V}_{\mathrm{IF}}\left(t\right)=u_{\mathrm{RF}}·U_{\mathrm{L01}}·\mathrm{sgn}\left(S_1\left(t\right)\right)·U_{\mathrm{L02}}·\mathrm{sgn}\left(S_2^{\prime }\left(t\right)\right)\\ =U_{\mathrm{RF}}·\cos \left(2\pi f_{\mathrm{RF}}t\right)U_{\mathrm{L01}}·\mathrm{sgn}\left(S_1\left(t\right)\right)U_{\mathrm{L02}}·\mathrm{sgn}\left(S_2\left(t\right)\right)\\ =\left(16/{\pi }^2\right)U_{\mathrm{RF}}·U_{\mathrm{L01}}·U_{\mathrm{L02}}·\cos \left(2\pi f_{\mathrm{RF}}t\right)[\cos \left(2\pi f_{\mathrm{L01}}t\right)-\\ \left(1/3\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)+\left(1/5\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)-\\ \left(1/7\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)+\cdots ]\times [-\sin \left(2\pi f_{\mathrm{L02}}t\right)+\\ \left(1/3\right)\sin \left(6\pi f_{\mathrm{L02}}t\right)-\left(1/5\right)\sin \left(10\pi f_{\mathrm{L02}}t\right)+\\ \left(1/7\right)\sin \left(14\pi f_{\mathrm{L02}}t\right)+\cdots ]\\ =-\left(16/{\pi }^2\right)U_{\mathrm{RF}}·U_{\mathrm{L01}}·U_{\mathrm{L02}}·\\ [\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\sin \left(2\pi f_{\mathrm{L02}}t\right)+\\ \left(1/9\right)\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\sin \left(6\pi f_{\mathrm{L02}}t\right)+\\ \left(1/25\right)\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\sin \left(10\pi f_{\mathrm{L02}}t\right)+\\ \left(1/49\right)\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\sin \left(14\pi f_{\mathrm{L02}}t\right)-\\ \left(1/3\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\sin \left(2\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\sin \left(6\pi f_{\mathrm{L02}}t\right)\}+\\ \left(1/5\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\sin \left(2\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\sin \left(10\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/7\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\sin \left(2\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(2\pi f_{\mathrm{L01}}t\right)\sin \left(14\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/15\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\sin \left(10\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\sin \left(6\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/21\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(6\pi f_{\mathrm{L01}}t\right)\sin \left(14\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\sin \left(6\pi f_{\mathrm{L02}}t\right)\}-\\ \left(1/35\right)\{\cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(10\pi f_{\mathrm{L01}}t\right)\sin \left(14\pi f_{\mathrm{L02}}t\right)+\\ \cos \left(2\pi f_{\mathrm{RF}}t\right)\cos \left(14\pi f_{\mathrm{L01}}t\right)\sin \left(10\pi f_{\mathrm{L02}}t\right)\}+\cdots ]\end{array}& \left(41\right)\end{array}$

From the formula (41), when the RF signal U_RF is switched by the polarities of the local oscillator signals S₁(t) and S₂′(t), basic waves and harmonic waves of the three signals are obtained.

For example, four frequency components such as sin{2π(f_RF−f_LO1−f_LO2)t}, sin{2π(f_RF−f_LO1+f_LO2)t}, sin{2π(f_RF+f_LO1+f_LO2)t} and sin{2π(f_RF+f^LO1−f_LO2)t} are obtained from the triple product cos(2πf_RFt)cos(2πf_LO1t) sin(2πf_LO2t) of basic waves with reference to the formula of triple product of trigonometric functions shown in the formula (17).

In this case, if the doubly-polarity switching mixer of FIG. 9 constitutes a down-conversion system where a low-pass filter follows this doubly-polarity switching mixer, only the low frequency component sin{2π(f_RF−f_LO1−f_LO2)t} is obtained. Note that a difference |f_LO1−f_LO2| is defined as an intermediate frequency f_IF.

Also, four frequency components such as sin[2π{f_RF−(2m+1)f_LO1−(2m′+1)f_LO2}t], sin[2π{f_RF−(2m+1)f_LO1+(2m′+1)f_LO2}t], sin[2π{f_RF+(2m+1)f_LO1+(2m′+1)f_LO2}t] and sin[2π{f_RF+(2m+1)f_LO1−(2m′+1)f_LO2}t] are obtained from triple products cos(2πf_RFt) cos{2π(2m+1)f_LO1t}sin{2π(2m′+1)f_LO2t} (m, m′=1, 2, . . . ) of the RF signal V_RF(t) and the odd-higher harmonic waves of the local oscillator signals S₁(t) and S₂(t) with reference to the formula of triple product of trigonometric functions shown in the formula (17). In this case, these triple products decay with a coefficient of 1/{(2m+1) (2m′+1)}. Also, the frequencies of these triple products are on the odd-higher order. For example, when f_LO1=2nf_LO2 (n=2, 3, . . . ) and f_RF=f_LO1+f_LO2, the closest frequency is 7f_LO1/4. Here, if f_LO2=f_LO1/2 and m=m′=1, 3f_LO1/2=f_RF.

In FIG. 10A, which illustrates a detailed circuit diagram of the doubly-polarity switching mixer of FIG. 9, each of the switches 91, 92, 93 and 94 are constructed by emitter-coupled pairs (current switches) of bipolar transistors Q1, Q2; Q3, Q4; Q5, Q6; and Q7, Q8. In this case, one of the transistors Q1 and Q2, one of the transistors Q3 and Q4, one of the transistors Q5 and Q6, and one of the transistors Q7 and Q8 are turned ON, while the other of the transistors Q1 and Q2, the other of the transistors Q3 and Q4, the other of the transistors Q5 and Q6, and the other of the transistors Q7 and Q8 are turned OFF, so that differential currents I_RF⁺ and I_RF⁻ generated from a linear differential circuit 101 by the RF signal V_RF are switched. Thus, an intermediate signal V_IF is obtained. Note that the doubly-polarity switching mixer of FIG. 10A also serves as an analog circuit and accordingly, serves as a three-input multiplier.

In FIG. 10B, which is a detailed circuit diagram of the linear differential circuit 101 of FIG. 10A, a current mirror circuit is formed by bipolar transistors Q9 and Q10 and another current mirror circuit is formed by bipolar transistors Q11 and Q12. Collectors of the bipolar transistors Q10 and Q11 are connected to emitters of bipolar transistors Q13 and Q14, respectively, whose bases receive the RF signal V_RF. Also, an emitter degeneration resistor R_EE is connected between the emitters of the bipolar transistors Q13 and Q14. Further, one current source I_O is connected to each of the collectors of the bipolar transistors Q13 and Q14, and also, bases of bipolar transistors Q15 and Q16 serving as current sources to the current mirror circuits (Q9, Q10; Q11, Q12) are connected to the collectors of the bipolar transistors Q13 and Q14, respectively.

In FIG. 10B, the following differential currents I_RF⁺ and I_RF⁻ are generated from the collectors of the bipolar transistors Q9 and Q12.I_RF⁺=I_O+V_RF/R_EE  (42)I_RF⁻=I_O−V_RF/R_EE  (43)

That is, the linear characteristics of the linear differential circuit 101 are determined by the emitter degeneration resistor R_EE. Note that a resistance manufactured by a semiconductor manufacturing process has excellent linear characteristics. Therefore, the linear differential circuit 101 has excellent linear characteristics.

In FIG. 11A, which illustrates another detailed circuit diagram of the doubly-polarity switching mixer of FIG. 9, each of the switches 91, 92, 93 and 94 are constructed by source-coupled pairs (current switches) of MOS transistors M1, M2; M3, M4; M5, M6; and M7, M8. In this case, one of the transistors M1 and M2, one of the transistors M3 and M4, one of the transistors M5 and M6, and one of the transistors M7 and M8 are turned ON, while the other of the transistors M1 and M2, the other of the transistors M3 and M4, the other of the transistors M5 and M6, and the other of the transistors M7 and M8 are turned OFF, so that differential currents I_RF⁺ and I_RF⁻ generated from a linear differential circuit 111 by the RF signal V_RF are switched. Thus, an intermediate signal V_IF is obtained. In the doubly-polarity switching mixer of FIG. 11A, the transconductance does not change monotonously for a small signal change. Therefore, the doubly-polarity switching mixer of FIG. 11A does not serve as an analog circuit and accordingly, does not serve as a three-input multiplier.

In FIG. 11B, which is a detailed circuit diagram of the linear differential circuit 111 of FIG. 11A, a current mirror circuit is formed by MOS transistors M9 and M10 and another current mirror circuit is formed by MOS transistors M11 and M12. Drains of the MOS transistors M10 and M11 are connected to sources of MOS transistors M13 and M14, respectively, whose gates receives the RF signal V_RF. Also, a source degeneration resistor R_EE is connected between the sources of the MOS transistors M13 and M14. Further, one current source I_O is connected to each of the drains of the MOS transistors M13 and M14, and also, gates of MOS transistors M15 and M16 serving as current sources to the current mirror circuits (M9, M10; M11, M12) are connected to the drains of the MOS transistors M13 and M14, respectively.

Even in FIG. 11B, the differential currents I_RF⁺ and I_RF⁻ represented by the formulae (42) and (43) are generated from the drains of the MOS transistors M9 and M12.

That is, the linear characteristics of the linear differential circuit 111 are determined by the source degeneration resistor R_EE. Note that a resistance manufactured by a semiconductor manufacturing process has excellent linear characteristics. Therefore, the linear differential circuit III has excellent linear characteristics.

Generally, in frequency mixers, the suppression of high-order distortion characteristics such as second-order and third-order distortion characteristics, i.e., second-order and third-order intercept point characteristics are important. In the doubly-polarity switching mixers of FIGS. 10A and 11A, the high-order distortion characteristics can be sufficiently suppressed by the linear differential circuit of FIGS. 10B and 11B. However, in the doubly-polarity switching mixers of FIGS. 10A and 11A, the power supply voltage V_CC or V_DD needs to be higher than 2V.

In FIG. 12, which illustrates a modification of the doubly-polarity switching mixer of FIG. 10A, two triple tail cells C1 and C2 are provided. In the triple tail cell C1, three emitter-coupled bipolar transistors Q1′, Q2′ and Q2′ are driven by one tail current I_RF⁺, while in the triple tail cell C2, three emitter-coupled bipolar transistors Q4′, Q5′ and Q6′ are driven by one tail current I_RF⁻. The tail currents I_RF⁺ and I_RF⁻ are generated by a V-I conversion circuit 121 which can be easily constructed by bipolar transistors whose emitters are grounded via emitter resistors, where use is made of base voltage-to-collector current characteristics as V-I characteristics. In FIG. 12, the power supply voltage V_CC can be lower than 1V.

In more detail, the transistors Q1′, Q2′, Q4′ and Q5′ are switched by the local oscillator signal V_LO1 or S₁(t), and the transistors Q3′ and Q6′ are switched by the local oscillator signal V_LO2 or S₂(t). In this case, when the transistor Q3′ is turned ON by the local oscillator signal V_LO2, the tail current I_RF⁺ needs to be supplied from the power supply voltage V_CC regardless of whether the transistors Q1′ and Q2′ are turned ON or OFF. On the other hand, when the transistor Q6′ is turned ON by the local oscillator signal V_LO2, the tail current I_RF⁻ needs to be supplied from the power supply voltage V_CC regardless of whether the transistors Q4′ and Q5′ are turned ON or OFF. For this purpose, the transistors Q3′ and Q6′ are increased in size or the amplitude of the local oscillator signal V_LO2 is larger than that of the local oscillator signal V_LO1.

Thus, the doubly-polarity switching mixer of FIG. 12 is equivalent to the doubly-polarity switching mixer of FIGS. 10A and 10B and has an advantage in that the power supply voltage V_CC is low.

In FIG. 13, which illustrates a modification of the doubly-polarity switching mixer of FIG 11A, two triple tail cells C1 and C2 are provided. In the triple tail cell C1, three source-coupled MOS transistors M1′, M2′ and M2′ are driven by one tail current I_RF⁺, while, in the triple tail cell C2, three emitter-coupled MOS transistors M4′, M5′ and M6′ are driven by one tail current I_RF⁻. The tail currents I_RF⁺ and I_RF⁻ are generated by a V-I conversion circuit 131 which can be easily constructed by MOS transistors whose sources are grounded via source resistors, where use is made of gate voltage-to-drain current characteristics as V-I characteristics. In FIG. 13, the power supply voltage V_DD can be lower than 1V.

In more detail, the transistors M1′, M2′, M4′ and M5′ are switched by the local oscillator signal V_LO1 or S₁(t), and the transistors M3′ and M6′ are switched by the local oscillator signal V_LO2 or S₂(t). In this case, when the transistor M3′ is turned ON by the local oscillator signal V_LO2, the tail current I_RF⁺ needs to be supplied from the power supply voltage V_DD regardless of whether the transistors M1′ and M2′ are turned ON or OFF. On the other hand, when the transistor M6′ is turned ON by the local oscillator signal V_LO2, the tail current I_RF⁻ needs to be supplied from the power supply voltage V_DD regardless of whether the transistors M4′ and M5′ are turned ON or OFF. For this purpose, the transistors M3′ and M6′ are increased in size or the amplitude of the local oscillator signal V_LO2 is larger than that of the local oscillator signal V_LO1.

Thus, the doubly-polarity switching mixer of FIG. 13 is equivalent to the doubly-polarity switching mixer of FIGS. 11A and 11B and has an advantage in that the power supply voltage V_DD is low.

In FIG. 14A, which illustrates a modification of the doubly-polarity switching mixer of FIG. 12, two triple tail cells C3 and C4 are added, and a dual linear differential circuit 141 is provided instead of the V-I conversion circuit 121 of FIG. 12. The dual linear differential circuit 141 which is similar to the linear differential circuit 101 of FIG. 10B is illustrated in detail in FIG. 14B. In the triple tail cell C3, three emitter-coupled bipolar transistors Q7′, Q8′ and Q9′ are driven by one tail current V_RF⁺, and in the triple tail cell C4, three emitter-coupled bipolar transistors Q10′, Q11′ and Q12′ are driven by one tail current I_RF⁻. In this case, the intermediate signal V_IF is obtained by a difference between a sum current flowing through the transistors Q1′, Q5′, Q8′ and Q10′ and a sum current flowing through the transistors Q2′, Q4′, Q7′ and Q11′.

In FIG. 15A, which illustrates a modification of the doubly-polarity switching mixer of FIG. 13, two triple tail cells C3 and C4 are added, and a dual linear differential circuit 151 is provided instead of the V-I conversion circuit 131 of FIG. 13. The dual linear differential circuit 151 which is similar to the linear differential circuit 101 of FIG. 11B is illustrated in detail in FIG. 15B. In the triple tail cell C3, three emitter-coupled bipolar transistors M7′, M8′ and M9′ are driven by one tail current V_RF⁺, and in the triple tail cell C4, three emitter-coupled bipolar transistors M10′, M11′ and M12′ are driven by one tail current I_RF⁻. In this case, the intermediate signal V_IF is obtained by a difference between a sum current flowing through the transistors M1′, M5′, M8′ and M10′ and a sum current flowing through the transistors M2′, M4′, M7′ and M11′.

In FIG. 16, which illustrates a third embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention, a 1/2n(n=1, 2, . . . ) frequency divider 161 is provided instead of the ½-frequency divider 71 and 72 of FIG. 7. In the ¼-frequency divider 161 where n=1 is constructed by a Johnson counter, the quadrature mixer circuit of FIG. 16 is the same as that of FIG. 7.

In FIG. 16,f_RF=(2n+1)f_LO/(2n)

Also, since f_LO2=f_LO/(2n) the frequency f_LO2 of the output signals of the 1/2n-frequency divider 161 is represented byf_RF=(2n+1)f_LO2  (44)

As apparent from the formulae (28) and (38), since the output signals of the ½n-frequency divider 161 are rectangular, odd-higher order harmonic frequencies (2j−1) f_LO2 (j=2, 3, . . . ) are included therein. From the formula (44), some of such harmonic frequencies always coincide with the frequency f_RF of the RF signal V_RF, which would increase the DC offset and the reception trouble as in the prior art.

In FIG. 17, which illustrates a fourth embodiment of the quadrature mixer circuit applied to a direct conversion type wireless receiver according to the present invention, a 1/m-frequency divider 171 is connected to the voltage controlled oscillator 33″ to generate a first local oscillator signal u_LO1, and a 1/m′-frequency divider 172 is connected to the voltage controlled oscillator 33″ to generate a second local oscillator signal u_LO2. In this case,f_LO1=f_LO/mf_LO2=f_LO/m′then,f_RF=f_LO1+f_LO2=f_LO/m+f_LO/m′=(m+m′)/(m m′)·f_LO  (45)

The frequency component of the first local oscillator signal u_LO1 includes (2i−1) f_LO/m (i=1, 2, . . . ) and the frequency component of the second local oscillator signal u_LO2 includes (2j−1) f_LO/m′(j=1, 2, . . . ). These frequencies should not coincide with the frequency f_RF of the RF signal. That is,(m+m′)/(m m′)≠1(2i−1)/m≠(m+m′)/(m m′)(2j−1)/m′≠(m+m′)/(m m′)

In other words, 1/m+ 1/m′≠1  (46)i≠m/(2m′)+1  (47)j≠m′/(2m)+1  (48)

From the formulae (47) and (48), i. j=2, 3, . . . .

The values of m and m′ satisfying the formulae (46), (47) and (48) are shown in FIG. 18. In this case, any even-ordered high order waves of the local oscillator signals do not coincide with the frequency f_RF of the RF signal. Also, in any cases, f_LO>f_RF. For example, the values m and m′ satisfying that f_RF<f_LO<2f_RF are as follows:f_LO=6f_RF/5 at (m, m′)=(3, 2)f_LO=10f_RF/7 at (m, m′)=(5, 2)f_LO=14f_RF/9 at (m, m′)=(7, 2)f_LO=18f_RF/11 at (m, m′)=(9, 2)f_LO=5f_RF/3 at (m, m′)=(10, 2)f_LO=22f_RF/13 at (m, m′)=(11, 2)f_LO=26f_RF/15 at (m, m′)=(13, 2)f_LO=7f_RF/4 at (m, m′)=(14, 2)f_LO=52f_RF/17 at (m, m′)=(15, 2)f_LO=12f_RF/7 at (m, m′)=(3, 4)f_LO=3f_RF/2 at (m, m′)=(2, 6)

Also, the values of m and m′ satisfying that f_RF<f_LO<1.5f_RF are as follows:f_LO=6f_RF/5 at (m, m′)=(3, 2)f_LO=10f_RF/7 at (m, m′)=(5, 2)

In FIG. 17, if the 1/m′-frequency divider 172 includes a Johnson counter (½-frequency divider),m′=2n(n=1, 2, . . . )Then,f_RF=(m+2n)(2mn)·f_LO  (50)

Even in this case, the frequency component of the first local oscillator signal u_LO1 includes (2i−1) f_LO/m (i=1, 2, . . . ) and the frequency component of the second local oscillator signal u_LO2 includes (2j−1) f_LO/(2n)(j=1, 2, . . . ). These frequencies should not coincide with the frequency f_RF of the RF signal. That is,(m+2n)/(2m n)≠1(2i−1)/m≠(m+2n)/(2m n)(2j−1)/2n≠(m+2n)/(2m n)

In other words,1/m+½n≠1  (51)i≠m/(4n)+1  (52)j≠n/m+1  (53)

From the formulae (52) and (53), i. j=2, 3, . . . .

The values of m and 2n satisfying the formulae (51), (52) and (53) are shown in FIG. 19. In this case, any even-ordered high order waves of the local oscillator signals do not coincide with the frequency f_RF of the RF signal. Also, in any cases, f_LO>f_RF. For example, the values m and n satisfying that f_RF<f_LO<2f_RF are as follows:f_LO=6f_RF/5 at (m, n)=(3, 1)f_LO=10f_RF/7 at (m, n)=(5, 1)f_LO=14f_RF/9 at (m, n)=(7, 1)f_LO=18f_RF/11 at (m, n)=(9, 1)f_LO=5f_RF/3 at (m, n)=(10, 1)f_LO=22f_RF/13 at (m, n)=(11, 1)f_LO=26f_RF/15 at (m, n)=(13, 1)f_LO=7f_RF/4 at (m, n)=(14, 1)f_LO=52f_RF/17 at (m, n)=(15, 1)f_LO=12f_RF/7 at (m, n)=(3, 2)f_LO=3f_RF/2 at (m, n)=(2, 3)

Also, the values of m and m′ satisfying that f_RF<f_LO<1.5f_RF are as follows:f_LO=6f_RF/5 at (m, n)=(3, 1)f_LO=10f_RF/7 at (m, n)=(5, 1)

According to the inventor's calculation the noise factor (NF) of the quadrature mixer circuit according to the present invention was about 7 dB while the NF of the first prior art quadrature mixer circuit as illustrated in FIG. 1 was 3.01 dB. Such deterioration of the NF can be compensated for by increasing the power gain of the AGC amplifiers 6 and 7 in a direct conversion type wireless receiver, and accordingly, there is no actual problem.

Also, the present invention can be applied to a low IF type wireless receiver.

As explained hereinabove, according to the present invention, the DC offset can be decreased, the reception trouble can be suppressed, and the number of components can be decreased.

Claims

1. A quadrature mixer circuit for receiving a radio frequency signal to generate first and second quadrature output signals, comprising: a first three-input mixer for receiving said radio frequency signal, a first local signal having a first frequency and a second local signal having a second frequency to generate said first quadrature output signal; anda second three-input mixer for receiving said radio frequency signal, said first local signal and said second local signal to generate said second quadrature output signal,wherein said second local signal received by said first three-input mixer and said second local signal received by said second three-input mixer are out of phase by π/2 from each other;wherein each of said three-input mixers comprises a doubly-polarity switching mixer; andwherein said first and second local signals are rectangular.

2. The quadrature mixer circuit as set forth in claim 1, further comprising: an oscillator for generating said first local signal; anda frequency divider, connected to said oscillator, for generating said second local signal.

3. The quadrature mixer circuit as set forth in claim 2, wherein said frequency divider comprises a ½-frequency divider for generating said second local signal and a phase-shifted signal of said second local signal by a phase of π/2.

4. The quadrature mixer circuit as set forth in claim 1, further comprising: an oscillator;a 1/m′-frequency divider, connected to said oscillator, for generating said first local signal;a 1/m′-frequency divider, connected to said oscillator, for generating said second local signal.

5. The quadrature mixer circuit as set forth in claim 4, wherein the values of m and m′ satisfy: 1/m+ 1/m′≠1,i≠m(2m′)+1, andj≠m′/(2m)+1where i,j=2,3 . . . .

6. The quadrature mixer circuit as set forth in claim 1, further comprising: an oscillator;a 1/m-frequency divider, connected to said oscillator, for generating said first local signal;a 1/(2n)-frequency divider, connected to said oscillator, for generating said second local signal.

7. The quadrature mixer circuit as set forth in claim 6, wherein the values of m and n satisfy: 1/m+1/(2n)≠1,i≠m/(4n)+1, andj≠n/m+1where i,j=2,3 . . . .

8. The quadrature mixer circuit as set forth in claim 1, wherein said doubly-polarity switching mixer comprises non-linear elements showing three-order transfer characteristics for said radio frequency signal, said first local signal and said second local signal.

9. The quadrature mixer circuit as set forth in claim 8, wherein said non-linear elements comprise bipolar transistors.

10. The quadrature mixer circuit as set forth in claim 8, wherein said non-linear elements comprise MOS transistors.

11. The quadrature mixer circuit as set forth in claim 1, wherein said doubly-polarity switching mixer comprises: a linear differential circuit for generating first and second differential currents in response to said radio frequency signal;first and second bipolar transistors having a common emitter for receiving said first differential current, said first local signal being supplied between bases of said first and second bipolar transistors;third and fourth bipolar transistors having a common emitter for receiving said second differential current, said first local signal being supplied between bases of said third and fourth bipolar transistors;fifth and sixth bipolar transistors having a common emitter connected to a common collector of said first and third bipolar transistors, said second local signal being supplied between bases of said fifth and sixth bipolar transistors;seventh and eighth bipolar transistors having a common emitter connected to a common collector of said second and fourth bipolar transistors, said second local signal being supplied between bases of said seventh and eighth bipolar transistors;a respective one of said first and second quadrature output signals being a difference between a sum current flowing through said sixth and eighth bipolar transistors and a sum current flowing through said fifth and seventh bipolar transistors.

12. The quadrature mixer circuit as set forth in claim 1, wherein said doubly-polarity switching mixer comprises: a linear differential circuit for generating first and second differential currents in response to said radio frequency signal;first and second MOS transistors having a common source for receiving said first differential current, said first local signal being supplied between gates of said first and second MOS transistors;third and fourth MOS transistors having a common source for receiving said second differential current, said first local signal being supplied between gates of said third and fourth MOS transistors;fifth and sixth MOS transistors having a common source connected to a common drain of said first and third MOS transistors, said second local signal being supplied between gates of said fifth and sixth MOS transistors;seventh and eighth MOS transistors having a common source connected to a common drain of said second and fourth MOS transistors, said second local signal being supplied between gates of said seventh and eighth MOS transistors;a respective one of said first and second quadrature output signals being a difference between a sum current flowing through said sixth and eighth MOS transistors and a sum current flowing through said fifth and seventh MOS transistors.

13. The quadrature mixer circuit as set forth in claim 1, wherein said doubly-polarity switching mixer comprises: a voltage-to-current conversion circuit for converting said radio frequency signal into first and second differential currents;first, second and third bipolar transistors having a common emitter for receiving said first differential current, said first local signal being supplied between bases of said first and second bipolar transistors, a collector of said third bipolar transistor being connected to a power supply terminal; andfourth, fifth and sixth bipolar transistors having a common emitter for receiving said second differential current, said first local signal being supplied between bases of said fourth and fifth bipolar transistors, a collector of said sixth bipolar transistor being connected to said power supply terminal,said second local signal being supplied between bases of said third and sixth bipolar transistors,a respective one of said first quadrature output signals being a difference between a sum current flowing through said first and fifth bipolar transistors and a sum current flowing through said second and fourth bipolar transistors.

14. The quadrature mixer circuit as set forth in claim 1, wherein said doubly-polarity switching mixer comprises: a voltage-to-current conversion circuit for converting said radio frequency signal into first and second differential currents;first, second and third MOS transistors having a common source for receiving said first differential current, said first local signal being supplied between gates of said first and second MOS transistors, a drain of said third MOS transistor being connected to a power supply terminal; andfourth, fifth and sixth MOS transistors having a common source for receiving said second differential current, said first local signal being supplied between gates of said fourth and fifth MOS transistors, a drain of said sixth MOS transistor being connected to said power supply terminal,said second local signal being supplied between gates of said third and sixth MOS transistors,a respective one of said first quadrature output signals being a difference between a sum current flowing through said first and fifth MOS transistors and a sum current flowing through said second and fourth MOS transistors.

15. The quadrature mixer circuit as set forth in claim 1, wherein said doubly-polarity switching mixer comprises: a dual linear differential circuit for generating first, second, third and fourth differential currents, said first and second differential currents being similar to each other, said third and fourth differential currents being similar to each other;first, second and third bipolar transistors having a common emitter for receiving said first differential current, said first local signal being supplied between bases of said first and second bipolar transistors, a collector of said third bipolar transistor being connected to a power supply terminal;fourth, fifth and sixth bipolar transistors having a common emitter for receiving said second differential current, said first local signal being supplied between bases of said fourth and fifth bipolar transistors, a collector of said sixth bipolar transistor being connected to said power supply terminal;seventh, eighth and ninth bipolar transistors having a common emitter for receiving said third differential current, said first local signal being supplied between bases of said seventh and eighth bipolar transistors, a collector of said ninth bipolar transistor being connected to said power supply terminal; andtenth, eleventh and twelfth bipolar transistors having a common emitter for receiving said fourth differential current, said first local signal being supplied between bases of said tenth and eleventh bipolar transistors, a collector of said twelfth bipolar transistor being connected to said power supply terminal;said second local signal being supplied between bases of said third and sixth bipolar transistors,said second local signal being supplied between bases of said ninth and twelfth bipolar transistors,a respective one of said first quadrature output signals being a difference between a sum current flowing through said first, fifth, eighth and tenth bipolar transistors and a sum current flowing through said second, fourth, seventh and eleventh bipolar transistors.

16. The quadrature mixer circuit as set forth in claim 1, wherein said doubly-polarity switching mixer comprises: a dual linear differential circuit for generating first, second, third and fourth differential currents, said first and second differential currents being similar to each other, said third and fourth differential currents being similar to each other;first, second and third MOS transistors having a common source for receiving said first differential current, said first local signal being supplied between gates of said first and second MOS transistors, a drain of said third MOS transistor being connected to a power supply terminal;fourth, fifth and sixth MOS transistors having a common source for receiving said second differential current, said first local signal being supplied between gates of said fourth and fifth MOS transistors, a drain of said sixth MOS transistor being connected to said power supply terminal;seventh, eighth and ninth MOS transistors having a common source for receiving said third differential current, said first local signal being supplied between gates of said seventh and eighth MOS transistors, a drain of said ninth MOS transistor being connected to said power supply terminal; andtenth, eleventh and twelfth MOS transistors having a common source for receiving said fourth differential current, said first local signal being supplied between gates of said tenth and eleventh MOS transistors, a drain of said twelfth MOS transistor being connected to said power supply terminal;said second local signal being supplied between gates of said third and sixth MOS transistors,said second local signal being supplied between gates of said ninth and twelfth MOS transistors,a respective one of said first quadrature output signals being a difference between a sum current flowing through said first, fifth, eighth and tenth MOS transistors and a sum current flowing through said second, fourth, seventh and eleventh MOS transistors.

17. The quadrature mixer circuit as set forth in claim 1, applied to a direct conversion type wireless receiver.

18. The quadrature mixer circuit as set forth in claim 1, applied to a low intermediate frequency type wireless receiver.