Quadrature mixer circuit including three-input local mixers

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.

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.

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 .

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 by

f ( u )= a 0 +a 1 u+a 2 u 2 + . . . +a n u n +  (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 z u z of the formula (1) is represented by

a z u 2 =a 2 ( u RF +u LO ) 2 =a 2 ( u RF 2 +u LO 2 +2 u RF u LO )  (2)

In this case,

u RF =U RF ·cos(2 πf RF t )  (3)

u LO =U LO ·cos(2 πf LO t )  (4)

Therefore, the third term of the formula (2) is represented by

2 U RF ·u LO =2 U RF ·U LO ·cos(2 πf RF t )·cos(2 πf LO t )= 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/(1 2 +1 2 )) 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 by

u LO ′=U LO ·sin(2 πf LO t )  (6)

Therefore, the third term of the formula (2) is represented by

2 U RF ·u LO ′=2 U RF ·U LO ·cos(2 πf RF t )·sin(2 πf LO t )= 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 X 1 and X 2 , respectively. The reflected local oscillator signals X 1 and X 2 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 a 3 u 3 of the formula (1) is represented by

a 3 u 3 =a 3 ( u RF +u LO ) 3 =a 3 ( u RF 3 +u LO 3 +3 u RF 2 u LO +3 u RF u LO 2 )  (8)

The fourth term of the formula (8) is represented by

3 u RF u LO 2 =3 U RF ·U LO 2 ·cos(2 πf RF t )cos 2 (2 πf LO t )=3 U RF ·U LO 2 ·cos(2 πf RF t ){1+cos(4 πf LO t )}/2=3 U RF ·U LO 2 ·{cos(2 πf RF t )+cos(2 πf RF t )cos(4 πf LO t )}/2=3 U RF ·U LO 2 [2 cos(2 πf RF t )+cos{2π( f RF t −2 f LO ) t }+cos{2π( f RF t +2 f 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 LO 2 cos{2π(f RF t−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/(2 2 +1 2 +1 2 )) 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 by

u LO ′=u LO ·cos(2 πf LO t +π/4)  (10)

Therefore, the fourth-order term of the formula (8) is represented by

3 u RF u LO ′ 2 =3 U RF ·U LO 2 ·cos(2 πf RF t )cos 2 (2 πf LO t +π/4)=3 U RF ·U LO 2 ·cos(2 πf RF t ){1+cos(4 πf LO t +π/2)}/2=3 U RF ·U LO 2 ·cos(2 πf RF t ){1−sin(4 πf LO t )}/2=3 U RF ·U LO 2 ·{cos(2 πf RF t )−cos(2 πf RF t )sin(4 πf LO t )}/2=3 U RF ·U LO 2 ·[2 cos(2π f RF t )+sin{2π( f RF t −2 f LO ) t }+sin{2π( f RF t +2 f 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 LO 2 ·sin{2π(f RF t−2f LO )t} of the formula (14) is obtained. Note that a difference |f RF t−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 a 3 u 3 of the formula (1) is replaced by

a 3 u 3 =a 3 ( u RF +u LO1 +u LO2 ) 3 =a 3 ( u RF 3 +u LO1 3 +u LO2 3 +3 u RF 2 u LO1 +3 u RF 2 u LO2 +3 u RF u LO1 2 +3 u LO1 2 u LO2 +3 u RF u LO2 2 +3 u LO1 u LO2 2 +6 u RF u LO1 u LO2 )  (13)

In this case, the RF signal u RF and the local oscillator signals u LO1 and u LO2 are represented by

u RF =u RF ·cos(2 πf RF t )  (14)

u LO1 =u LO1 ·cos(2 πf LO1 t )  (15)

u LO2 =u LO2 ·cos(2 πf LO2 t )  (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 by

6 u RF u LO1 u LO2 =6 U RF ·U LO1 ·U LO2 ·cos(2 πf RF t )cos(2 πf LO1 t )cos(2 πf LO2 t )=3 U 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 by

u LO2 ′=U LO2 sin(2 πf LO2 t )  (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 by

6 u RF u LO1 u LO2 =6 U RF ·U LO1 ·U LO2 ·cos(2 πf RF t )cos(2 πf LO1 t )sin(2 πf LO2 t )=3 U 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 a 3 u 3 of the formula (1) is replaced by

a 3 u 3 =a 3 ( u RF +u LO +u LO/4 ) 3 =a 3 ( u RF 3 +u LO 3 +u LO/4 3 +3 u RF 2 u LO +3 u RF 2 u LO/4 +3 u RF u LO 2 +3 u LO 2 u LO/4 +3 u RF u LO/4 2 +3 u LO u LO/4 2 +6 u RF u LO u LO/4 )  (22)

In this case, the local oscillator signal u LO/4 is represented by

u LO/4 =u LO/4 ·cos{2π( f LO /4) t}   (23)

Therefore, the tenth term of the formula (22) is represented by

6 u RF u LO u LO/4 =6 U RF ·U LO ·U LO/4 ·cos(2 πf RF t )cos(2 πf LO t )cos{2π( f LO /4) t }=3 U RF ·U LO ·U LO/4 ·[cos{2π( f RF −5 f LO /4) t }+cos{2π( f RF −3 f LO /4) t }+cos{2π( f RF +5 f LO /4) t }+cos{2π( f RF +3 f 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 by

u LO4 ′U LO/4 ·sin{2π( f LO /4) t}   (25)

Therefore, the tenth term of the formula (24) is represented by

6 u RF u LO u LO/4 =6 U RF ·U LO ·U LO/4 ·cos(2 πf RF t )cos(2 πf LO t )sin{2π( f LO /4)t}=3 U RF ·U LO ·U LO/4 ·[sin{2π( f RF −5 f LO /4) t }+sin{2π( f RF −3 f LO /4) t }+sin{2π( f RF +5 f LO /4) t }+sin{2π( f RF +3 f 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 , 10 A, 10 B, 11 A, 11 B, 12 , 13 , 14 A, 14 B, 15 A and 15 B.

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 S 1 (t) and S 2 (t). That is, switches 91 , 92 , 93 and 94 controlled by the local oscillator signals S 1 (t) and S 2 (t) receive the RF signal V RF (t) to generate an intermediate frequency signal V IF . If the local oscillator signal S 1 (t) has an amplitude of ±U LO1 and a frequency of f LO1 and the local oscillator signal S 2 (t) has an amplitude of ±U LO2 and a frequency of f LO2 , S 1 (t) and S 2 (t) are represented by

S 1 ( t )= U LO1 (4/π)[cos(2 πf LO1 t )−(⅓)cos(6 πf LO1 t )+(⅕)cos(10 πf LO1 t )−( 1/7)cos(14 πf LO1 t )+ . . . ]  (27)

S 2 ( t )= U LO2 (4/π)[cos(2 πf LO2 t )−(⅓)cos(6 πf LO2 t )+(⅕)cos(10 πf LO2 t )−( 1/7)cos(14 πf LO2 t )+ . . . ]  (28)

Also, if S 1 (t)=U LO1 sgn(S 1 (t)) and S 2 (t)=U LO2 sgn(S 2 (t)),

V IF ( t )= u RF ·U LO1 sgn ( S 1 ( t ) U LO2 sgn ( S 2 (t))  (29)

u RF =U RF cos(2 πf RF t )  (30)

Therefore,

V IF ( t )= u RF ·U LO1 sgn ( S 1 ( t )) U LO2 sgn ( S 2 (t))= u RF cos(2 πf RF t ) U LO1 sgn ( S 1 ( t )) U LO2 sgn ( S 2 ( t ))  (31)

Since sgn(S 1 (t)) and sgn(S 2 (t)) are −1 and +1, and −1 and +1, respectively, their absolute values are represented by

| sgn ( S 1 ( t ))|=1  (32)

| sgn ( S 2 ( t ))|=1  (33)

Therefore, sgn(S 1 (t)) and sgn(S 2 (t)) are represented by

sgn ( S 1 ( t ))=(4/π)[cos(2 πf LO1 t )−(⅓)cos(6 πf LO1 t )+( 5/1)cos(10 πf LO1 t )−( 1/7)cos(14 πf LO1 t )+ . . . ]  (34)

sgn ( S 2 ( t ))=(4/π)[cos(2 πf LO2 t )−(⅓)cos(6 πf LO2 t )+( 5/1)cos(10 πf LO2 t )−( 1/7)cos(14 πf LO2 t )+ . . . ]  (35)

Thus, the formula (31) is represented by

Apparent from the formula (36), when the RF signal U RF is switched by the polarities of the local oscillator signals S 1 (t) and S 2 (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−f LO2 )t} are obtained from the triple product cos(2πf RF t)cos(2πf LO1 t) cos(2πf LO2 t) 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 RF t) cos{2π(2m+1)f LO1 t}cos{2π(2m′+1)f LO2 t} (m, m′=1, 2, . . . ) of the RF signal V RF (t) and the odd-higher harmonic waves of the local oscillator signals S 1 (t) and S 2 (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( 1/1 2 +⅓ 2 +⅕ 2 + . . . )( 1/1 2 +⅓ 2 +⅕ 2 + . . . )=π 4 /16=6.088) from the formula (36). Note 1/1 2 +⅓ 2 +⅕ 2 + . . . +1/(2i+1) 2 + . . . π 2 /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 S 2 ′(t) out of phase by π/2 from the local oscillator signal S 2 (t) instead of the local oscillator signal S 2 (t) is input to the doubly-polarity switching mixer of FIG. 9 , the formulae (27), (28), (34) and (35) are replaced by

S 1 ( t )= U LO1 (4/π)[cos(2 πf LO1 t )−(⅓)cos(6 πf LO1 t )+(⅕)cos(10 πf LO1 t )−( 1/7)cos(14 πf LO1 t )+ . . . ]  (37)

S 2 ( t )′ U LO2 (4/π)[sin(2 πf LO2 t )+(⅓)sin(6 πf LO2 t )+(⅕)sin(10 πf LO2 t )+( 1/7)sin(14 πf LO2 t )+ . . . ]  (38)

sgn ( S 1 ( t ))=(4/π)[cos(2 πf LO1 t )−(⅓)cos(6 πf LO1 t )+( 5/1)cos(10 πf LO1 t )−( 1/7)cos(14 πf LO1 t )+ . . . ]  (39)

sgn ( S 2 ( t )′)=(4/π)[−sin(2 πf LO2 t )+(⅓)sin(6 πf LO2 t )−( 5/1)sin(10 πf LO2 t )+( 1/7)sin(14 πf LO2 t )+ . . . ]  (40)

Therefore, the formula (31) is represented by

From the formula (41), when the RF signal U RF is switched by the polarities of the local oscillator signals S 1 (t) and S 2 ′(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 RF t)cos(2πf LO1 t) sin(2πf LO2 t) 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 RF t) cos{2π(2m+1)f LO1 t}sin{2π(2m′+1)f LO2 t} (m, m′=1, 2, . . . ) of the RF signal V RF (t) and the odd-higher harmonic waves of the local oscillator signals S 1 (t) and S 2 (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 Q 1 , Q 2 ; Q 3 , Q 4 ; Q 5 , Q 6 ; and Q 7 , Q 8 . In this case, one of the transistors Q 1 and Q 2 , one of the transistors Q 3 and Q 4 , one of the transistors Q 5 and Q 6 , and one of the transistors Q 7 and Q 8 are turned ON, while the other of the transistors Q 1 and Q 2 , the other of the transistors Q 3 and Q 4 , the other of the transistors Q 5 and Q 6 , and the other of the transistors Q 7 and Q 8 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 Q 9 and Q 10 and another current mirror circuit is formed by bipolar transistors Q 11 and Q 12 . Collectors of the bipolar transistors Q 10 and Q 11 are connected to emitters of bipolar transistors Q 13 and Q 14 , 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 Q 13 and Q 14 . Further, one current source I O is connected to each of the collectors of the bipolar transistors Q 13 and Q 14 , and also, bases of bipolar transistors Q 15 and Q 16 serving as current sources to the current mirror circuits (Q 9 , Q 10 ; Q 11 , Q 12 ) are connected to the collectors of the bipolar transistors Q 13 and Q 14 , respectively.

In FIG. 10B , the following differential currents I RF + and I RF − are generated from the collectors of the bipolar transistors Q 9 and Q 12 .

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 M 1 , M 2 ; M 3 , M 4 ; M 5 , M 6 ; and M 7 , M 8 . In this case, one of the transistors M 1 and M 2 , one of the transistors M 3 and M 4 , one of the transistors M 5 and M 6 , and one of the transistors M 7 and M 8 are turned ON, while the other of the transistors M 1 and M 2 , the other of the transistors M 3 and M 4 , the other of the transistors M 5 and M 6 , and the other of the transistors M 7 and M 8 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 M 9 and M 10 and another current mirror circuit is formed by MOS transistors M 11 and M 12 . Drains of the MOS transistors M 10 and M 11 are connected to sources of MOS transistors M 13 and M 14 , 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 M 13 and M 14 . Further, one current source I O is connected to each of the drains of the MOS transistors M 13 and M 14 , and also, gates of MOS transistors M 15 and M 16 serving as current sources to the current mirror circuits (M 9 , M 10 ; M 11 , M 12 ) are connected to the drains of the MOS transistors M 13 and M 14 , 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 M 9 and M 12 .

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 C 1 and C 2 are provided. In the triple tail cell C 1 , three emitter-coupled bipolar transistors Q 1 ′, Q 2 ′ and Q 2 ′ are driven by one tail current I RF + , while in the triple tail cell C 2 , three emitter-coupled bipolar transistors Q 4 ′, Q 5 ′ and Q 6 ′ 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 Q 1 ′, Q 2 ′, Q 4 ′ and Q 5 ′ are switched by the local oscillator signal V LO1 or S 1 (t), and the transistors Q 3 ′ and Q 6 ′ are switched by the local oscillator signal V LO2 or S 2 (t). In this case, when the transistor Q 3 ′ 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 Q 1 ′ and Q 2 ′ are turned ON or OFF. On the other hand, when the transistor Q 6 ′ 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 Q 4 ′ and Q 5 ′ are turned ON or OFF. For this purpose, the transistors Q 3 ′ and Q 6 ′ 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 11 A, two triple tail cells C 1 and C 2 are provided. In the triple tail cell C 1 , three source-coupled MOS transistors M 1 ′, M 2 ′ and M 2 ′ are driven by one tail current I RF + , while, in the triple tail cell C 2 , three emitter-coupled MOS transistors M 4 ′, M 5 ′ and M 6 ′ 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 M 1 ′, M 2 ′, M 4 ′ and M 5 ′ are switched by the local oscillator signal V LO1 or S 1 (t), and the transistors M 3 ′ and M 6 ′ are switched by the local oscillator signal V LO2 or S 2 (t). In this case, when the transistor M 3 ′ 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 M 1 ′ and M 2 ′ are turned ON or OFF. On the other hand, when the transistor M 6 ′ 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 M 4 ′ and M 5 ′ are turned ON or OFF. For this purpose, the transistors M 3 ′ and M 6 ′ 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 C 3 and C 4 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 C 3 , three emitter-coupled bipolar transistors Q 7 ′, Q 8 ′ and Q 9 ′ are driven by one tail current V RF + , and in the triple tail cell C 4 , three emitter-coupled bipolar transistors Q 10 ′, Q 11 ′ and Q 12 ′ 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 Q 1 ′, Q 5 ′, Q 8 ′ and Q 10 ′ and a sum current flowing through the transistors Q 2 ′, Q 4 ′, Q 7 ′ and Q 11 ′.

In FIG. 15A , which illustrates a modification of the doubly-polarity switching mixer of FIG. 13 , two triple tail cells C 3 and C 4 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 C 3 , three emitter-coupled bipolar transistors M 7 ′, M 8 ′ and M 9 ′ are driven by one tail current V RF + , and in the triple tail cell C 4 , three emitter-coupled bipolar transistors M 10 ′, M 11 ′ and M 12 ′ 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 M 1 ′, M 5 ′, M 8 ′ and M 10 ′ and a sum current flowing through the transistors M 2 ′, M 4 ′, M 7 ′ and M 11 ′.

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 =(2 n +1) f LO /(2 n )

Also, since f LO2 =f LO /(2n) the frequency f LO2 of the output signals of the 1/2n-frequency divider 161 is represented by

f RF =(2 n +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 /m

f 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

(2 i −1)/ m ≠( m+m ′)/( m m ′)

(2 j −1)/ m ′≠( m+m ′)/( m m ′)

In other words,

1 /m + 1 /m ′≠1  (46)

i≠m /(2 m ′)+1  (47)

j≠m ′/(2 m )+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 =6 f RF /5 at ( m, m ′)=(3, 2)

f LO =10 f RF /7 at ( m, m ′)=(5, 2)

f LO =14 f RF /9 at ( m, m ′)=(7, 2)

f LO =18 f RF /11 at ( m, m ′)=(9, 2)

f LO =5 f RF /3 at ( m, m ′)=(10, 2)

f LO =22 f RF /13 at ( m, m ′)=(11, 2)

f LO =26 f RF /15 at ( m, m ′)=(13, 2)

f LO =7 f RF /4 at ( m, m ′)=(14, 2)

f LO =52 f RF /17 at ( m, m ′)=(15, 2)

f LO =12 f RF /7 at ( m, m ′)=(3, 4)

f LO =3 f 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 =6 f RF /5 at ( m, m ′)=(3, 2)

f LO =10 f 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′ =2 n ( n =1, 2, . . . )

Then,

f RF =( m +2 n )(2 mn )· 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 +2 n )/(2 m n )≠1

(2 i −1)/ m ≠( m +2 n )/(2 m n )

(2 j −1)/2 n ≠( m +2 n )/(2 m n )

In other words,

1 /m +½ n≠ 1  (51)

i≠m /(4 n )+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 =6 f RF /5 at ( m, n )=(3, 1)

f LO =10 f RF /7 at ( m, n )=(5, 1)

f LO =14 f RF /9 at ( m, n )=(7, 1)

f LO =18 f RF /11 at ( m, n )=(9, 1)

f LO =5 f RF /3 at ( m, n )=(10, 1)

f LO =22 f RF /13 at ( m, n )=(11, 1)

f LO =26 f RF /15 at ( m, n)=( 13, 1)

f LO =7 f RF /4 at ( m, n )=(14, 1)

f LO =52 f RF /17 at ( m, n )=(15, 1)

f LO =12 f RF /7 at ( m, n )=(3, 2)

f LO =3 f 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 =6 f RF /5 at ( m, n )=(3, 1)

f LO =10 f 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.