US 6,987,425 B1Grant
Low phase noise MOS LC oscillator
Issue Date:2006-01-17
•15 Claims
•13 Drawing Sheets
Abstract
A cross-coupled differential MOS oscillator having reduced phase noise is applicable to a RF communication device such as a transmitter or receiver. The oscillator having low phase noise is formed of a frequency dependent amplifier to amplify a signal having a fundamental frequency; a frequency dependent feedback device that is connected between an output of the frequency dependent amplifier and an input of the frequency dependent amplifier to feed a portion of an amplified signal having the fundamental frequency to an input of the frequency dependent amplifier to stimulate oscillation; and a attenuating device in communication with the frequency dependent amplifier. The attenuating device reduces the gain of the frequency dependent amplifier for signals having frequencies much, much less than the fundamental frequency to decrease the phase noise.
Metadata
Assignee
- Marvell International Ltd.
Inventor
- Sehat Sutardja
Application Information
Application Number:US 10/780,754
Filing Date:2004-02-19
Priority Date:2000-05-17
Art Unit:2817
Classifications
IPC:
H03L5/00
Patent Drawings (13 sheets)
Description
[0001] This application is a continuation of, and hereby claims priority under 35 U.S.C. §120 on, U.S. application Ser. No. 09/699,772 filed on Sep. 26, 2000, now U.S. Pat. No. 6,376,101 the entire contents of which are hereby incorporated herein by reference, which is based on, and which claims priority under 35 U.S.C. §120 on, provisional patent application, 60/204,885, filed on May 17, 2000.
Background of the Invention
[0002] 1. Field of the Invention
[0003] This invention relates to high frequency oscillator circuits. More particularly, this invention relates to metal oxide semiconductor (MOS) oscillators having low phase noise.
[0004] 2. Description of the Related Art
[0005] Inductive/capacitive (LC) oscillators are important elements of any Radio Frequency (RF) communication devices, such as transmitters, where the LC oscillators are used as master oscillators, or as receivers where the LC oscillators are used as local oscillators. An important performance benchmark of an LC oscillator is the phase noise characteristic. An oscillator with a lower phase noise indicates that the oscillator produces lower spurious energy outside the desired fundamental signal tone.
[0006] Phase noise is produced as a result of low frequency noise signal found in active elements used in the oscillator. This low frequency signal is modulated (up converted) by the fundamental signal tone, resulting in the spreading of the oscillator frequency energy beyond the intended target frequency. This low frequency noise signal source is often referred to as flicker noise (commonly referred to in the literature as 1/f) in bipolar and Metal Oxide Semiconductor (MOS) transistors. The 1/f noise energy in bipolar transistors is known to be significantly less than that of MOS transistors. This is the reason why practically all low phase noise LC oscillators are built using bipolar transistors or even more esoteric transistors such as Galium-Arsenide devices.
[0007] Complementary MOS (CMOS) based LC oscillators are now being investigated again for application to systems-on-a-chip (SOC) devices for RF communication applications. LC oscillators of the prior art fall far short of the minimum performance requirements of many of today's wireless communication systems.
[0008] A typical example of an LC oscillator in MOS technology is shown in FIG. 2. It is based on cross-coupled NMOS transistors M1 and M2, a pair of inductors L1 and L2, and capacitor C1 and C2 tuning elements. PMOS transistors, which usually have slightly lower 1/f noise characteristics, can be used to replace the NMOS transistors M1 and M2 at a slight increase in power dissipation and lower maximum operating frequency.
[0009] A review of a general form of the criteria for designing an oscillator circuit of the prior art is shown in FIG, 1. The necessary components of an oscillator are a frequency dependent gain circuit 100, a frequency dependent feedback circuit 105, and a combining block 110. The output V0 120 of the gain circuit 100 is the input to the feedback circuit 105. The input signal V1 115 is combined in the combining block 110 with the output Vfb 107 of the feedback circuit 105 to form the input 112 of the gain circuit 100.
[0010] The gain of the gain block 100 is designated G(jω) and the gain of the feedback circuit 105 is designated H(jω). These gains G(jω) and H(jω) describe the relationship of their respective output signals Vo 120 and Vfb 107 to their respective input signals 112 and Vo 120. Therefore, the output signal Vo 120 becomes
[0011] For an oscillator, the output signal Vo 120 must be nonzero even if the input voltage V1 115 is zero. For this to be true, then
1+G(jω)H(jω)=0
or
G(jω)H(jω)=−1.
That is, the magnitude of the open-loop transfer function must be equal to 1 and the phase shift of the gain circuit 100 and the feedback circuit 105 must be 180°.
1+G(jω)H(jω)=0
or
G(jω)H(jω)=−1.
That is, the magnitude of the open-loop transfer function must be equal to 1 and the phase shift of the gain circuit 100 and the feedback circuit 105 must be 180°.
[0012] In FIG. 2 , the gain circuit of the oscillator is formed by the differentially cross-connected pair of transistors M1 and M2 and the constant current source I1. The frequency dependent gain determining impedances are formed by the inductors L1 and L2 and the capacitors C1 and C2.
[0013] The feedback circuit is accomplished by the connecting of the drain of the NMOS transistor M1 to the gate of the NMOS transistor M2 and the drain of the NMOS transistor M2 to the gate of the NMOS transistor M1. This forms a cross-coupled differential oscillator.
[0014] A CMOS oscillator of the prior art is illustrated in FIG. 3. In this case, the gain circuit is formed by the differentially connected pair of NMOS transistors M1 and M2, the differentially connected pair of PMOS transistors M3 and M4, and the current sources I1 and I2. As described above, the frequency dependent gain determining impedances are formed by the inductors L1 and L2 and capacitors C1 and C2.
[0015] The fundamental frequency f0 of a cross coupled differential oscillator is determined by the formula:
-
- [0016] where:
- [0017] Leff is the value of the effective inductance of the inductors L1 and L2.
- [0018] Ceff is the value of the effective capacitance of the capacitors C1 and C2.
For the structure of the design where the inductors are mutually coupled then the effective inductance is:
Leff=4L1=4L2.
The effective capacitance of the capacitors C1 and C2 is the parallel combination of the two capacitors C1 and C2 and is:
Ceff=½C1=½C2
Combining the above, the frequency of the oscillators ofFIGS. 2 and 3 is:
It should be noted that the capacitances C1 and C2 included the parasitic capacitances of the oscillator circuit.
- [0016] where:
[0019] It is well known in the art that phase noise is the result of small perturbations in phase due to small random shifts in oscillator frequency. These shifts are caused by thermal noise, shot noise, and flicker noise (1/f noise). These noises are functions of the device characteristics of the NMOS transistors M1 and M2 of FIGS. 2 and 3 and the PMOS transistors M3 and M4 of FIG. 3. The phase noise is modeled as small voltage sources Vn1 and Vn2 at the gates of the NMOS transistors M1 and M2 of FIGS. 2 and 3 and voltage sources Vp1 and Vp2 at the gates of the PMOS transistors M4 and M4 of FIG. 3.
[0020] This flicker noise (1/f noise) is a function of the active device characteristics of the NMOS transistors M1 and M2 of FIGS. 1 and 2 and PMOS transistors M3 and M4 of FIG. 3.
[0021] The advancements in scaling of the device features in semiconductor processing allow multi-gigahertz operating frequencies to be readily achievable. Unfortunately, the same scaling down of MOS transistors have the opposite effect on the 1/f noise characteristics. The smaller device geometries are, the higher the 1/f noise components, leading to higher phase noise on the final oscillator.
[0022] “A 1.8 Ghz CMOS Voltage-Controlled Oscillator”, —Razavi, B., Digest of Technical Papers. 43rd ISSCC, 1997, pp. 388-389 and shown in FIG. 4 describes a structure of having multiple oscillators OSC1 and OSC2 coupled together to oscillate in quadrature or 90° out-of-phase. The oscillators OSC1 and OSC2 are structured and function as described in FIG. 2. The differential pair of NMOS transistors M3 and M4 and the current source I2 form a first coupling circuit. The first coupling circuit has an in-phase input that is formed by the gate of the NMOS transistors M3 and a out-of-phase input that is formed by the gate of the NMOS transistor M4. The first coupling circuit has a in-phase output that is formed by the drain of the NMOS transistor M4 and an out-of-phase output that is formed by the drain of the NMOS transistor M3. The in-phase input of the first coupling circuit is connected to the drain of the NMOS transistor M5 and the gate of the NMOS transistor M6. The out-of-phase input of the first coupling circuit is connected to the drain of the NMOS transistor M6 and the gate of the NMOS transistor M5. The in-phase output of the first coupling circuit is connected to the drain of the NMOS transistor M2 and the gate of the NMOS transistor M1. The out-of-phase output of the first coupling circuit is connected to the drain of the NMOS transistor M1 and the gate of the NMOS transistor M2.
[0023] The differential pair of NMOS transistors M7 and M8 and the current source I4 form a second coupling circuit. The second coupling circuit has an in-phase input that is formed by the gate of the NMOS transistors M7 and a out-of-phase input that is formed by the gate of the NMOS transistor M8. The second coupling circuit has a in-phase output that is formed by the drain of the NMOS transistor M8 and an out-of-phase output that is formed by the drain of the NMOS transistor M7. The in-phase input of the second coupling circuit is connected to the drain of the NMOS transistor M2 and the gate of the NMOS transistor M1. The out-of-phase input of the second coupling circuit is connected to the drain of the NMOS transistor M1 and the gate of the NMOS transistor M2. The in-phase output of the second coupling circuit is connected to the drain of the NMOS transistor M6 and the gate of the NMOS transistor M5. The out-of-phase output of the second coupling circuit is connected to the drain of the NMOS transistor M6 and the gate of the NMOS transistor M5.
[0024] The structure as shown generates two oscillatory signals, one between the drains of the NMOS transistors M1 and M2 and one between the drains of the NMOS transistors M5 and M6. The two oscillatory signals are in quadrature or 90° out of phase. The quadrature oscillator as described is subject to the phase noise problems as above-described.
[0025] “Design Issues In CMOS Differential LC Oscillators,” Hajimiri, A., Lee, T. H., IEEE Journal of Solid-State Circuits, pp. 717-724, May 1999 Vol. 34 Issue No. 5, presents an analysis of phase noise in differential cross-coupled inductance-capacitance (LC) oscillators. The effect of tail current and tank power dissipation on the voltage amplitude is shown. Various noise sources in the complementary cross-coupled pair are identified, and their effect on phase noise is analyzed.
[0026] “Phase Noise In CMOS Differential LC Oscillators”, Hajimiri, A., Lee, T. H., Digest of Technical Papers —1998 Symposium on VLSI Circuits, 1998, pp. 48-51, describes an analysis of phase noise in differential cross-coupled tuned tank voltage controlled oscillators. The effect of active device noise sources as well as the noise due to the passive elements is taken into account.
[0027] U.S. Pat. No. 5,475,345 (Gabara) teaches a CMOS coupled-tank oscillator having two inverters coupled, input-to-output, by inductances that may be simply wires, and a capacitance acting in parallel with each inverter that may be, simply, the inverter's gate capacitance.
[0028] U.S. Pat. No. 5,850,163 (Drost, et al.) discusses an active inductor oscillator with wide frequency range. The active inductor oscillator includes a tank circuit, buffer and integrating circuit that use differential transistor pairs that reduce phase jitter due to external common-mode noise sources.
[0029] U.S. Pat. No. 5,959,504 (Wang) describes a voltage controlled oscillator CMOS circuit using back gate terminals of CMOS transistors to vary the parasitic capacitances of the transistors. The back gate terminals receive a signal from a variable voltage source so that oscillation can be controlled by adjusting the variable voltage.
[0030] “A Low-Noise, 900-MHz VCO in 0.6-um CMOS” (Park, et al), IEEE Journal Of Solid-State Circuits, Vol. 34, pp. 586-591, May 1999, Issue No. 5, describes a low-noise, 900-MHz, voltage controlled oscillator (VCO) fabricated in a 0.6-um CMOS technology. The VCO consists of four-stage fully differential delay cells performing full switching. It utilizes dual-delay path techniques to achieve high oscillation frequency and obtain a wide tuning range.
[0031] “10 MHz CMOS OTA-C Voltage-Controlled Quadrature Oscillator,” Linares-Barranco, et al., IEEE Electronics Letters, June 1989, pp. 765-767, Vol. 25, Issue No. 12, details a quadrature-type voltage-controlled oscillator with operational transconductance amplifiers and capacitors (OTA-C).
[0032] “RC Sequence Asymmetric Polyphase Networks for RF Integrated Transceivers,” Galal et al, Transactions On Circuits And Systems —II: Analog And Digital Signal Processing, January 2000, pp. VOL 47, Issue No. 1, describes Resistance-Capacitance (RC) sequence asymmetric polyphase networks. A sequence of asymmetric polyphase networks provide the generation of highly matched wide-band quadrature signals which are immune to components mismatch, and suppression of the image signals without the need for highly selective RF filters and without employing image-reject mixing techniques.
[0033] U.S. Pat. No. 5,714,911 (Gilbert) describes a quadrature oscillator that includes an amplitude control circuit. The amplitude control circuit is that is based upon the trigonometric identity sin2(Ωt)+cos2(Ωt)=1. The amplitude control circuit, referred to as a Pythagorator, includes two squaring circuits. Each squaring circuit receives a respective quadrature oscillator signal and squares it. The outputs of the two squaring circuits are joined together so as to sum the outputs of the two squaring circuits to produce a sum of squares signal. This signal, a current in the preferred embodiment, is provided to damping diodes coupled to the outputs of the quadrature oscillator. The damping diodes produce a shunt positive resistance at the outputs of the quadrature oscillator in response to this current that has the effect of canceling the shunt negative resistance of the regenerative elements of the oscillator thereby establishing the amplitude of the quadrature oscillator signals at a desired amplitude.
[0034] U.S. Pat. No. 5,949,295 (Schmidt) teaches an integratable tunable resonant circuit for use in filters and oscillators. The circuit incorporates differential amplifier stage with a pair of differentially connected transistors with two negative feedback resistors. The two negative feedback resistors increase the linearity range of an input voltage of the differential amplifier stage.
[0035] U.S. Pat. No. 6,008,701 (Gilbert) details a quadrature oscillator using inherent nonlinearities of impedance cells to limit amplitude. The quadrature oscillator based on two cross-coupled integrator cells utilizing the inherent nonlinearity of positive and negative impedance cells to control the amplitude of oscillation. The oscillator is simplified thus eliminating the need for an outer control loop. A negative impedance cell is coupled to each integrator cell for assuring proper start-up and enhancing the amplitude of oscillation. A positive impedance cell is also coupled to each integrator cell to dampen the amplitude of oscillation. The transconductance of each impedance cell varies in response to the bias current provided to the cell. Thus, by controlling the bias currents through the cells, the negative and positive impedances seen by each integrator cell can made to cancel at the desired oscillation amplitude, so that the circuit oscillates without any damping or enhancement. By utilizing the inherent nonlinearity of positive and negative impedance cells, the bias currents provided to the impedance cells can remain fixed for a given frequency of operation, thereby simplifying the design of the oscillator and providing precise, robust control.
Summary of the Invention
[0036] An object of this invention is to provide a cross-coupled differential MOS oscillator.
[0037] Another object of this invention is to provide a cross-coupled differential MOS oscillator having reduced phase noise.
[0038] Another object of this invention is to provide a RF communication device, e.g., a transmitter or receiver, having a cross-coupled differential MOS oscillator.
[0039] To accomplish these and other objects, an oscillator having low phase noise that is formed of a frequency dependent amplifier to amplify a signal having a fundamental frequency; a frequency dependent feedback device that is connected between an output of the frequency dependent amplifier and an input of the frequency dependent amplifier to feed a portion of an amplified signal having the fundamental frequency to an input of the frequency dependent amplifier to stimulate oscillation; and a attenuating device in communication with the frequency dependent amplifier. The attenuating device reduces the gain of the frequency dependent amplifier for signals having frequencies much, much less than the fundamental frequency to decrease the phase noise.
[0040] The frequency dependent amplifier has an amplifying means. The amplifying means has an input and an output, whereby a signal at the input is amplified by a gain factor to form a signal at the output. The frequency dependent amplifier further, has a frequency dependent gain determining in communication with the amplifying means. The frequency dependent gain determining impedance determines the frequency at which the maximum gain of the frequency dependent amplifier occurs.
[0041] The amplifying means is composed of a pair of cross-coupled MOS transistors. The drain of each MOS transistor is connected to a gate of the other MOS transistor and to a port of the frequency dependent gain determining impedance. A first current source is connected to a source of one of the MOS transistors and to a ground reference point and to a first port of the attenuating device. A second current source is connected to a source of the other MOS transistor and to a second port of the attenuating device.
[0042] The frequency dependent determining impedance is formed by at least one inductor in communication with the amplifying means and a power supply voltage source, and at least one capacitor in communication with the amplifying means and a ground reference point.
[0043] The attenuating device is in the preferred embodiment, a capacitor in communication with the sources of the cross-coupled MOS transistors. The value of the capacitor is selected such that the fundamental frequency of oscillation is from approximately 10 times to approximately 20 times the high pass bandwidth of the cross-coupled MOS transistors.
[0044] Alternately, the amplifying means is formed of a cross-coupled pair of MOS transistors of the first conductivity type and a cross-coupled pair of MOS transistors of the second conductivity type to form a CMOS amplifying means. The drain of each MOS transistor of the first conductivity type is connected to a gate of the other MOS transistor of the first conductivity type and to a port of the frequency dependent gain determining impedance. A first current source is connected to a source of one of the MOS transistors of the first conductivity type and to a first port of the attenuating device, and a second current source is connected to a source of the other MOS transistor of the first conductivity type and to a second port of the attenuating device.
[0045] The drain of each MOS transistor of the second conductivity type is connected to a gate of the other MOS transistor of the second conductivity type and to one port of the frequency dependent gain determining impedance. A third current source in communication with a source of one of the MOS transistors of the second conductivity type and to a third port of the attenuating device, and a fourth current source in communication with a source of the other MOS transistor of the second conductivity type and to a fourth port of the attenuating device.
[0046] The attenuating device in the CMOS embodiment of the amplifying means is composed of a first capacitor connected from the first port to the second port of the gain attenuating means and a second capacitor in communication with the third and fourth ports of the gain attenuating means.
[0047] An application of the cross-coupled differential MOS oscillator is as the carrier oscillator of an RF transmitter. Alternately, the cross-coupled differential MOS oscillator is the local oscillator of an RF receiver that is used to demodulate the incoming RF signal.
Brief Description of the Drawings
[0048] FIG. 1 is a system block diagram of a frequency dependent system with feedback of the prior art.
[0049] FIG. 2 is a schematic diagram of a cross-coupled differential NMOS oscillator of the prior art.
[0050] FIG. 3 is a schematic diagram of a cross-coupled differential CMOS oscillator of the prior art.
[0051] FIG. 4 is a schematic diagram of a quadrature oscillator of the prior art.
[0052] FIGS. 5 a and 5b are schematic diagrams of two embodiments of cross-coupled differential MOS oscillators of this invention.
[0053] FIGS. 6 a and 6b are schematic diagrams of the cross-coupled differential MOS oscillator of this invention (FIG. 5 a) operating at low frequencies (FIG. 6 a) and at high frequencies (FIG. 6 b).
[0054] FIG. 7 is a schematic diagram of a cross-coupled differential CMOS oscillator of this invention.
[0055] FIG. 8 is a schematic diagram of a quadrature oscillator having low phase noise of this invention having.
[0056] FIG. 9 is a block diagram of a multiple frequency transforming circuit having low phase noise of this invention.
[0057] FIG. 10 is a schematic diagram of a differential amplifier having low phase noise of this invention.
[0058] FIG. 11 is a plot of the spectral density of the phase noise versus the frequency offset from the fundamental frequency.
[0059] FIG. 12 a is a schematic diagram of an ideal current source implemented by a biased MOSFET.
[0060] FIG. 12 b is a schematic diagram of a current source of FIG. 12 a having the noise component represented as a voltage source.
[0061] FIG. 12 c is a schematic diagram of a current source of FIG. 12 a having the noise component represented as a parallel current source.
[0062] FIG. 13 a is a current source implemented as a programmable resistance.
[0063] FIG. 13 b is an example of a programmable resistance of FIG. 13 a.
[0064] FIG. 14 a is a current source implemented as an inductance and programmable resistance.
[0065] FIG. 14 b is an example of a programmable resistance of FIG. 14 a.
[0066] FIG. 14 c is an example of a programmable inductance/resistance of FIG. 14 a.
Detailed Description of the Invention
[0067] Refer now to FIG. 5 a for a discussion of the cross-coupled differential NMOS oscillator having low phase noise of this invention. The frequency dependent gain amplifier is formed by the NMOS transistors M1 and M2 and the constant current sources I1 and I2. The frequency dependent gain determining impedance is formed by the inductors L1 and L2 and the capacitors C1 and C2.
[0068] The inductor L1 is connected from the drain of the NMOS transistor M1 to the reference voltage source Vcc and the inductor L2 is connected from the drain of the NMOS transistor M2 to the reference voltage source Vcc. The capacitor C1 is connected from the drain of NMOS transistor M1 to the ground reference point and the capacitor C2 is connected from the drain of the NMOS transistor M2 to the ground reference point. It is apparent to those skilled in the art that, while the capacitors C1 and C2 are connected to the ground reference point, the capacitors C1 and C2 may be connected to any reference voltage source or to any power supply voltage source and not effect the operation of the oscillator as explained above.
[0069] The fundamental frequency f0 of the cross-coupled differential oscillator of this invention is determined as:
-
- [0070] where:
- [0071] L1 is the value of the inductance of the inductor L1 or L2.
- [0072] C1 is the value of the capacitance of the capacitor C1 or C2.
- [0070] where:
[0073] The drain of the NMOS transistor M1 is connected to the gate of the NMOS transistor M2 and the drain of the NMOS transistor M2 is connected to the gate of the NMOS transistor M1. This cross-coupling of the drains to the gates of the NMOS transistors M1 and M2 forms the feedback circuit of the oscillator.
[0074] The source of the NMOS transistor M1 is connected to the constant current source I1 and the source of the NMOS transistor M2 is connected to the constant current source I2. The decoupling capacitor Cc is connected between the sources of the NMOS transistors M1 and M2 to act as a gain-attenuating device.
[0075] Refer now to FIGS. 6 a and 6b to understand the operation of the cross-coupled differential oscillator of this invention. The decoupling capacitor Cc is chosen to have very high impedance at frequencies much, much lower than the fundamental frequency f0 of the cross-coupled differential NMOS oscillator of FIG. 6 a. At frequencies much lower than the fundamental frequency f0, the cross-coupled differential NMOS oscillator of this invention functions as shown in FIG. 6 a. The current sources are separated and the gain of the frequency dependent gain circuit formed by the NMOS transistors M1 and M2 and the current sources I1 and I2 becomes much, much less than one, preventing the flicker noise or 1/f noise of the noise voltage sources Vn1 and Vn2 from being amplified and being added to the output signal of the cross-coupled differential NMOS oscillator of this invention.
[0076] At the fundamental frequency f0, the decoupling capacitor Cc is chosen to have an impedance that is very low. Thus, the cross-coupled differential NMOS oscillator of this invention functions as shown in FIG. 6 b. The frequency dependent gain circuit formed by the NMOS transistors M1 and M2 and the constant current sources I1 and I2 function as described in FIG. 2. The constant current sources I1 and I2 are summed together to form effectively one current source (I1+I2). Thus, the frequencies at the fundamental frequency f0 are amplified. The frequency dependent gain determining impedance formed by the inductors L1 and L2 and the capacitors C1 and C2 insure that the peak gain of the frequency dependent gain circuit is at the fundamental frequency f0 and that the higher and lower frequencies are attenuated.
[0077] The noise voltage sources Vn1 and Vn2 are, as described above, the models of the flicker or 1/f noise that is caused by the device characteristics of the NMOS transistors M1 and M2. The noise voltage sources Vn1 and Vn2 having frequency content that is much less than the fundamental frequency f0 and thus will be attenuated as shown in FIG. 6 a.
[0078] The high pass bandwidth (BW) of the cross-coupled differential oscillator is a function of the transconductance of the NMOS M1 and M2 and the value of the decoupling capacitor Cc and is determined by the formula:
[0079] The high pass bandwidth BW must be maintained at a level that is much, much smaller than the cutoff frequency of the cross-coupled differential oscillator to prevent loss of the fundamental frequency signal. The decoupling capacitor Cc should be chosen such that the fundamental frequency f0 of the cross-coupled differential oscillator is from approximately ten times to approximately twenty times the high pass bandwidth BW of the cross-coupled oscillator.
[0080] FIG. 5 b illustrates a second embodiment of a cross-coupled differential NMOS oscillator of this invention. The frequency dependent gain amplifier in this case is formed by the NMOS transistors M1 and M2 and the resistors R1 and R2. The resistor R1 is connected between the source of the NMOS transistor M1 and the ground reference point. The resistor R2 is connected between the source of the NMOS transistor M2 and the ground reference point.
[0081] The inductor L1 is connected from the drain of the NMOS transistor M1 to the reference voltage source Vcc and the inductor L2 is connected from the drain of the NMOS transistor M2 to the reference voltage source Vcc. The capacitor C1 is connected from the drain of NMOS transistor M1 to the ground reference point and the capacitor C2 is connected from the drain of the NMOS transistor M2 to the ground reference point. As described above, it is apparent to those skilled in the art that, while the capacitors C1 and C2 are connected to the ground reference point, the capacitors C1 and C2 may be connected to any reference voltage source or to any power supply voltage source and not effect the operation of the oscillator.
[0082] The decoupling capacitor Cc2 is connected between the sources of the NMOS transistors M1 and M2 and acts as gain attenuating device as above-described.
[0083] A third embodiment of this invention, as shown in FIG. 7 , implements the frequency dependent gain circuit as a cross-coupled differential CMOS amplifier. The frequency dependent gain circuit is formed by the NMOS transistors M1 and M2, the P-type MOS (PMOS) transistors M3 and M4, and the current sources I1, I2, I3, and I4.
[0084] The drain of the NMOS transistor M1 is connected to the gate of the NMOS transistor M2 and the drain of the NMOS transistor M2 is connected to the gate of the NMOS transistor M1. Similarly, the drain of the PMOS transistor M3 is connected to the gate of the PMOS transistor M4 and the drain of the PMOS transistor M4 is connected to the gate of the PMOS transistor M3. The cross-coupling of the drains and gates of the NMOS transistors M1 and M2 and the PMOS transistors M3 and M4 forms the feedback circuit of the oscillator.
[0085] The inductor L1 is connected between the drains of the NMOS and PMOS transistors M1 and M3 and the reference voltage source VCT. The inductor L2 is connected between the drains of the NMOS and PMOS transistors M2 and M4 and the reference voltage source VCT. The capacitor C1 is connected between the drains of the NMOS and PMOS transistors M1 and M3 and the ground reference point. The capacitor C2 is connected between the drains of the NMOS and PMOS transistors M2 and M4 and the ground reference point. Again, as described above, it is apparent to those skilled in the art that, while the capacitors C1 and C2 are connected to the ground reference point, the capacitors C1 and C2 may be connected to any reference voltage source or to any power supply voltage source and not effect the operation of the oscillator.
[0086] The inductors L1 and L2 and the capacitors C2 and C2 form the frequency dependent gain determining impedance.
[0087] The constant current source I1 is connected to the source of the NMOS transistor M1, and the constant current source I2 is connected to the source of the NMOS transistor M2. Similarly, the constant current source I3 is connected to the source of the PMOS transistor M3 and the constant current source I4 is connected to the source of the PMOS transistor M4.
[0088] The gain-attenuating circuit is formed by the decoupling capacitors Cc3 and Cc4. The decoupling capacitor Cc3 is connected between the sources of the NMOS transistors M1 and M2. The decoupling capacitor Cc4 is connected between the sources of the PMOS transistors M3 and M4.
[0089] The gain-attenuating circuit (Cc3 and Cc4) functions much as described in FIGS. 6 a and 6b. For frequencies much, much less than the fundamental frequency f0, the decoupling capacitors Cc3 and Cc4 have a large impedance and force the gain of the frequency dependent gain circuit to a level much, much less than one to attenuate the low frequency flicker or 1/f noise. Conversely, for frequencies equal to the fundamental frequency fo, the decoupling capacitors Cc3 and Cc4 have low impedance and the frequency dependent gain circuit functions equivalently to that as described in FIG. 3. The constant current sources I1 and I2 are summed as described in FIG. 5 b and, similarly, the constant current sources I3 and I4 are summed together to function equivalently to the description of FIG. 3.
[0090] The high pass bandwidth (BW) of the cross-coupled differential oscillator is a function of the transconductance of the NMOS M1 and M2 and the value of the decoupling capacitor Cc3 and the transconductance of the PMOS transistors M3 and M4 and the value of the decoupling capacitor Cc4 and is determined by the formula:
[0091] The high pass bandwidth BW must be maintained, as described above, at a level that is much, much smaller than the cutoff frequency of the cross-coupled differential oscillator to prevent loss of the fundamental frequency signal f0. The decoupling capacitor Cc should be chosen such that the fundamental frequency f0 of the cross-coupled differential oscillator is from approximately ten times to approximately twenty times the high pass bandwidth BW of the cross-coupled oscillator.
[0092] FIG. 8 illustrates a quadrature oscillator having low phase noise of this invention. The cross-coupled differential oscillators OSC1 and OSC2 are structured and function as cross-coupled differential oscillators as described in FIG. 5 a. The NMOS transistors M3 and M4 and the current sources I3 and I4 form a first coupling circuit. The current source I3 is connected between the source of the NMOS transistor M3 and the ground reference point. The current source I4 is connected between the source of the NMOS transistor M4 and the ground reference point. The gate of the NMOS transistor M3 functions as the in-phase input of the first coupling circuit and the gate of the NMOS transistor M4 functions as the out-of-phase input of the first coupling circuit. The drain of the NMOS transistor M4 functions as the in-phase output of the first coupling circuit and the drain of the NMOS transistor M3 functions as the out-of-phase output of the first coupling circuit. The decoupling capacitor Cc6 is connected between the sources of the NMOS transistors M3 and M4. The decoupling capacitor Cc6 is chosen to function similar to the decoupling capacitor Cc of FIG. 5 a to eliminate the phase noise from the first coupling circuit.
[0093] The in-phase input of the first coupling circuit is connected to the drain of the NMOS transistor M5 and the gate of the NMOS transistor M6 of the second cross-coupled differential oscillator OSC2. The out-of-phase input of the first coupling circuit is connected to the drain of the NMOS transistor M6 and the gate of the NMOS transistor M5 of the second cross-coupled differential oscillator OSC2. The in-phase output of the first coupling circuit is connected to the drain of the NMOS transistor M2 and the gate of the NMOS transistor M1 of the first cross-coupled differential oscillator OSC1. The out-of-phase output of the first coupling circuit is connected to the drain of the NMOS transistor M1 and the gate of the NMOS transistor M2 of the first cross-coupled differential oscillator OSC1.
[0094] The NMOS transistors M7 and M8 and the current sources I7 and I8 form the second coupling circuit. The current source I7 is connected between the source of the NMOS transistor M7 and the ground reference point. The current source I8 is connected between the source of the NMOS transistor M8 and the ground reference point. The gate of the NMOS transistor M7 functions as the in-phase input of the second coupling circuit and the gate of the NMOS transistor M4 functions as the out-of-phase input of the second coupling circuit. The drain of the NMOS transistor M7 functions as the in-phase output of the second coupling circuit and the drain of the NMOS transistor M8 functions as the out-of-phase output of the second coupling circuit. The decoupling capacitor Cc8 is connected between the sources of the NMOS transistors M7 and M8. The decoupling capacitor Cc8 is chosen to function similar to the decoupling capacitor Cc of FIG. 5 a to eliminate the phase noise from the first coupling circuit.
[0095] The in-phase input of the second coupling circuit is connected to the drain of the NMOS transistor M1 and the gate of the NMOS transistor M2 of the first cross-coupled differential oscillator OSC1. The out-of-phase input of the second coupling circuit is connected to the drain of the NMOS transistor M2 and the gate of the NMOS transistor M1 of the first cross-coupled differential oscillator OSC1. The in-phase output of the second coupling circuit is connected to the drain of the NMOS transistor M6 and the gate of the NMOS transistor M5 of the second cross-coupled differential oscillator OSC2. The out-of-phase output of the second coupling circuit is connected to the drain of the NMOS transistor M5 and the gate of the NMOS transistor M6 of the second cross-coupled differential oscillator OSC2.
[0096] The in-phase and the out-of-phase of the first coupling circuit are transposed relative to the similar in-phase and out-of-phase connections of the second coupling circuit. This transposition is to force the necessary phase shift to cause the cross-coupled differential oscillators OSC1 and OSC2 to oscillate in quadrature or 90° out of phase as described above in Razavi.
[0097] The structure of the oscillator of FIG. 8 is generalized to a structure as shown in FIG. 9. This circuit is used to create multiple phased oscillators, mixers, modulators, demodulators, and any circuit requiring the transforming of the an input signal with multiple frequencies. The frequency transforming circuit of FIG. 9 has multiple coupling elements CE1, CE2, . . . , CEn that are serially connected output to input. The frequency transforming circuit, further, has multiple cross-coupled differential oscillators OSC1, OSC2, . . . , OSCn. The output of each of the multiple cross-coupled differential oscillators OSC1, OSC2, . . . , OSCn is connected to an input of one of the coupling elements coupling elements CE1, CE2, . . . , CEn.
[0098] The input signal is developed between the input terminals IN+ and IN− and is transferred to the first coupling element CE1. The input signal is then combined with the oscillatory signal from the first cross-coupled differential oscillator OSC1. The signal at the output of the first coupling element CE1 is transferred to the input the second coupling element CE2 where it is combined with the second oscillatory signal from the second cross-coupled differential oscillator OSC2. The signal at the output of the second coupling element CE2 is transferred to the following coupling elements CEn for combination with the oscillatory signals from the subsequent oscillators OSCn. The signal from the final coupling element CEn is transferred to subsequent circuitry. In the alternative, the output of the last coupling element CEn maybe connected to the input of the first coupling element CE1 to feedback the output signal (or a portion of the output signal) to the input of the circuit.
[0099] The coupling elements coupling elements CE1, CE2, . . . , CEn, in addition to combining the oscillatory signals from the multiple cross-coupled differential oscillators OSC1, OSC2, . . . , OSCn, may provide phase shifting for a multiple phased oscillator, or any appropriate filtering, integrating, differentiating function.
[0100] Further, the outputs of each of the coupling elements CE1, CE2, . . . , CEn is connected to an input of a buffering amplifier BUF1, BUF2, . . . , BUFn. Each of the a buffering amplifiers BUF1, BUF2, . . . , BUFn capture the output of one of the coupling elements CE1, CE2, . . . , CEn and amplifies and isolates the signal to form the output signals Φ1, Φ2, . . . , Φn that are transferred to external circuitry.
[0101] Each cross-coupled differential oscillators OSC1, OSC2, . . . , OSCn, each coupling element CE1, CE2, . . . , CEn, and each buffering amplifier BUF1, BUF2, . . . , BUFn has a differential amplifier with low phase noise of this invention as shown in FIG. 10. The differential amplifier is formed by the NMOS transistors M1 and M2 and the current sources I1 and I2.
[0102] The gates of the NMOS transistors M1 and M2 respectively form the in-phase input IN+ and the out-of-phase input IN−. The drains of the NMOS transistors M1 and M2 respectively form the in-phase output OUT+ and the out-of-phase output OUT−.
[0103] The current source I1 is connected between the source of the NMOS transistor M1 and the ground reference point. The current source I2 is connected between the source of the NMOS transistor M2 and the ground reference point.
[0104] The decoupling capacitor Cc is connected between the sources of the NMOS transistors M1 and M2 to provide the necessary gain attenuating to eliminate the phase noise. When the differential amplifier is operating at sufficiently high frequency, the impedance of the decoupling capacitor Cc is very low and the current sources I1 and I2 combine. The differential amplifier operates as a true differential amplifier having very high gain. However, if the frequency of operation is sufficiently low, the impedance of the decoupling capacitor Cc is very high and the gain of the differential amplifier is very low, thus attenuating the signals of the phase noise.
[0105] The high pass bandwidth BW of the differential amplifier of this invention is a function of the transconductance (gm) of the NMOS transistors “looking” into the sources and is determined by the formula:
For the most successful operation of the differential amplifier the decoupling capacitor Cc should be chosen such that the fundamental frequency f0 of the cross-coupled differential oscillator is from approximately ten times to approximately twenty times the high pass bandwidth BW of the cross-coupled oscillator. This insures that the fundamental frequency f0 is not affected by the operation of the decoupling capacitor Cc.
For the most successful operation of the differential amplifier the decoupling capacitor Cc should be chosen such that the fundamental frequency f0 of the cross-coupled differential oscillator is from approximately ten times to approximately twenty times the high pass bandwidth BW of the cross-coupled oscillator. This insures that the fundamental frequency f0 is not affected by the operation of the decoupling capacitor Cc.
[0106] FIG. 11 is plots 700 and 750 that illustrate the spectral density of the phase noise of the output signal versus the frequency offset from the fundamental frequency f0 of cross-coupled differential oscillators of this invention 700 and the prior art. As can be seen, the spectral density of the phase noise of the cross-coupled differential NMOS transistor is lower than an equivalent design of the prior art.
[0107] FIG. 12 a is an example of an ideal current source utilized by the present invention. In FIG. 12 a the ideal current source is implemented as a MOS transistor which is biased so that the MOS transistor operates in the saturation region. Such a current source may generate a 1/f noise component, which can be significant in MOS devices. This problem is exacerbated at higher frequencies, in which the oscillator of the present invention is designed to operate. Additionally, as the device geometry becomes small the 1/f noise becomes more pronounced. FIG. 12 c illustrates an equivalent representation showing a current source I and a noise component current source Inoise.
[0108] A conventional solution to reduce or eliminate the 1/f noise is to utilize a resistor as the current source. However, it is difficult to set the appropriate amount of resistance for the oscillator to function properly. In accordance with an embodiment of the present invention a programmable resistance R is utilized as the current source, as shown in FIG. 13 a. The programmable resistance can insure the appropriate amount of resistance to provide the current to the oscillator. The programmable resistance may be implemented as a switched resistor array. One example of the resistor array is shown in FIG. 13 b. The resistor array shown therein comprises resistors R1-Rn and associated switches S1-Sn. Of course as will be appreciated by one of ordinary skill in the art, other resistor configurations may be employed and are within the scope and spirit of the present invention.
[0109] An alternative embodiment of the current source in accordance with the present invention is to utilize an inductance L in series with a programmable resistance R, as shown in FIG. 14 a. As with the previous embodiment there is no 1/f noise, since the inductance and resistance are passive components. This configuration behaves like a constant current source regardless of the input voltage, especially if the inductance is sufficiently high, at high frequencies the current is essentially constant (due to the inductance properties). In this embodiment the programmable resistance may be implemented as a switched resistor array. One example shown therein comprises resistors R1-Rn and associated switches S1-Sn. Of course as will be appreciated by one of ordinary skill in the art, other resistor configurations may be employed and are within the scope and spirit of the present invention. The inductance L inherently has some resistance. Accordingly, the inductance and programmable resistance may be alternatively be implement by a switched inductance array, wherein each inductance inherently has the appropriate amount of resistance.
[0110] It will be apparent to those skilled in the art that the NMOS transistors M1 and M2 of FIG. 5 a can be replaced by PMOS transistors with appropriate changes to the power supply voltage source Vcc and the ground reference point. Further, it would be apparent that the NMOS transistors could be replaced by bipolar junction transistors or other field effect transistors constructed of materials such as Galium-arsenide and still be in keeping with this invention.
[0111] While this invention has been particularly shown and described with reference to the preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made without departing from the spirit and scope of the invention.
Claims
Certificate of Correction present
The claims shown below may not include correction changes. Use the corrected PDF for authoritative text.
What is claimed is:
1. A method for designing a MOS oscillator, comprising:
arranging a frequency dependent amplifier and a frequency dependent feedback device, such that the frequency dependent feedback device is in communication with an output and an input of the frequency dependent amplifier; and
designing an attenuating device to be in communication with the frequency dependent amplifier to attenuate flicker noise, wherein said attenuating device has a characteristic such that the fundamental frequency is from approximately ten times to twenty times a high pass bandwidth of a combination of the frequency dependent amplifier and the attenuating device.
2. The method of claim 1, wherein said frequency dependent amplifier is arranged to amplify an input by a predetermined gain factor, wherein said frequency dependent feedback device is arranged to include a frequency dependent gain determining impedance in communication with the frequency dependent amplifier, and wherein a maximum gain of said frequency dependent amplifier occurs at the fundamental frequency.
3. The method of claim 2, wherein the frequency dependent amplifier is designed to include,
a pair of cross-coupled MOS transistors, a drain of each of the pair of cross-coupled MOS transistors being in communication with a gate of the other of the pair of cross-coupled MOS transistors and to a corresponding terminal of said frequency dependent gain determining impedance;
a first current source having a first terminal in communication with a source of a first one of said pair of cross-coupled MOS transistors and to a first terminal of the attenuating device; and
a second current source having a first terminal in communication with a source of a second one of said pair of cross-coupled MOS transistors and to a second terminal of the attenuating device.
4. The method of claim 2, wherein said frequency dependent gain determining impedance is designed to include,
at least one inductor in communication with the frequency dependent amplifier and a first terminal of a voltage source; and
at least one capacitor in communication with the frequency dependent amplifier and a second terminal of the voltage source.
5. The method of claim 3, wherein the attenuating device is designed to include a capacitor.
6. A method for designing an RF communication device, comprising:
designing an MOS oscillator, having a fundamental frequency, the MOS oscillator including a frequency dependent amplifier and a frequency dependent feedback device, wherein the frequency dependent feedback device is in communication with an output and an input of the frequency dependent amplifier; and
designing an attenuating device to be in communication with the frequency dependent amplifier to attenuate flicker noise signals having a frequency much less than the fundamental frequency, wherein said attenuating device has a characteristic such that the fundamental frequency is from approximately ten times to twenty times a high pass bandwidth of a combination of the frequency dependent amplifier and the attenuating device.
7. A method for designing an RF communication device, comprising:
designing an MOS oscillator, having a fundamental frequency, the MOS oscillator including a frequency dependent amplifier and a frequency dependent feedback device, wherein the frequency dependent feedback device is in communication with an output and an input of the frequency dependent amplifier; and
designing an attenuating device to be in communication with the frequency dependent amplifier to attenuate flicker noise signals having a frequency much less than the fundamental frequency, wherein the RF communication device is designed to include an RF receiver and the oscillator is designed to include a local oscillator to demodulate a carrier frequency signal received by the RF receiver.
8. A method for designing an RF communication device, comprising:
designing an MOS oscillator, having a fundamental frequency, the MOS oscillator including a frequency dependent amplifier and a frequency dependent feedback device, wherein the frequency dependent feedback device is in communication with an output and an input of the frequency dependent amplifier; and
designing an attenuating device to be in communication with the frequency dependent amplifier to attenuate flicker noise signals having a frequency much less than the fundamental frequency, wherein the frequency dependent amplifier is designed to amplify an input signal by a predetermined gain factor, wherein the frequency dependent feedback device is designed to include a frequency dependent gain determining impedance in communication with the frequency dependent amplifier, and wherein a maximum gain of the frequency dependent amplifier occurs at the fundamental frequency.
9. The method of claim 8, wherein the frequency dependent amplifier is designed to include,
a pair of cross-coupled MOS transistors, a drain of each of the pair of cross-coupled MOS transistors being in communication with a gate of the other of the pair of cross-coupled MOS transistors and to a corresponding terminal of the frequency dependent gain determining impedance;
a first current source having a first terminal in communication with a source of a first one of the pair of cross-coupled MOS transistors and with a first terminal of the attenuating device; and
a second current source having a first terminal in communication with a source of a second one of said pair of cross-coupled MOS transistors and with a second terminal of the attenuating device.
10. The method of claim 8, wherein the frequency dependent gain determining impedance is designed to include,
at least one inductor in communication with the amplifier and a first terminal of a voltage source; and
at least one capacitor in communication with the amplifier and a second terminal of the voltage source.
11. The method of claim 8, wherein the attenuating device is designed to include a capacitor.
12. A method for designing an RF communication device, comprising:
designing an MOS oscillator, having a fundamental frequency, the MOS oscillator including a frequency dependent amplifier and a frequency dependent feedback device, wherein the frequency dependent feedback device is in communication with an output and an input of the frequency dependent amplifier; and designing an attenuating device to be in communication with the frequency dependent amplifier to attenuate flicker noise signals having a frequency much less than the fundamental frequency, wherein frequency dependent amplifier is designed to include;
a first pair of cross-coupled MOS transistors of a first conductivity type, a drain of each of the first pair of cross-coupled MOS transistors being in communication with a gate of the other of the first pair of cross-coupled MOS transistors and to a corresponding terminal of the frequency dependent gain determining impedance;
a first current source in communication with a source of one of the first pair of cross-coupled MOS transistors of the first conductivity type and with a first terminal of the attenuating device;
a second current source in communication with a source of a second one of the first pair of cross-coupled MOS transistors of the first conductivity type and with a second terminal of the attenuating device;
a second pair of cross-coupled MOS transistors of a second conductivity type wherein a drain of each of the second pair of cross-coupled MOS transistors is connected to a gate of the other of the second pair of cross-coupled MOS transistors and to one terminal of said frequency dependent gain determining impedance;
a third current source in communication with a source of one of the second pair of cross-coupled MOS transistors and with a first terminal of the second attenuating device; and
a fourth current source in communication with a source of the other of the second pair of cross-coupled MOS transistors of the second conductivity type and to a second terminal of the second attenuating device.
13. The method of claim 6, wherein the attenuating device includes a first capacitor.
14. The method of claim 7, wherein the attenuating device includes a first capacitor.
15. The method of claim 8, wherein the attenuating device includes a first capacitor.
Patent Citations (7)
Non-Patent Literature (8)
- “10MHz CMOS OTA-C Voltage-Controlled Quadrature Oscillator”, Linares-Barranco et al., IEEE Electronics Letter, Jun. 1989, pp. 765-767, vol. 25, Issue No. 12.Applicant
- RC Sequence Asymmetric Polyphase Networks for RF Intergrate Transceivers, Galal et al., Transaction On Circuits and Systems-11: Analog and Digital Signal Processing, Jan. 2000, vol. 47, Issue No. 1, pp. 18-27.Applicant
- Phase Noise In CMOS Differential LC Oscillators, Hajimiri, A. et al., Digest of Technical Papers—1998 Symposium on VLSI Circuits, 1998, pp. 48-51.Applicant
- A Low-Noise, 900 MHz VCO in 6-um CMOS (Park et al.), IEEE Journal of Solid-State Circuits, vol. 34, pp. 586-591, May 1999, Issue No. 5.Applicant
- A 1.8 Ghz CMOS Voltage-Controlled Oscillator, B. Razavi, Digest of Technical Papers., 43rd ISSCC, 1997, pp. 388-389, and Fig. 4.Applicant
- “Design Issues in CMOS Differential LC Oscillators”, Hajimiri, A. et al. IEEE Journal of Solid-State Circuits, pp. 717-724, May 1999, vol. 34, Issue No. 5.Applicant
- B. Linares-Barranco et al., “10 MHz CMOS OTA-C Voltage-Controlled Quandrature Oscillator”, Jun. 8, 1989, vol. 25, No. 12, pp. 765-767.Applicant
- Sherif H. Galal et al., “RC Sequence Asymmetric Polyphase Networks for RF Integrated Transceivers”, IEEE Transactions on Circuits and Systems-II: Analog and Digital Signal Processing, Jan. 2000, vol. 47, No. 1.Applicant