Patent Grant
8223067
This patent application is being concurrently filed with commonly owned utility U.S. patent application Ser. No. 12/590,117, entitled “Method of Down Converting High-frequency Signals”, by Christopher E. Stewart, Grant E. Moulton, and Steven H. Goody; and with commonly owned utility U.S. patent application Ser. No. 12/590,116, entitled “Continuous-Wave Field Disturbance Sensing System”, by Christopher E. Stewart, Grant E. Moulton, and Steven H. Goody; and with commonly owned design U.S. patent application Ser. No. 29/316,807, entitled “Hand-Held Radar Device”, by Steven H. Goody, Joie L. Puckett, Jr., Grant E. Moulton, Christopher E. Stewart.
None.
None.
None.
This invention relates generally to electromagnetic field disturbance sensing techniques, such as radar systems.
Continuous Wave (CW) Coherent Radar uses frequency conversion to compare the phase of a transmitted signal with the reflection of that signal from a moving target. The phase of the wave reflected from the target changes as a function of the changing distance to that target. If the target velocity remains constant, the phase of the reflected signal changes at a constant rate. A constant rate of phase change corresponds to a constant frequency. Thus the returning reflected signal appears at a frequency offset from the transmitted signal that is proportional to the relative velocity between the transmitter and the target.
Comparing the transmitted signal and the received signal with a frequency downconverter delivers the difference frequency between the transmitted and received signals at the converter output. Practical implementation of a radar seeks to optimize the cost and size of the equipment required to compare the phase (or frequency) of the transmitted and received signals, while obtaining the greatest detection range to target possible for that cost and size.
Many conventional portable radar guns use a Gunn diode driving a cavity oscillator with an integral diode peak detector which functions as a frequency downconverter or mixer, using either one or multiple detector diodes. The cavity oscillator/mixer is coupled to a horn antenna used to transmit the incident signal and to receive the reflected signal. The cavity drives the diode detector with a local oscillator (“LO”) signal from the transmitter and couples the received RF signal to the same diode. The diode detector mixes the RF and LO signals, creating an IF signal at their difference frequency. The diode detector typically matches to a relatively high impedance, often hundreds or even thousands of ohms, and conversion loss can approach 0 dB. Matching to LO and RF signals is accomplished by moving the diode location within the cavity to optimize the coupling for optimal system performance.
The detector diode also rectifies the LO power in the cavity, and any variations in the amplitude due to either coherent amplitude modulation (“AM”) or to AM noise will show up at the IF output. Because of this problem, designers typically use Gunn diode oscillators adjusted to the point of minimal conversion of diode bias supply voltage input to amplitude variation. This minimizes the AM noise on the LO and thus also minimizes the detected LO AM noise on the IF output allowing for sufficiently sensitive RF detection.
The cavity based radar devices typically require a horn antenna up to several inches long and a cavity oscillator at least one or more cubic inches in size for operation at the 10 GHz or 24 GHz ISM bands (e.g., the X, and K bands). Both of these factors cause the system to have significant, weight and size, which is undesirable for a small hand-held application. Furthermore, the optimum Gunn diode bias point often requires substantial current draw, limiting the useful operating time for portable, battery-powered applications. Alternatively, the radar size must increase to accommodate larger batteries.
Another design approach to small sized radar devices uses planar or “patch” antenna arrays. These devices either use cavity stabilized Gunn oscillator/detectors or use traditional switching mixers where the LO signal switches the RF signal phase to the IF output dependent upon LO phase. The switching type of mixer typically shows 6 dB or more conversion loss, and must be a balanced configuration to cancel any AM noise of the local oscillator. Diodes used in conventional mixer-based systems act like switches that provide either an open circuit or a closed switch in a signal path. The LO signal drives the mixer diode(s) to turn the diode “on”, or low impedance, for about a half cycle and “off”, or high impedance, for the other half cycle.
The balanced or double or triple balanced switching diode mixer suffers from imperfect AM noise cancelation due to variations in manufacturing and remains sensitive to the AM noise of most oscillators. The down-converted local oscillator AM noise obscures the incoming RF signal, even while the local oscillator phase noise cancels due to the short time required for the round trip on the radar path or the path inside the mixer itself. Conventional (incoherent) receivers do not typically see the AM noise of the LO as the phase noise typically dominates the AM noise by tens of dB. Only in coherent reception (such as used for CW radar) does the phase noise of the LO cancel and allow the AM noise to dominate.
Additionally, the IF output of a switching diode mixer typically requires termination with a low noise IF amplifier with low input impedance, usually equal to 50 ohms. The noise voltage of that amplifier with 6 dB mixer loss is equivalent to twice that noise voltage measured at the antenna input. Diodes typically add another 0.5 to 1 dB to the input noise of the mixer above the conversion loss, further degrading the receive signal to noise ratio as seen at the antenna RF port. This type of radar does not typically deliver good long range performance compared with the Gunn and horn antenna alternatives without the addition of other components such as additional antennas or an RF preamplifier.
Other devices constructed using planar patch antenna arrays have used a Gunn-based cavity oscillator for the transmitter source and a detector diode for the receive mixer. These can provide reasonable AM noise from the Gunn source, but are limited in miniaturization by the size of the oscillator resonant cavity.
Components for radar systems and other applications overcoming the deficiencies of the prior art are desirable.
A detector system has a first detector configured to detect a first high-frequency signal having amplitude-modulated (“AM”) noise to produce a first detected signal having at least a first detected AM noise signal component and a demodulated signal component and a second detector configured to detect a second high-frequency signal having the AM noise to produce a second detected signal having at least a second detected AM noise signal component. An algebraic combining network combines the first detected signal and the second detected signal to cancel the first detected AM noise signal component with the second detected AM noise signal component to produce an output signal including the demodulated signal component.
In a further embodiment, the first high-frequency signal includes a local high-frequency signal and a received high-frequency signal, the demodulated signal component being at a difference frequency between the local high-frequency signal and the received high-frequency signal. In a particular embodiment, the second high-frequency signal is the local high-frequency signal. In an alternative embodiment, the second high-frequency signal includes the local high-frequency signal and the received high-frequency signal, and the second detected signal further includes a second demodulated signal component at the difference frequency. The output signal of the algebraic combining network can include the second demodulated signal component added to the demodulated signal component.
In some embodiments, the received high-frequency signal is a reflected signal of the local high-frequency signal.
A further embodiment includes a high-frequency algebraic combining network combining the local high-frequency signal and the received high-frequency signal and provides the local high-frequency signal and the received high-frequency signal to the first detector.
In a particular embodiment, the algebraic combining network has an inverting input and a non-inverting input, the first detected signal is provided to the non-inverting input and the second detected signal is provided to the inverting input.
In a particular embodiment, the first high-frequency signal includes a local high-frequency signal and a received high-frequency signal, the output signal being at a difference frequency.
In some embodiments, a detector system includes a high-frequency algebraic combining network configured to receive a local high-frequency signal and a received high-frequency signal. The high-frequency algebraic combining network provides the local high-frequency signal and the received high-frequency signal to the first detector, and provides the local high-frequency signal and an inverse of the received high-frequency signal to the second detector. The second detector produces the second detected signal having the second detected AM noise signal component and an inverse demodulated signal component.
In some embodiments, the algebraic combining network has an inverting input and a non-inverting input. The first detected signal is provided to the non-inverting input and the second detected signal is provided to the inverting input to produce an inverse AM noise signal component and a second demodulated signal component. The output signal is a sum of the demodulated signal component and the second demodulated signal component, and the inverse AM noise signal component cancels the AM noise signal component.
The detectors can be a single-diode detector, a dual-diode detector, a diode voltage multiplier, for example.
In a particular embodiment, the high-frequency algebraic combining network has a first received signal path through the high-frequency algebraic combining network, a second received signal path through the high-frequency algebraic combining network, a first LO signal path through the high-frequency algebraic combining network, and a second LO signal path through the high-frequency algebraic combining network. A first high-frequency combiner combines a received signal in the first received signal path and an LO signal in the first LO signal path, and provides a first combined signal to the first detector. A second high-frequency combiner combines the received signal in the second RF signal path and the LO signal in the second LO signal path and provides a second combined signal to the second detector. In a particular embodiment, the first combined signal is the LO signal plus the received signal and the second combined signal is the LO signal minus the received signal.
In a particular embodiment, a detector system includes a second algebraic combining network disposed between the first detector and the algebraic combining network. The second algebraic combining network has a first positive output connected to a first non-inverting input of the algebraic combining network, and a first negative output connected to a first inverting input of the algebraic combining network. In a further embodiment, the second detected signal includes a second demodulated signal component, and a third algebraic combining network disposed between the second detector and the algebraic combining network has a second positive output connected to a second inverting input of the algebraic combining network. A second negative output is connected to a second non-inverting input of the algebraic combining network.
In a further embodiment, an adjustable gain stage is between the algebraic combining network and one of the second algebraic combining network and the third algebraic combining network. The adjustable gain stage allows adjustment of one of the detected AM noise signals so that the second detected AM noise signal component is equal to the first detected AM noise signal times a gain factor. For example, the adjustable gain stage divides the second detected AM noise signal component by the gain factor.
Field disturbance sensing systems (e.g., radar systems, motion sensors, automatic door openers, automotive sensors, and low-IF Tx/Rx systems) according to embodiments achieve small size and improved range, sensitivity and signal-to-noise ratio. Systems according to some embodiments are mobile, battery-powered systems. Other embodiments are fixed installations with AC (mains) power or external power supplies. A noise-canceling down-converting detector according to an embodiment cancels the AM noise from a first signal (signal 1, which will be referred to as an “LO signal”, although it performs quite differently from an LO signal in a switching mixer system) or transmitter signal.
Some embodiments of detectors provide significant conversion gain when properly matched to the detector circuit compared to a conventional switching mixer-based system. In some embodiments, greater than 10 dB of conversion voltage gain has been measured. Improved conversion gain allows use of a simple, low cost IF amplifier as an alternative to the more expensive low noise IF or RF amplifiers used in conventional switched diode mixer systems because of the superior signal to noise ratio embodiments achieve. In some embodiments, low noise IF amplifiers are used to boost the converted signal with minimal additional noise. As used herein, “noise canceling” means AM noise detected at one detector is subtracted from essentially the same AM noise detected at another detector.
A noise canceling down-converting detector according to an embodiment detects an incoming (reflected or other RF) signal while canceling AM noise on the local (transmitted or LO) signal. The noise canceling down-converting detector does not operate as a traditional switching mixer in that it does not switch or gate the RF signal to an IF port with the LO signal controlling the switching action. Instead, the noise canceling down-converting detector operates as two or more separate envelope detectors.
The detectors detect the instantaneous value (voltage) of two high-frequency signals. One of these signals includes the average (steady state) LO power (generating an essentially DC detected output), the instantaneous RF power, which adds to or subtracts from the LO signal to produce a beat (IF) signal, and the instantaneous LO AM noise power, which can also add to or subtract from the average LO power, thus introducing inaccuracy into the peak detected voltage (potentially obscuring or adding uncertainty to the detected RF signal). The signal-of-interest (e.g., the reflected signal or a received signal from a transmitter) is demodulated into a component of the detected signal (“demodulated component”).
The other signal includes at least the average LO power and the instantaneous LO AM noise power (and optionally the RF signal or other signals). The detected steady-state LO signals are DC, and are rejected or otherwise canceled (e.g., do not affect an RC detector load). The detected AM noise components are subtracted from each other, thus canceling the effect of LO AM noise on the resultant detected (IF) signal. The RF signal, which in some embodiments is the reflected signal from a moving object, adds to or subtracts from the instantaneous LO signal at the detector and increases or decreases the peak voltage, and hence the detector output. A convenient visualization is that the reflected signal is “sliding past” the LO signal, and the detector produces an IF signal at a beat frequency in base band.
In conventional switching mixers, the IF signal is the sum or difference between two frequencies (i.e., the RF and LO signals). In other words, a conventional switching mixer can operate as an upconverter or a downconverter. Conventional switching mixers typically operate in a relatively low impedance system (e.g., a system with a characteristic impedance of fifty ohms) at all mixer ports; LO, RF and IF. Noise canceling down-converting detectors are not limited to systems with such low impedance at the IF port.
In a continuous-wave (“CW”) radar system, a high-frequency (LO) signal is sent to an antenna and transmitted to a target, which reflects a portion of the energy (the “reflected” or “RF” signal) that is received by the same antenna or by a second antenna. Relative motion between the target and antenna(s) is detected as a frequency shift between the reflected signal and the transmitted (LO) signal. The detected frequency shift is processed to determine the speed of the relative motion. The LO signal is typically much higher power than the reflected signal, and AM noise on the LO signal can obscure the RF component.
The outputs 114, 116 from the detectors 104, 112 are provided to a non-inverting input 117 and an inverting input 115 of a second algebraic combining network 118 to produce a detected output signal 120, which will be referred to as an intermediate frequency (“IF”) signal for purposes of convenient discussion. The second algebraic combining network 118 can be implemented in any of several types of circuits, such as an operational amplifier, differential amplifier, or a digital circuit technique incorporating analog-to-digital converters and subsequent digital signal processing in hardware or software. The second algebraic combining network combines the first detected signal and the second detected signal to cancel the detected AM noise. The first detected signal 116 includes a detected AM noise component (LOd) and a second detected signal component (RFd). The second detected signal 114 includes a detected AM noise component (LOd) nominally equal to the detected AM noise component in the first detected signal 116, thus, the detected AM noise cancels. The output signal (IF) 120 is the second detected signal component (RFd).
The IF signal of
The IF signal 120 equals the detected LO signal at input 117, minus the detected LO signal at input 115, plus the detected RF signal at input 117. The system is setup (e.g., matched design or calibrated) such that the output signals at 114 and 116 are equal in amplitude and phase for a given incident power at the LO port 102, thus canceling the LO AM noise on the signal output. In some systems, the amplitude of the LO signal is much higher than the RF signal and the LO AM noise can dominate the differences in the peak signal voltage arising from the RF signal. The system 100 avoids the problem of LO AM noise dominating the RF signal by subtracting the detected LO signal and associated LO AM noise from a detected copy of that same signal.
In a particular embodiment, the first and second detectors are single diode detectors. In alternative embodiments, the first and second detectors are multiple diode detectors or other types of detectors. In a particular embodiment the first and second detectors are substantially identical to each other so that the detected LO signals are substantially identical and provide good cancelation of the LO AM noise. Manufacturing tolerances can result in minor differences, and some embodiments include calibration techniques, as discussed below in reference to
The RF signal path (“first RF path”) 220 to the first detector 208 has a first associated phase delay Φ1 and a first gain (or loss, which will be expressed as negative gain (or −dB) for purposes of discussion) g1, and the RF signal path (“second RF path”) 222 to the second detector 210 has a second associated phase delay Φ2 and second gain g2. The LO signal path (“first LO path”) 224 to the first detector 208 has an associated phase shift Φ3 and gain g3, and the LO signal path (“second LO path”) 226 to the second, detector 210 has a phase shift Φ4 and gain g4.
The RF and LO signals are combined in high-frequency combiners 228, 230, and coupled to the detectors 208, 210. The detectors 208, 210 provide low-frequency signals to outputs 236, 238, and to the detected signal algebraic combining network 218. The detected signals are subject to phase delay Φ5, Φ6, and gain g5, g6, before being combined in the low-frequency summer 240 that adds the detected signals to produce a combined IF output 242.
Differences in the phase shifts and gains in complementary paths (i.e., the first and second RF paths 220, 222 and the first and second LO paths 224, 226) arising from component variation and manufacturing tolerances in the combining network 202 can result in different high-frequency output signals 232, 234 being supplied to the detectors 208, 210. Furthermore, differences in the detector components can result in different detector outputs 236, 238, even if the combined high-frequency signals are equal. In some embodiments, one or more of the gain values in the high-frequency combining network 202 is adjustable. In a particular embodiment, a gain value of an LO signal path (e.g., g3, g4) is adjustable to balance the detected LO power from each detector 236, 238, which allows nearly complete cancelation of the detected LO AM noise.
The added common mode noise at the either differential outputs 306, 308 or differential outputs 307, 309 will cancel at the output 320. Noise may enter the differential output sum from power supply rails or from other inputs. The system 300 cancels the common mode noise; however, it does not cancel differential noise showing up between outputs 306 and 308, or between outputs 307 and 309.
IF=(LOd+RFd)+(−(LOd−(RFd)/K)−(LOd−(RFd)/K)−(−(LOd+RFd)) (Eq. 1)
=LOd−LOd−LOd+LOd+RFd+(RFd)/K+(RFd)/K+RFd (Eq. 2)
=(2+2/K)RFd+2LOd−2LOd (Eq. 3)
=(2+2/K)RFd (Eq. 4)
The gain control 33 allows canceling the LO AM noise, while increasing the detected RF signal by approximately a factor of four (when the two detected LO signals are approximately equal, i.e., when K is close to unity).
In a particular embodiment, a downconverting system is calibrated to achieve cancelation of LO AM signal or noise by adding an amount of AM signal to each LO signal path of the system. By changing the gain K to create identical levels of detected LO AM signals at both detector outputs, the detected AM signal or noise seen at the IF will drop in amplitude. The LO AM canceling is done at baseband (IF/audio), which allows for very precise LO AM calibration/cancelation compared to calibrating at high frequencies (i.e., before detection), where mismatch errors degrade the calibration accuracy. In a particular embodiment, a downconverting system has a built-in calibration source, such as an AM signal or noise source (calibration standard), look-up table (“LUT”), or variable gain or attenuation stage. In a further embodiment, the downconverting system performs an automatic LO AM noise calibration according to firmware instructions (self-calibration). LO AM noise calibration provides improved signal-to-noise performance, which can provide superior range to a radar system.
IF=8RFd+4LOd−4LOd=8RFd (Eq. 5)
The detectors D1 through D4 may each see as much as half of the incident power to the NCDD, if they are configured as dual diode detectors. They detect peak voltage and add their outputs as voltage. If they are paired as a positive and negative peak detector, their low frequency detected output voltage may nearly double for the same power input. Thus, the four-detector system 400 cancels the detected LO AM noise while increasing the detected RF signal several times over what would be detected by a single-diode system. Uncorrelated diode-generated (i.e., detector generated) noise adds as power. In single diode detector systems, this uncorrelated noise is added to the detected signal and degrades range/sensitivity. In this four-detector system, the four detected RF output signals add as correlated voltages while the four noise outputs add as uncorrelated power, improving the signal to noise ratio and providing low-cost detection systems capable of detecting low or very low RF signals, whether reflected or remotely generated.
Switching mixer-based systems require relatively high LO power to drive the mixer diodes. Using a diode as a detector, rather than as a mixer, allows operating the system with lower LO power, which results in lower total system power consumption and wider design choice in LO design. A low LO power requirement at the detector diode also allows splitting the LO power to drive multiple detectors for LO AM noise canceling. Noise-canceling techniques according to embodiments provide downconverting systems with improved signal-to-noise performance over switching mixers or single diode detecting mixers. Using multiple RF detectors can further improve signal-to-noise performance.
The noise-canceling down-converting detector RF coupler 600 uses a ring hybrid coupler 602 and two diode detectors 604, 606. The diode detectors 604, 606 may be single or multiple diode detectors. The noise-canceling down-converting detector 600 is particularly desirable for coherent self-demodulated radar where the LO signal and the transmitted radar signal 610 are at the same frequency with a constant phase difference. The RF signal 612 is the portion of the transmitted (LO) signal 610 that is reflected by the target 614 with a frequency shift (i.e., Doppler shift) due to the target's velocity towards or away from the transmitted signal source (transmitted signal 610). Thus the RF signal 612 is shifted by only a small amount from the LO frequency (compared with the coupler bandwidth) and will have nearly the same wavelength as that transmitted signal and thus similar phase shift, and the system can be designed for a known frequency, which is generally the transmitted frequency, which in a particular embodiment is in one or more of the L-, S-, C-, X-, K-, Ku-, Ka-Band or other frequency. The details of the radar transmitting and receiving antennas are not shown in
The ring hybrid coupler 602 has four ports 616, 618, 620, 622. The ring hybrid coupler 602 splits the LO signal 615 arriving at the LO port 616 into two equal signals 624, 626 and sends those LO signals to two detector ports 618, 620 that are essentially identical. The LO signals travel one quarter or an odd multiple of one quarter wavelengths of the LO signal from the LO port to either diode port (clockwise to diode port 620 and counterclockwise to diode port 618). The RF port 622 is located on the ring hybrid coupler 602 a distance of an even multiple of one half wavelength clockwise from the LO port 616 and an odd multiple of one half wavelength of the RF signal in the counter-clockwise direction around the ring hybrid coupler 602.
The LO signals 624, 626 cancel at the RF port 622, as the minimum signal path for 624 is ½λ (180 degrees), while the minimum signal path for 626 is λ (360 degrees), thus the LO signals 624 and 626 arrive 180 degrees out of phase. Alternative embodiments use other multiples of wavelengths. In an embodiment, the RF port 622 sits on the ring hybrid coupler 602 separated from one detector 604 by one quarter wavelength in the counter-clockwise direction and 5/4λ in the clockwise direction, and separated from the second detector 606 by three quarters of a wavelength in either direction. Thus the RF signal arrives at each detector in phase from either direction around the ring. The closest distance between the two detector ports, 618,620 is ½λ. Thus, the signal from the RF port 622 splits and arrives at the two detectors 604, 606 with a one-hundred and eighty degree phase difference. The same phase relationship can also be accomplished by scaling the ring by odd multiples of ¼λ.
The RF signal 612 and LO signal 615 are not at exactly the same frequency, but are separated by a very low frequency due to the Doppler shift of the moving target of the radar. This can be visualized as a large LO signal adding with a smaller RF signal which appears nearly identical in frequency to the LO signal, but that moves slowly in phase over time. For some cycles of the waveform, the RF and LO signals as seen by either of the detectors add in phase and increase the total amplitude of the waveform. One half-cycle of the difference frequency later, the RF and LO will add out of phase and decrease the total amplitude of the waveform seen by that detector. This results in a low frequency output (i.e., baseband, or “audio”) from the diode detector at the difference frequency between RF and LO (either LO−RF or RF−LO), which is the frequency that results from the change of the phase in the path from the transmitted signal to the moving target and back. This constant change in phase (for a constant relative velocity of the target) is indistinguishable from a change in frequency and is understood as the Doppler Effect.
The two detectors. 604, 606 see the same phase of the signal from the LO port, but the small signal from the RF port adds to the amplitude of the LO signal at the first detector while it subtracts from the amplitude of the LO signal at the second detector. One half cycle of the difference frequency later, the phase of the RF signal has changed by 180 degrees versus the LO signal, and the first detector which saw an addition of the RF magnitude and the LO magnitude will now see a subtraction of the RF magnitude from the LO magnitude. Thus if one detector sees a higher output, the other detector will see a lower output due to the same RF signal. Coherent AM or AM noise of the LO signal will also show up as an addition or subtraction of the signals at each detector which will add or subtract at each detector identically (i.e., in phase) such that the algebraic combination of both detectors will see an increase in detected RF output level, while detected LO AM noise at that combined output is canceled.
An algebraic combining network (see
The differential detector and differential IF amplifier work together to deliver high sensitivity to the received RF signal while canceling the AM noise on the LO signal, which would otherwise limit the sensitivity of the downconverter. In a further embodiment, the differential IF amplifier provides adjustment of the amplitude of one IF signal compared to the other IF signal (see, e.g.,
Some embodiments include noise calibration to correct for variations in the detector gain and differences in signal path loss. Correct design of the RF transmission circuit and appropriate process control (repeatability) insures that the phase between LO signal portions and the RF signal portions remain balanced. Even if the RF path has some imbalance, sufficient LO AM noise is canceled to result in only a slight modification in the gain of the RF signals, while providing a large increase in the received signal to noise ratio.
The detectors 604, 606 do not operate like the diode switches used in typical mixers. The detectors 604, 606 can detect the input signal voltage with higher gain (voltage output for voltage input) if they are narrow bandwidth diode detectors, compared to the much wider bandwidth diode switches used in switching mixer circuits. Narrow-bandwidth diode detectors are easier to match than wider bandwidth diode switches, providing improved detection efficiencies and improved signal-to-noise ratio.
In an alternative embodiment, the RF port is connected to the coupler such that the two detectors see the RF signal portions arrive in phase with each other, but see the LO signal portions arrive 180 degrees out of phase with each other. The ring coupler geometry creates a differential structure that cancels the AM noise on the LO signal. The algebraic combining network still subtracts the two detector outputs in order to cancel the two detected LO AM noise portions, since the low frequency variation in AM noise will show up in phase at the two diode outputs regardless of the relative phase of the LO signals. Since the RF signals add to both detectors, but the LO signals on those detectors are 180 degrees out of phase with each other, the combination of LO and RF signals create a low frequency variation on the output of the detectors that is 180 degrees out of phase and at the frequency difference between the LO and RF signals. The algebraic combining network effectively adds the amplitudes of the two detected RF signal portions. This implementation may still require amplitude adjustment of one of the detector outputs to obtain the maximum cancelation of the LO AM signals or noise. Embodiments can incorporate single-diode or multiple-diode detectors or use alternative detector methods.
The detector outputs are coupled to a summing network (not shown, see
The diode uses a DC path to develop a current and provide a detected output (VOUT) from the incident high-frequency power (i.e., LO and RF signals). A network with fan lines 642, 646 on opposite ends of a high-impedance transmission line 644 provides both a high impedance at the LO frequency and a DC path for diode current, as is well known in the art of high-frequency hybrid microcircuit design. Any of several matching and bias structures are suitable in various embodiments, and some embodiments may utilize different matching and biasing techniques on different detector circuits. Accordingly, the detector circuit 630 is merely exemplary. Alternative embodiments use detector circuits with one or more diodes in a shunt configuration. It is generally desirable to provide a DC connection to the output of the diode without loading the resonance of the diode inductance and fan line 642 capacitance. Many mixer circuits operate in a relatively low characteristic impedance system, such as a 50-ohm system. Detectors in some embodiments operate in circuits with much higher impedance, which avoids loading the resonance and delivers higher voltage gain from the detector.
A second fan line 646 connects to a network 650 that has a series resistor 652, shunt capacitor 654 and a shunt resistor 656. The series resistor 652 should have a resistance large enough to provide isolation between the fan 646 and the following shunt capacitor 654, yet have a resistance sufficiently lower than the shunt resistor 656 to ground, so as not to unduly attenuate the output voltage VOUT. For example, with a shunt resistor 656 in the range of 1 to 5 K ohms, the series resistor 652 is between about 20 ohms and about 200 ohms.
The shut capacitor 654 is chosen to have a self-resonant frequency above the highest expected IF frequency, and below the LO frequency. In a particular, embodiment with an LO frequency of about 24 GHz, the shunt capacitor is chosen to have a self-resonant frequency not greater than about two GHz. This provides some immunity from adjacent signal sources, such as other radar units or communications devices, from affecting the output voltage VOUT. The shunt capacitor 654 and shunt resistor 656 in conjunction with the resistance and capacitance presented by the following amplifier connected to Vout will determine the IF bandwidth, which must be set large enough to enable reception of the highest frequency IF signal expected to be received. The shortened fan lines 642, 646 and series inductances of the diode and package, line 644, and other leads, provide attenuation at LO and RF frequencies. Several other networks are alternatively used, as would be appreciated by one of skill in the art.
Transmission lines 658, 660, 662 are used in the matching structure 634 to match the impedance of the diode/fan line resonance to the system impedance of the ring coupler. Other transmission line matching structures are alternatively used. For example, an alternative design uses a single one quarter wavelength long transmission line with impedance equal to the geometric mean of the source and load impedances. In an alternative embodiment, discrete components are used in a matching circuit.
The outputs VOUT1, VOUT2 of the two diodes are connected to the inputs of a differential amplifier 676. The output 678 of the differential amplifier 676 is connected to the input of a differential IF amplifier 682. The dual diode detector circuit 660 operates similarly to the single diode detector circuit of
Each ring is optimized to maintain to best performance of the overall system. In a particular embodiment, the local oscillator 720 is a dielectric resonator oscillator (“DRO”, also known as dielectric resonator stabilized oscillators (“DSOs”)). DROs are low-cost, compact, and consume relatively little power consumption; however, DROs often have too much AM noise to be used in conventional diode detector CW radar systems without noise cancelling. Embodiments of the invention canceling LO AM noise allow DROs to be used in many different radar applications, including radar applications requiring low power consumption, long range, or measurement accuracy. Alternative embodiments use a transmission line resonator oscillator or other oscillator.
The first ring coupler 708 receives the LO signal at a first port (“LO input port”) 724 and distributes the LO signal to an antenna port 726 for transmission to the target (not shown) and an LO port 728. The portion of the LO signal coupled to the LO port 728 is transmitted through a transmission line 730 to deliver the LO signal to the LO port 616 (“of the second ring coupler 702. The portion of the LO/RF signal coupled to the RF port 732, which is directly across from the LO port 728 and isolated from LO signal is coupled to the RF port 622 of the first ring coupler 702. Other details of the radar system 700 are well known in the art of RF circuit design and are omitted for clarity of illustration.
The LO input port 724 is separated from the LO output port 728 by ¼λ, from the antenna port 726 by ¼λ and from the RF output port 732 by ½λ CW and 1λ CCW for a difference of 180 degrees. This allows the LO signal to split into 2 paths, to the antenna and to the differential detector. The second ring hybrid coupler sends the LO signal to the two detector diodes nominally in phase and the RF input signal from the antenna nominally 180 degrees out of phase at the two diode detectors. In a particular embodiment, the LO 720 is fabricated on the opposite side (“second side”) of the substrate 723 from the side (“first side”) of the substrate that the ring couplers 702, 708 and antenna 722 are fabricated on. A ground plane 734 on the first side overlies the LO circuit on the second side (not shown), and the LO signal is brought from the second side, where it is generated, to the first side through a plated via 736. A ground plane on the second side (not shown) underlies the antenna 722 and generally the RF circuitry on the first side, as is known in the art of RF microstrip design.
The antenna 752 and receives a signal (RF signal) 758 reflected off the target 755. The reflected signal is combined with the LO signal in a combining network 757 (see, e.g.,
The outputs of the first and second detectors 760, 762 are provided to an algebraic summing network 768. Differential outputs of the summing network 768 are coupled to an amplifier and filter 770, which convert the differential inputs to a single-ended signal 771. An analog-to-digital converter 772 converts the signal, representing the detected voltage, into a digital value 773 that is processed by a controller 774, which renders the measured speed (between the target and system) to an electronic display screen 776, such as a liquid crystal display screen. The system 750 optionally includes a user interface 778 for communicating information such as the measured speed to an external device, accepting a trigger to start the radar speed measurement process, modifying the function of the radar to report the speed in alternative units or optimizing the radar speed measurement process to report the speed of a particular type of object (i.e. optimized for measuring a vehicle or a baseball) or report the measured speed within a particular limit of high and low speeds. In an alternative embodiment, a system does not include an integrated display, and the controller provides relative speed data to a device (not shown) outside of the system.
In a particular embodiment, the system 750 is a portable system powered by a battery(s) 784. In a further embodiment, the portable system is intended to be operated as a hand-held system. In an alternative embodiment, the battery is not included in the system, and power is supplied from an external source, such as an external battery (e.g., a vehicle battery) or mains power (e.g. from a transformer connected to the mains power). A power supply circuit 786 provides voltage regulation and similar functions to generate the appropriate voltages and supply currents to power the system components. Individual power lines to each powered component are omitted for simplicity and clarity of illustration. In a particular embodiment, operation of the power supplies are monitored by the controller 774 and A/D converter 772, and adjusted or turned off and on, if necessary, through a power supply control line or lines 787.
The controller 774 provides an amplitude modulation control signal 788 to the oscillator circuit 756, which creates amplitude modulation of the oscillator 756 to enable calibration of the NCDD to minimize the detected level of AM noise of the oscillator 756 by the NCDD. In a particular embodiment, the controller 774 includes memory, such as calibration tables, or the system 750 includes separate memory (not shown) that cooperates with the controller. In a particular embodiment, the controller 774 includes a signal processing block and an optional automatic calibration block that works in cooperation with an AM signal generator (i.e., a signal on the AM control line 788 that modulates the oscillator 756 in a known fashion). In a particular embodiment, the system is integrated in a housing 790 containing the other system elements for hand-held application. In a particular embodiment, the housing 790 is a rectangular plastic housing having approximate dimensions of 2.25 inches by 4.5 inches by one inch. Alternative systems are provided as original-equipment manufacturer (“OEM”) systems, and incorporated into other products wherein the housing may be omitted.
In a particular embodiment, the second detector detects the same phase of the RF signal as the first detector, and the opposite phase (i.e., 180 degrees out of phase) of the LO signal. One of the detected RF signals is inverted at the detector output and subtracted from the other detected RF signal, while the detected LO AM signals cancel due to the same subtraction. In an alternative embodiment, the second detector detects the opposite phase of the RF signal as the first detector, and the same phase of the LO AM signal. The output of one of the detector outputs is subtracted from the other to add the detected (downconverted) out of phase RF signals and cancel the detected in phase (demodulated) LO AM signals. In both cases the demodulated amplitude modulation of the LO signals comes out of the detectors in phase at the detector outputs and the downconverted RF signal comes out of the detectors out of phase at the detector outputs.
The first detected output is provided to a first algebraic combining network and the second detected output is concurrently provided to a second algebraic combining network (step 826). The first algebraic combining network produces a first differential signal and an inverse first differential signal and the second algebraic combining network produces a second differential signal and an inverse second differential signal (step 832). The first differential signal is provided to a positive input (i.e., non-inverting input) of a third algebraic combining network, the inverse first differential signal is provided to a negative input (i.e., inverting input) of the third algebraic combining network, the second differential signal is provided to a second negative input of the third algebraic combining network and the inverse second differential signal is provided to a second positive input of the third algebraic combining network. The third algebraic combining network produces an output (e.g., the IF output) (step 830). In a further embodiment, the combined detected reflected signal is processed to derive a relative speed between the radar system and the target (step 832).
In a further embodiment, gain (including negative gain, which is also known as attenuation) is applied to the output of the second algebraic combining network (step 834) to match the detected LO signal from the first detector to the detected LO signal from the second detector so as to cancel AM noise detected on the LO signals.
Gain adjustments are stepped through a selected range of settings (see, e.g.,
Gain is adjusted in various ways in alternative embodiments. For example, an adjustable gain element is included in one of the detected signal paths to increase or decrease the gain of that detected signal relative to another detected signal. Alternatively, the bias of elements, such as the bias level to one or more transistors, is changed. As the gain in one of the detector paths changes, the level of noise will fall to a minimum value. By observing the noise level for each bias control step, one can identify the gain setting where the noise begins to increase as the bias control leaves that optimum area (e.g., increasing or decreasing bias control voltage above or below the bias control voltage at the minimum noise condition). In a particular embodiment, a threshold is defined at a selected excursion from the minimum point. The gain settings at which the noise equals the threshold is determined (i.e., the minimum noise will be between the two threshold settings, but may be relatively “flat”, making a direct measurement of the minimum noise point less precise). The optimum point for best LO noise rejection is set by choosing the gain setting between the gain settings for the two thresholds identified above. In embodiments that use more than two detectors (see, e.g.,
A calibration signal that mimics AM noise of the LO, or a signal at an operating frequency (e.g., the LO frequency plus sidebands) is applied to a first detector and to a second detector of a noise-canceling down-converting detector system (see, e.g.,
While the invention has been described with reference to a preferred embodiment or embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention not be limited to the particular embodiment disclosed as the best mode contemplated for carrying out this invention, but that the invention will include all embodiments falling within the scope of the appended claims.
| Number | Name | Date | Kind |
|---|---|---|---|
| 3942178 | Hackett | Mar 1976 | A |
| 4115774 | Tresselt | Sep 1978 | A |
| 4176351 | DeVita et al. | Nov 1979 | A |
| 4254505 | Bossert | Mar 1981 | A |
| 4323898 | Barnes et al. | Apr 1982 | A |
| 4649388 | Atlas | Mar 1987 | A |
| 4856084 | Richards, Jr. | Aug 1989 | A |
| 4888557 | Puckette et al. | Dec 1989 | A |
| 4896374 | Waugh et al. | Jan 1990 | A |
| 4926443 | Reich | May 1990 | A |
| 5262783 | Philpott et al. | Nov 1993 | A |
| 5423080 | Perret et al. | Jun 1995 | A |
| 5570094 | Armstrong | Oct 1996 | A |
| 5596325 | Maas | Jan 1997 | A |
| 5806804 | Goodzeit et al. | Sep 1998 | A |
| 5861837 | Richardson et al. | Jan 1999 | A |
| 5864061 | Dilz, Jr. | Jan 1999 | A |
| 5940025 | Koehnke et al. | Aug 1999 | A |
| 6091355 | Cadotte, Jr. et al. | Jul 2000 | A |
| 6121919 | Ameen et al. | Sep 2000 | A |
| 6134281 | Green et al. | Oct 2000 | A |
| 6809681 | Niechayev | Oct 2004 | B1 |
| 7079400 | Inamoto et al. | Jul 2006 | B2 |
| 7099642 | Fernandez-Corbaton et al. | Aug 2006 | B2 |
| 7436356 | Bochkovskiy et al. | Oct 2008 | B2 |
| 8085185 | Stewart et al. | Dec 2011 | B2 |
| 8089394 | Stewart et al. | Jan 2012 | B2 |
| 20020102958 | Buer | Aug 2002 | A1 |
| 20040257268 | Aker et al. | Dec 2004 | A1 |
| 20060181449 | Aker | Aug 2006 | A1 |
| 20070296625 | Bruzzone et al. | Dec 2007 | A1 |
| 20080106460 | Kurtz et al. | May 2008 | A1 |
| 20110102240 | Stewart et al. | May 2011 | A1 |
| 20110102241 | Stewart et al. | May 2011 | A1 |
| 20110102246 | Moulton et al. | May 2011 | A1 |
| Number | Date | Country |
|---|---|---|
| 0098339 | Jan 1984 | EP |
| 53005590 | Jan 1978 | JP |
| 4184184 | Jul 1992 | JP |
| 2000-028702 | Jan 2000 | JP |
| 2004-247831 | Sep 2004 | JP |
| 10-2005-0056932 | Jun 2005 | KR |
| WO 96-07931 | Mar 1996 | WO |
| WO 01-26215 | Apr 2001 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20110102246 A1 | May 2011 | US |
