Nothing Special   »   [go: up one dir, main page]
Next Article in Journal
Electrochemical Sensors Based on Self-Assembling Peptide/Carbon Nanotube Nanocomposites for Sensitive Detection of Bisphenol A
Next Article in Special Issue
Subcranial Encephalic Temnograph-Shaped Helmet for Brain Stroke Monitoring
Previous Article in Journal
Time Delay Study of Ultrasonic Gas Flowmeters Based on VMD–Hilbert Spectrum and Cross-Correlation
Previous Article in Special Issue
An Intermodulation Radar for Non-Linear Target and Transceiver Detection
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sensors
Volume 24
Issue 5
10.3390/s24051460
sensors-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Fully Integrated 24-GHz 1TX-2RX Transceiver for Compact FMCW Radar Applications

by
Goo-Han Ko
1,
Seung-Jin Moon
1,
Seong-Hoon Kim
1,
Jeong-Geun Kim
2 and
Donghyun Baek
1,*
1
School of Electrical Engineering, Chung-Ang University, Seoul 06974, Republic of Korea
2
Department of Electronic Engineering, Kwangwoon University, Seoul 01897, Republic of Korea
*
Author to whom correspondence should be addressed.
Sensors 2024, 24(5), 1460; https://doi.org/10.3390/s24051460
Submission received: 19 December 2023 / Revised: 7 February 2024 / Accepted: 15 February 2024 / Published: 23 February 2024
(This article belongs to the Special Issue Short-Range Radar-Based Techniques for Remote Monitoring and Medical Related Applications)
Browse Figures
Figure 1
<p>(<b>a</b>) FMCW radar principle and (<b>b</b>) angle estimation based on the monopulse phase comparison method.</p> "> Figure 2
<p>Block diagram of the proposed 24 GHz 1TX-2RX FMCW radar transceiver.</p> "> Figure 3
<p>(<b>a</b>) Block diagram and (<b>b</b>) parameter values of the FMCW synthesizer.</p> "> Figure 4
<p>(<b>a</b>) Detailed schematics and parameter values of the VCO with the driving buffer, (<b>b</b>) the schematic of the three-bit cap bank, (<b>c</b>) layout of the triple-coupled transformer, (<b>d</b>) measured frequency range, and (<b>e</b>) measured phase noise.</p> "> Figure 4 Cont.
<p>(<b>a</b>) Detailed schematics and parameter values of the VCO with the driving buffer, (<b>b</b>) the schematic of the three-bit cap bank, (<b>c</b>) layout of the triple-coupled transformer, (<b>d</b>) measured frequency range, and (<b>e</b>) measured phase noise.</p> "> Figure 5
<p>(<b>a</b>) Block diagram and (<b>b</b>) timing diagram of the FSG.</p> "> Figure 6
<p>Schematic of the three-bit attenuator.</p> "> Figure 7
<p>Simulation results: (<b>a</b>) relative attenuation state and (<b>b</b>) return loss according to switch state.</p> "> Figure 8
<p>Schematic of the two-stage DA and one-stage PA.</p> "> Figure 9
<p>Simulation results of the DA and PA: (<b>a</b>) gain and Pout versus Pin, (<b>b</b>) PAE, and (<b>c</b>) IP3.</p> "> Figure 10
<p>Overall architecture of the RX.</p> "> Figure 11
<p>(<b>a</b>) Schematic and (<b>b</b>) simulation results of the two-stage LNA, and (<b>c</b>) front end conversion gain without TIA.</p> "> Figure 12
<p>(<b>a</b>) Schematic of the TIA with DCOC and (<b>b</b>) schematic of the one-stage BPF.</p> "> Figure 13
<p>Simulation results: (<b>a</b>) gain variation and (<b>b</b>) the HPF cut-off frequency variation.</p> "> Figure 14
<p>Chip microphotograph.</p> "> Figure 15
<p>Measured phase noise at 24 GHz carrier frequency.</p> "> Figure 16
<p>Measured chirp transient when chirp bandwidth is 250 (black line), 500 (red line), and 750 (blue line) MHz: (<b>a</b>) triangular waveform with hold-chirp time, (<b>b</b>) sawtooth waveform with hold-chirp time, (<b>c</b>) triangular waveform without hold-chirp time, and (<b>d</b>) sawtooth waveform without hold-chirp time.</p> "> Figure 17
<p>Measurement results: (<b>a</b>) output power versus the operation frequency and (<b>b</b>) chirp spectrum.</p> "> Figure 18
<p>Measurement results: (<b>a</b>) TRX loopback test, (<b>b</b>) the HPF cut-off frequency variation, and (<b>c</b>) gain variation when the HPF cut-off frequency was 500 kHz.</p> "> Figure 19
<p>(<b>a</b>) Implemented prototype FMCW radar module using the proposed transceiver and Yagi antenna and (<b>b</b>) stack up of the radar module.</p> "> Figure 20
<p>(<b>a</b>) Design configuration, (<b>b</b>) simulation results of the Yagi antenna, and (<b>c</b>) simulated H-plane and E-plane beam patterns.</p> "> Figure 21
<p>Measurement setup.</p> "> Figure 22
<p>Range measurement: (<b>a</b>) 2 m, (<b>b</b>) 4 m, (<b>c</b>) 6 m, (<b>d</b>) 8 m, (<b>e</b>) 10 m, and (<b>f</b>) 12 m.</p> "> Figure 23
<p>Range measurement: distance error between real distance and measured distance.</p> "> Figure 24
<p>Angle measurement: (<b>a</b>) 0 degree, (<b>b</b>) −30 degree, and (<b>c</b>) 30 degree.</p> ">
Versions Notes

Abstract

:
A fully integrated 24-GHz radar transceiver with one transmitter (TX) and two receivers (RXs) for compact frequency modulated continuous wave (FMCW) radar applications is here presented. The FMCW synthesizer was realized using a fractional-N phase-locked loop (PLL) and programmable chirp generator, which are completely integrated in the proposed transceiver. The measured output phase noise of the synthesizer is −80 dBc/Hz at 100 kHz offset. The TX consists of a three-bit bridged t-type attenuator for gain control, a two-stage drive amplifier (DA) and a one-stage power amplifier (PA). The TX chain provides an output power of 13 dBm while achieving <0.5 dB output power variation within the range of 24 to 24.25 GHz. The RX with a direct conversion I-Q structure is composed of a two-stage low noise amplifier (LNA), I-Q generator, mixer, transimpedance amplifier (TIA), a two-stage biquad band pass filter (BPF), and a differential-to-single (DTS) amplifier. The TIA and the BPF employ a DC offset cancellation (DCOC) circuit to suppress the strong reflection signal and TX-RX leakage. The RX chain exhibits an overall gain of 100 dB. The proposed radar transceiver is fabricated using a 65 nm CMOS technology. The transceiver consumes 220 mW from a 1 V supply voltage and has 4.84 mm2 die size including all pads. The prototype FMCW radar is realized with the proposed transceiver and Yagi antenna to verify the radar functionality, such as the distance and angle of targets.

1. Introduction

In recent years, the demand to sense the surrounding environment has been rapidly increasing, not only in the automotive industry but also in appliance businesses, such as smart appliances, energy control, security, and emergency detection, with the development of the Internet of Things (IoT). Sensor types used in various applications include radar, laser, camera, infrared, and ultrasonic [1,2]. Among them, industrial, scientific, and medical (ISM)-band 24 GHz radar based on frequency-modulated continuous wave (FMCW) is in the spotlight because it is not sensitive to weather, light condition, and temperature, and can be installed invisibly [3,4]. However, FMCW radar systems have lacked competitiveness owing to their complex structure and high bill of materials (BOM) compared to other sensors. For example, an FMCW radar system being implemented by combining discrete radio frequency integrated chips (RFICs) on a printed circuit board (PCB) has large size, high cost, and high system complexity. In order to solve these problems, several works have been established to integrate discrete integrated chips (ICs), such as a transmitter (TX), a receiver (RX), and local oscillator (LO), into a single chip [5,6,7,8,9]. However, although the previous works have unique advantages in terms of the number of TX-RX channels and power consumption, there are still considerable shortcomings regarding the size and cost-effectiveness of FMCW radar applications. In [5,6], multi-channel TX-RX radar transceivers were presented for the FMCW radar. In [7,8], 1-TX and 1-RX radar transceivers with TRX leakage cancellation and the TRX novel switching technique were presented for a single antenna operation. However, they still require external ICs such as phase-locked loop (PLL) for chirp generation and analog baseband for intermediate frequency (IF) amplification. This constrains their usage in compact and low-cost FMCW radar applications. Another requirement for the FMCW radar is a fast chirp generation capability. In [9], a fully integrated 1TX and 2-RX radar transceiver was presented using complementary metal oxide semiconductor (CMOS) technology. However, the slow-chirp waveform with ramp duration > 1 ms was used in a target detection scenario. A radar using the slow chirp has range-doppler ambiguity and poor signal-to-noise ratio (SNR) in the IF signals because the IF signals suffer from the baseband flicker noise and the noise floor of the other targets. Thus, a fast chirp is required to improve these problems. The fast chirp is makes it possible to move the IF signal of the targets away from the flicker noise corner of the baseband and increase the frequency difference between the IF signals in a multi-target detection scenario [10,11].
In this paper, a fully integrated 24 GHz 1TX-2RX transceiver for compact FMCW radar applications is presented. An FMCW synthesizer being integrated in the proposed transceiver generates the fast chirp waveform through a fractional-N PLL, voltage-controlled oscillator (VCO), and programmable chirp generator. The FMCW synthesizer supports triangular and sawtooth waveforms with reconfigurable pulse repetition internal (PRI). The TX chain exhibits a high output power by utilizing a two-stage driving amplifier (DA) and a one-stage power amplifier (PA) with a common-source (CS) structure including a neutralization capacitor. Furthermore, the 0–7 dB attenuation stage for TX gain control is realized through a fully differential bridged T-type attenuator. The RX chain with a direct conversion I-Q structure demonstrates high linearity at the given chirp bandwidth by adopting the transformer-based gain-bandwidth enhancements technique. An analog baseband in the RX chain mitigates the problem of TX-RX leakage and short-range interference through a range compensation filter with a −40 dB slope.
The remainder of this paper is organized as follows. Section 2 presents the fundamental principle of the FMCW radar. Section 3 provides the system architecture and the details of circuit implementations of key building blocks. Further, Section 4 verifies the transceiver and radar functionality with the measurement results. Finally, Section 5 concludes this article.

2. FMCW Radar Principle

The FMCW radar transmits a continuous wave, whose frequency changes linearly, and receives the signal reflected from the targets. The FMCW radar compares the received signal to the transmitted signal by mixing them. Then, IF signals are extracted due to the time delay between two signals, as shown in Figure 1a. After the extracted IF signals are sampled by the analog-to-digital converter (ADC), 2-D fast Fourier transform (FFT) is performed to obtain information about the range and velocity of the targets [12,13].
An FMCW radar consisting of two receivers determines the angle of the targets using the mono-pulse phase comparison method, as shown in Figure 1b. The angle is determined by comparing the phase difference between the received signals from two separate antenna arrays, as given by Equation (1):
θ = sin − 1 ⁡ [ ʎ · ( φ R X 1 − φ R X 2 ) 2 π · d ]  
where ʎ is wavelength, φ R X 1 and φ R X 2 are the phases of the IF signals for the targets, and d is the distance between the antenna arrays, respectively.

3. Design of 24 GHz 1TX-2RX FMCW Radar Transceiver

The proposed transceiver consists of a one-channel TX, two-channel RXs, LO and supporting circuits such as a bandgap reference (BGR), a master current generator (MCG), a reference current generator (IREF), low-dropout regulators (LDRs), and a serial–peripheral interface (SPI) as shown in Figure 2. The FMCW synthesizer consists of a fractional-N PLL, a class-C VCO operating at 24 GHz, a buffer to drive the VCO output, and a frequency sweep generator (FSG) for the programmable chirp generation. The buffer that drives the VCO output provides 24 GHz differential signals to the first split buffer in the LO distribution network. The first split buffer output is connected to the second split buffer input to distribute the signals to the two-channel receiver, and is connected to the attenuator input in the TX. The TX includes three-bit attenuator for gain control with an attenuation step of 1 dB, two-stage DA and one-stage PA. Each RX includes a two-stage LNA, a passive LC I/Q generator, a mixer, a trans-impedance amplifier (TIA), and an analog baseband. The analog baseband consists of a two-stage biquad band pass filter (BPF) and a differential-to-single amplifier (DTS).

3.1. FMCW Synthesizer

The overall block diagram and parameter values of the FMCW synthesizer are shown in Figure 3. The FMCW synthesizer consists of three parts: a sigma-delta modulator (SDM)-based fractional-N PLL, an LC VCO for LO generation, and a FSG. The SDM-based fractional-N PLL is composed of a phase frequency detector (PFD), a charge pump (CP), a three-order loop filter (LF), a frequency divider stage, and a three-order multistage noised-shaping (MASH) SDM for fractional division. The frequency divider stage is realized through a fixed 1/4 and 1/16 current mode logic (CML) divider and an eight-bit multi-modulus divider (MMD). The fixed 1/4 CML divider provides the VCO output frequency divided by 4 to the MMD. The LC VCO for the LO generation is implemented with a triple-coupled transformer for the class-c bias technique, a three-bit cap bank for a wide frequency operation range, and a mos-varactor. The detailed schematics and parameter values of the VCO with the driving buffer are described in Figure 4a. Each inductor L1 and L2 is connected to the CML and the driving buffer, respectively. Figure 4b,c show the schematic of the three-bit cap bank and the layout of the triple-coupled transformer, respectively. The cap bank includes three T-gate switches and unit capacitors arrays. The triple-coupled transformer consists of three top metals. All transformers are stacked in a 1:1 transformer configuration. Figure 4d,e shows the measurement results of the LC class-c VCO. The measured frequency-tuning range of the VCO is 22.8–27.3 GHz. Further, the phase noise is −104 dBc/Hz at 1-MHz offset frequency.
The FSG generates a programmable modulated frequency division ratio code. The input of the FSG consists of 33-bit data, which determine a FMCW waveform parameter such as a chirp time and a chirp bandwidth. The output of the FSG consists of 24-bit data. The lower 16-bit controls the fractional division ratio (FDR) of the SDM, and the upper 8-bit controls the integer division ratio (IDR) of the MMD. When the FDR exceeds the integer 1-division ratio, the IDR of the MMD is increased to the 1-division ratio while maintaining the extra FDR exceeding the 1-division ratio. As a result, the FSG provides ΔN.F at the modulation start frequency, being determined as N.Finitial. Figure 5 shows a detailed block diagram and timing diagram of the FSG.
The FSG consists of a mode controller, a data register, a multiplexer array, a modulator, and an accumulator. The mode controller, implemented as a finite state machine, determines the chirp sequence according to the two-bit input code. It has three modes: 00 (Hold chirp), 01 (Up chirp–Down chirp), and 10 (Up chirp–Down chirp–Hold chirp). The data register has three datasets, which determine the FMCW waveform parameter. Each dataset is structured in a 33-bit format, which allocates 5 bits for the unit time step (NT), 12 bits for the unit frequency step (NF), and 16 bits for the Up-, Down-, and Hold-chirp period steps (Nc). The multiplexer array provides corresponding 33-bit data to the modulator depending on the mode. The modulator includes a TSTEP generator, a FSTEP generator, and a CSTEP generator. The TSTEP generator determines a clock TSTEP (= TDIVNT) to provide a reference clock to the FSTEP generator and the CSTEP generator. The FSTEP generator determines a frequency variation FSTEP (= FRESNF) to be changed according to TSTEP, where the FRES is the resolution frequency of the PLL. Finally, the CSTEP generator determines a clock CSTEP (= TSTEPNC) to shift the chirp sequence. Therefore, the chirp bandwidth BW is determined by the frequency variation FSTEP and NC as BW = FSTEP × NC.

3.2. Transmitter

The three-bit attenuator is designed as a differential bridged T-type for low mismatch and high linearity. A body floating technique is adopted to improve the insertion loss and prevent signal leakage. The schematic of the three-bit attenuator is depicted in Figure 6. R1 and R2 are determined by Equations (2) and (3) where RO is the characteristic impedance and ATTdB is the desired attenuation in the dB scale. The attenuation values of each stage are 1 dB, 2 dB, and 4 dB. Figure 7 shows simulation results for the relative attenuation state and return loss according to the switch state.
R 1 = R O ( 10 A T T d B 20 − 1 )
R 2 = R O ( 10 A T T d B 20 − 1 )
Figure 8 shows the overall schematic of the two-stage DA and one-stage PA. All stages adopt the transformers (T1, T2, T3, T4) to improve bandwidth. The topology of the DA and PA is designed as a CS amplifier with a neutralization capacitor. The neutralization capacitor is a capacitor that crosses a differential signal and acts as a negative capacitor because the phase of the differential signal is reversed. This cancels the parasitic gate-drain capacitance of the transistor, increasing the input impedance of the CS amplifier and eliminating the negative feedback path. In conclusion, linearity and isolation can be increased, and the maximum available gain (MAG) value can be improved by reducing S12. In addition, since the amplifier consists of one stack, unlike the cascode structure, the VOV consumption generated by the transistor is reduced and the output voltage swing is increased. The load impedance of the PA is optimized to the impedance that can transmit the highest output power through the load pull test under the conditions of the 50 ohm output port. The simulated optimal load pull impedance is (21 + j11) ohm. Figure 9 summarizes the simulation results of the overall stage, achieving a gain of about 32.5 dB, an output P1 dB of 13.8 dBm, a power-added efficiency (PAE) of 29.3%, and an output three-order intercept point (IP3) of about 34.5 dBm.

3.3. Receiver

The Figure 10 shows the overall architecture of the RX. The RX consists of a transformer-based two-stage LNA, a mixer, an IQ generator, a TIA, a two-stage BPF, and a DTS amplifier. The two-stage LNA adopts the transformer-based gain-bandwidth enhancement technique to realize wideband gain at the given chirp bandwidth. The wideband gain is generated from a magnetically coupled resonator with an output capacitance of the preceding stage’s LNA, the input capacitance of the next stage’s LNA, and the low coupling factor k of the transformer. This magnetically coupled resonator exhibits wide impedance over two resonance points depending on capacitance and transformer values. Figure 11a shows the schematic of the two-stage LNA. The first stage LNA is implemented with a single-input cascode amplifier topology, including a matching inductor L1, a source degeneration inductor L2, and variable resistor R1 for gain control. The input impedance of the first-stage LNA matches 50 ohm through an LC matching network. The matching network employs two capacitors (C1M and C2M) and two inductors (L1M and L2M). The second stage’s LNA is implemented with a differential cascode amplifier topology. The two transformers (T1 and T2) are optimized so that the gain of the front end has wideband characteristics. Figure 11b shows the simulation results of the two-stage LNA. When the two-stage LNA operates in high gain mode, it achieves an input P1 dB of −15.5 dBm and an input IP3 of 0 dBm. The mixer is a conventional double-balanced passive mixer [14], and the IQ generator is an LC-type poly-phase filter [15]. The I-Q LO signals drive the gate of the passive mixers, and the LNA output current signals are amplified through the TIA. In Figure 11b, the front-end conversion gain demonstrates wideband gain from 19 to 28 GHz, which is sufficient to cover the chirp bandwidth. The TIA is implemented with a DC offset cancellation circuit (DCOC), which cancels the voltage DC offset and forms a high-pass filter (HPF) with −20 dB slope. Likewise, the HPF of the two-stage BPF in analog baseband is also implemented as DCOC. The TIA and first-stage BPF create the HPF region with −20 dB slope at the same cut-off frequency. This can achieve an HPF with a total of −40 dB slope, which mitigates the problem of TX–RX leakage and short-range interference. The second-stage BPF forms the HPF region with −20 dB slope at the cut-off frequency of 156 Hz to further suppress TX to RX leakage. The cut-off frequency of the HPF with −40 dB slope is controlled using two-bit to realize the variation (68, 125, 250, and 500 kHz). The cut-off frequencies of the low-pass filter (LPF) of the TIA and the two-stage BPF are the same at 2 MHz. This establishes an LPF region with a total of 60 dB slope.
H ( s ) = R F R I ⋅ s 1 R F C F s 2 + s 1 R F C F + 1 R C R D C D C F 2
H H P F ( s ) = s s + R F C D R C R D
From the transfer function given by Equations (4) and (5), we can see that the gain of the two-stage BPF is the ratio of RF and RI and the LPF cut-off frequency is determined by RF and CF. The HPF cut-off frequency is determined by CD, RF, RC, and RD. Figure 12 shows a detailed schematic of the TIA and the one-stage BPF. The circuits employ the differential op-amps. The common-mode feedback (CMFB) of the op-amp is included in the op-amp itself. The phase margin of the CMFB in the op-amp is designed to be > 50 degrees, and the phase margins for TIA and BPF including the DCOC circuits are also secured to be more than 50 degrees. The simulation results of the TIA and two-stage BPF are summarized in Figure 13. Figure 13a shows the gain variation of the TIA and two-stage BPF. The receiver achieves the maximum gain of the TIA and BPF stages of 60 dB, and the controllable gain step is 1 dB. Figure 13b shows the HPF cut-off frequency variation induced by two-bit code control when the gain is 40 dB.

4. Measurement Results

4.1. Transceiver

The chip microphtograph of the proposed 1TX-2TX transceiver is shown in Figure 14. The chip size, including all pads, is 4.84 mm2. A 2.5 V supply voltage is applied to the internal LDRs, which provide the 0.9–1.2 V internal voltages to each sub-block circuit. All analog and power pads are protected from electrostatic discharge (ESD) using the gate-coupled MOSFET and clamp circuits. Table 1 shows the current consumption of each block. The total current consumption is 220 mA when all the block is turned on in the FMCW radar mode.
The proposed radar transceiver is measured using an on-chip measurement setup, wherein the transceiver chip is glued onto the PCB. The radar functionality is measured using a radar module including the proposed transceiver and Yagi antenna.
The phase noise performance is shown in Figure 15. The phase noise for the 24 GHz carrier frequency is −80 dBc/Hz, with an offset frequency of 100 kHz when using a 40 MHz crystal as the reference frequency. Figure 16 demonstrates the measured chirp transient of triangular waveforms and sawtooth waveforms using a signal source analyzer in the time domain. The difference between Figure 16ab–cd relates to whether the hold-chirp time is set or not.
The TX and RX were measured with the on-wafer measurement method using a ground–signal–ground (GSG) probe. The total calibrated loss due to the GSG probe, RF connector and cable is 4 dB. The measured output power of the TX reaches 13.3 dBm when the supply voltage is 1 V, provided by the LDR. The output power can be reconfigured by changing the supply voltage, as shown in Figure 17a. The measured output power variation in the chirp spectrum is less than 1 dB within the range of 24 to 24.25 GHz, as shown in Figure 17b.
The RX was measured through the output port of the DTS amplifier in the time domain and frequency domain to verify its functional operation. First, a loopback measurement was performed to check the voltage waveform of the down-converted IF signals. The proposed transceiver was operated in the FMCW mode. The TX output port links to the LNA input port of the RX through a cable. Subsequently, IQ IF signals corresponding to the cable length are generated as output. Figure 18a shows the measurement results for the parameter set at an Up-chirp time of 200 μs, a chirp bandwidth of 250 MHz, and a cable length of 4 m. The frequency of the IF signal is about 33 kHz. When substituted into the radar range equation, this frequency corresponds to a distance of 7.92 m. Figure 18b,c demonstrate the HPF cut-off frequency variation with a fixed gain of 40 dB, and gain variation with a fixed HPF cut-off frequency of 500 kHz. The signal generator was set to generate a 24 GHz signal, which was then directed to the RX while maintaining a fixed output power. Likewise, the proposed transceiver was operated in the continuous wave (CW) mode at 24 GHz. Then, alterations in the HPF cut-off frequency and gain were controlled through SPI. The output of the RX was observed with a spectrum analyzer in the frequency domain. The measured HPF cut-off frequencies were about 68, 125, 250, and 500 kHz according to the two-bit control code, while maintaining the LPF cut-off frequency at about 2 MHz. The maximum measured RX conversion gain was about 100 dB. A comparison of performance with previous works is summarized in Table 2.

4.2. Radar Functionality

Figure 19a shows the prototype FMCW radar module. The proposed transceiver was glued onto the module and gold-bonded in a chip-on-board (COB) layout. Furthermore, a 40 MHz crystal was mounted for the reference frequency. A total of three unit Yagi antennas were employed, and the distance between the two Yagi antenna for the RXs was set as ʎ /2. The prototype radar module consists of a total of 4 layers using FR-4, as shown in Figure 19b. The director of the Yagi antenna was placed on the top layer where the transceiver was glued, and the ground reflector was placed on the bottom layer. Figure 20 outlines detailed design parameters and simulation results related to the unit Yagi antenna.
The unit Yagi antenna consisted of 5 directors, a quarter wavelength transformer for impedance matching, a 50 ohm feeding line with coplanar waveguide with ground (CPWG), and a ground reflector. The Yagi antenna achieved a return loss of <−20 dB within the frequency range of 24 GHz to 25 GHz, as shown in Figure 20a. Finally, the simulated H-plane and E-plane beam patterns are depicted in Figure 20b,c. The gain was about 7 dB, and the half power beamwidth (HPBW) was 69.4° in the H-plane and 59.4° in the E-plane. The radar functionality has been verified in real environments by estimating the range and angle of a target. A corner reflector with a high radar cross section (RCS) was used as a target in the real environment. As shown in Figure 21, the corner reflector was placed in front of the radar module. The distance between the radar module and the reflector was varied at 2 m intervals. The chirp waveform characteristics in the measurement are as follows: chirp bandwidth 250 MHz, Up-chirp time 50 μs, Down-chirp time 5 μs, and Hold-chirp time 0 μs. Figure 22 shows the measured results, which were obtained by performing FFT on the IF signals of receiver channel-1, and Figure 23 shows the distance error between the real distance and the measured distance. For all distances the error was measured to be below 0.2 m. Finally, for angle detection, the corner reflector was placed 3 m away from the radar module. The angle detection measurements were made by changing the angles of the corner reflector with high RCS to −30, 0, and +30 degrees. Similar to range estimation, after performing FFT on the IF signals of receiver channels 1 and 2, the phase of the only IF signal with maximum magnitude was extracted, and then the angle was estimated through the mono-pulse angle estimation method. Figure 24 shows the measurement results of angle detection.

5. Conclusions

A fully integrated 1TX 2RX transceiver for a 24 GHz FMCW monopulse radar is here presented. The entire transceiver was realized with a small chip size of 4.84 mm2 through full integration. The FMCW synthesizer has been proposed for fast chirp generation and a reconfigurable chirp waveform. The synthesizer achieved various waveforms such as sawtooth and triangular waveform with chirp time and chirp bandwidth variation, and a phase noise of −80 dBc/Hz at the offset frequency of 100 kHz. The TX consists of a bridged T-type ATT, a DA and a PA. The PA optimized the load impedance through load pull testing and achieved an output power of 12.6 dBm up to 14.6 dBm, depending on the supply voltage of the LDR. The RX achieved a wideband gain in the front-end stage by using a magnetically coupled resonator with a low coupling factor k. The TIA and the two-stage BPF can be used to adjust the gain and cut-off frequency according to the target distance budget. The overall RX achieved high linearity characteristics and an overall maximum gain of 100 dB. The prototype FMCW radar module was developed using the proposed transceiver and Yagi antennas. The radar module has demonstrated the ability to detect the positions of targets in real environments.

Author Contributions

Conceptualization, G.-H.K., S.-J.M., S.-H.K. and D.B.; methodology, G.-H.K., S.-J.M., S.-H.K. and D.B.; writing—original draft preparation, G.-H.K., S.-J.M., S.-H.K. and D.B.; writing—review and editing, G.-H.K. and D.B.; supervision, J.-G.K. and D.B.; project administration, J.-G.K. and D.B.; funding acquisition, D.B. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Institute for Information and Communications Technology Planning and Evaluation (IITP) grant funded by the Korean Government (MSIT) (No. 2019-0-00138, Development of Intelligent Radar Platform Technology for Smart Environments), by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. RS-2023-00243744, Research on Future High-Pixel Image Radar Using Wideband/Multichannel 60-GHz CMOS FMCW Radar Transceivers), and by the Chung-Ang University Research Scholarship Grants in 2022.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data can be obtained from the authors on request.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Patole, S.M.; Torlak, M.; Wang, D.; Ali, M. Automotive radars: A review of signal processing techniques. IEEE Signal Process Mag. 2017, 34, 22–35. [Google Scholar] [CrossRef]
  2. Ma, T.; Deng, W.; Chen, Z.; Wu, J.; Zheng, W.; Wang, S.; Qi, N.; Liu, Y.; Chi, B. A CMOS 76–81-GHz 2-TX 3-RX FMCW Radar Transceiver Based on Mixed-Mode PLL Chirp Generator. IEEE J. Solid-State Circuits 2020, 55, 233–248. [Google Scholar] [CrossRef]
  3. Park, J.; Ryu, H.; Ha, K.W.; Kim, J.G.; Baek, D.H. 76–81-GHz CMOS Transmitter with a Phase-Locked-Loop-Based Multichirp Modulator for Automotive Rradar. IEEE Trans. Microw. Theory Tech. 2015, 63, 1399–1408. [Google Scholar] [CrossRef]
  4. Luo, T.N.; Wu, C.H.E.; Chen, Y.J.E. A 77 GHz CMOS Automotive Radar Transceiver with Anti-Interference Function. IEEE Trans. Circuits Syst. I Regul. Pap. 2013, 60, 3247–3255. [Google Scholar] [CrossRef]
  5. Chakraborty, A.; Hamouda, M.; Chen, Y.; Lautenschlager, C.; Englisch, D.; Qureshi, M.; Huynh, N.-H.; Hoenninger, K.; Forstner, H.-P. A scalable multimode 24-GHz radar transceiver for industrial and consumer applications in a 0.13 μm SiGe BiCMOS technology. In Proceedings of the 2019 IEEE BiCMOS and Compound Semiconductor Integrated Circuits and Technology Symposium (BCICTS), Nashville, TN, USA, 3–6 November 2019; pp. 1–4. [Google Scholar]
  6. Maiellaro., G.; Alessi, G.; Bruno, A.; Calcagno, A.; Micalizzi, N.; Di Mauro, A.; Susino, E.; Scuderi, A.; Scaccianoce, S. A 24-GHz transceiver with RF power envelope digital control for automotive radar ICs. In Proceedings of the 2017 12th European Microwave Integrated Circuits Conference (EuMIC), Nuremberg, Germany, 8–10 October 2017; pp. 45–48. [Google Scholar]
  7. Pyo, G.; Kim, C.-Y.; Hong, S. Single-antenna FMCW radar CMOS transceiver IC. IEEE Trans. Microw. Theory Tech. 2017, 65, 945–954. [Google Scholar] [CrossRef]
  8. Hamidian, A.; Ebelt, R.; Shmakov, D.; Vossiek, M.; Zhang, T.; Subramanian, V.; Boeck, G. 24 GHz CMOS transceiver with novel T/R switching concept for indoor localization. In Proceedings of the 2013 IEEE Radio Frequency Integrated Circuits Symposium (RFIC), Seattle, WA, USA, 2–4 June 2013; pp. 293–296. [Google Scholar]
  9. Su, Y.-P.; Huang, C.-Y.; Chen, S.-J. A 24-GHz fully integrated CMOS transceiver for FMCW radar applications. IEEE J. Solid-State Circuits 2021, 56, 3307–3317. [Google Scholar] [CrossRef]
  10. Deng, W.; Chen, Z.; Jia, H.; Yan, A.; Sun, S.; Chen, G.; Wang, Z.; Chi, B. A Self-Adapted Two-Point Modulation Type-II Digital PLL for Fast Chirp Rate and Wide Chirp-Bandwidth FMCW Signal Generation. IEEE J. Solid-State Circuits 2022, 57, 1162–1174. [Google Scholar] [CrossRef]
  11. Renukaswamy, P.T.; Markulic, N.; Wambacq, P.; Craninckx, J. A 12-mW 10-GHz FMCW PLL Based on an Integrating DAC With 28-kHz RMS-Frequency-Error for 23-MHz/μs Slope and 1.2-GHz Chirp-Bandwidth. IEEE J. Solid-State Circuits 2020, 55, 3294–3307. [Google Scholar] [CrossRef]
  12. Hyun, E.; Jin, Y.-S.; Lee, J.-H. A Pedestrian Detection Scheme Using a Coherent Phase Difference Method Based on 2D Range-Doppler FMCW Radar. Sensors 2016, 16, 124. [Google Scholar] [CrossRef] [PubMed]
  13. Hyun, E.; Jin, Y. Doppler-Spectrum Feature-Based Human–Vehicle Classification Scheme Using Machine Learning for an FMCW Radar Sensor. Sensors 2020, 20, 2001. [Google Scholar] [CrossRef] [PubMed]
  14. Razavi, B. RF Microelectronics, 2nd ed.; Prentice Hall: Hoboken, NJ, USA, 2013; p. 53. [Google Scholar]
  15. Christensen, K.T. LC quadrature generation in integrated circuits. In Proceedings of the 2001 IEEE International Symposium on Circuits and Systems, Sydney, NSW, Australia, 6–9 May 2001; pp. 41–44. [Google Scholar]
Sensors 24 01460 g001
Figure 1. (a) FMCW radar principle and (b) angle estimation based on the monopulse phase comparison method.
Figure 1. (a) FMCW radar principle and (b) angle estimation based on the monopulse phase comparison method.
Sensors 24 01460 g001
Sensors 24 01460 g002
Figure 2. Block diagram of the proposed 24 GHz 1TX-2RX FMCW radar transceiver.
Figure 2. Block diagram of the proposed 24 GHz 1TX-2RX FMCW radar transceiver.
Sensors 24 01460 g002
Sensors 24 01460 g003
Figure 3. (a) Block diagram and (b) parameter values of the FMCW synthesizer.
Figure 3. (a) Block diagram and (b) parameter values of the FMCW synthesizer.
Sensors 24 01460 g003
Sensors 24 01460 g004aSensors 24 01460 g004b
Figure 4. (a) Detailed schematics and parameter values of the VCO with the driving buffer, (b) the schematic of the three-bit cap bank, (c) layout of the triple-coupled transformer, (d) measured frequency range, and (e) measured phase noise.
Figure 4. (a) Detailed schematics and parameter values of the VCO with the driving buffer, (b) the schematic of the three-bit cap bank, (c) layout of the triple-coupled transformer, (d) measured frequency range, and (e) measured phase noise.
Sensors 24 01460 g004aSensors 24 01460 g004b
Sensors 24 01460 g005
Figure 5. (a) Block diagram and (b) timing diagram of the FSG.
Figure 5. (a) Block diagram and (b) timing diagram of the FSG.
Sensors 24 01460 g005
Sensors 24 01460 g006
Figure 6. Schematic of the three-bit attenuator.
Figure 6. Schematic of the three-bit attenuator.
Sensors 24 01460 g006
Sensors 24 01460 g007
Figure 7. Simulation results: (a) relative attenuation state and (b) return loss according to switch state.
Figure 7. Simulation results: (a) relative attenuation state and (b) return loss according to switch state.
Sensors 24 01460 g007
Sensors 24 01460 g008
Figure 8. Schematic of the two-stage DA and one-stage PA.
Figure 8. Schematic of the two-stage DA and one-stage PA.
Sensors 24 01460 g008
Sensors 24 01460 g009
Figure 9. Simulation results of the DA and PA: (a) gain and Pout versus Pin, (b) PAE, and (c) IP3.
Figure 9. Simulation results of the DA and PA: (a) gain and Pout versus Pin, (b) PAE, and (c) IP3.
Sensors 24 01460 g009
Sensors 24 01460 g010
Figure 10. Overall architecture of the RX.
Figure 10. Overall architecture of the RX.
Sensors 24 01460 g010
Sensors 24 01460 g011
Figure 11. (a) Schematic and (b) simulation results of the two-stage LNA, and (c) front end conversion gain without TIA.
Figure 11. (a) Schematic and (b) simulation results of the two-stage LNA, and (c) front end conversion gain without TIA.
Sensors 24 01460 g011
Sensors 24 01460 g012
Figure 12. (a) Schematic of the TIA with DCOC and (b) schematic of the one-stage BPF.
Figure 12. (a) Schematic of the TIA with DCOC and (b) schematic of the one-stage BPF.
Sensors 24 01460 g012
Sensors 24 01460 g013
Figure 13. Simulation results: (a) gain variation and (b) the HPF cut-off frequency variation.
Figure 13. Simulation results: (a) gain variation and (b) the HPF cut-off frequency variation.
Sensors 24 01460 g013
Sensors 24 01460 g014
Figure 14. Chip microphotograph.
Figure 14. Chip microphotograph.
Sensors 24 01460 g014
Sensors 24 01460 g015
Figure 15. Measured phase noise at 24 GHz carrier frequency.
Figure 15. Measured phase noise at 24 GHz carrier frequency.
Sensors 24 01460 g015
Sensors 24 01460 g016
Figure 16. Measured chirp transient when chirp bandwidth is 250 (black line), 500 (red line), and 750 (blue line) MHz: (a) triangular waveform with hold-chirp time, (b) sawtooth waveform with hold-chirp time, (c) triangular waveform without hold-chirp time, and (d) sawtooth waveform without hold-chirp time.
Figure 16. Measured chirp transient when chirp bandwidth is 250 (black line), 500 (red line), and 750 (blue line) MHz: (a) triangular waveform with hold-chirp time, (b) sawtooth waveform with hold-chirp time, (c) triangular waveform without hold-chirp time, and (d) sawtooth waveform without hold-chirp time.
Sensors 24 01460 g016
Sensors 24 01460 g017
Figure 17. Measurement results: (a) output power versus the operation frequency and (b) chirp spectrum.
Figure 17. Measurement results: (a) output power versus the operation frequency and (b) chirp spectrum.
Sensors 24 01460 g017
Sensors 24 01460 g018
Figure 18. Measurement results: (a) TRX loopback test, (b) the HPF cut-off frequency variation, and (c) gain variation when the HPF cut-off frequency was 500 kHz.
Figure 18. Measurement results: (a) TRX loopback test, (b) the HPF cut-off frequency variation, and (c) gain variation when the HPF cut-off frequency was 500 kHz.
Sensors 24 01460 g018
Sensors 24 01460 g019
Figure 19. (a) Implemented prototype FMCW radar module using the proposed transceiver and Yagi antenna and (b) stack up of the radar module.
Figure 19. (a) Implemented prototype FMCW radar module using the proposed transceiver and Yagi antenna and (b) stack up of the radar module.
Sensors 24 01460 g019
Sensors 24 01460 g020
Figure 20. (a) Design configuration, (b) simulation results of the Yagi antenna, and (c) simulated H-plane and E-plane beam patterns.
Figure 20. (a) Design configuration, (b) simulation results of the Yagi antenna, and (c) simulated H-plane and E-plane beam patterns.
Sensors 24 01460 g020
Sensors 24 01460 g021
Figure 21. Measurement setup.
Figure 21. Measurement setup.
Sensors 24 01460 g021
Sensors 24 01460 g022
Figure 22. Range measurement: (a) 2 m, (b) 4 m, (c) 6 m, (d) 8 m, (e) 10 m, and (f) 12 m.
Figure 22. Range measurement: (a) 2 m, (b) 4 m, (c) 6 m, (d) 8 m, (e) 10 m, and (f) 12 m.
Sensors 24 01460 g022
Sensors 24 01460 g023
Figure 23. Range measurement: distance error between real distance and measured distance.
Figure 23. Range measurement: distance error between real distance and measured distance.
Sensors 24 01460 g023
Sensors 24 01460 g024
Figure 24. Angle measurement: (a) 0 degree, (b) −30 degree, and (c) 30 degree.
Figure 24. Angle measurement: (a) 0 degree, (b) −30 degree, and (c) 30 degree.
Sensors 24 01460 g024
Table 1. Current consumption of sub blocks.
Table 1. Current consumption of sub blocks.
BlockCurrent ConsumptionUnit
FMCW Synthesizer40mA
1-channel TX70
2-channel RXs90
Supporting Circuit, LDRs20
Table 2. Performance comparison.
Table 2. Performance comparison.
BlockUnitThis Work[5][6][7][8][9]
Technology·65 nm CMOS130 nm SiGe130 nm SiGe130 nm CMOS130 nm CMOS55 nm CMOS
FrequencyGHz242424242424
The number of channel 1TX-2RX2TX-2RX1TX-3RX1TX-1RX1TX-1RX1TX-2RX
FMCW generation·Built-inExternal PLLExternal PLLExternal PLLExternal PLLBuilt-in
Modulation BW & MHz250200·250250250
Modulation Time & μs50150·200010002000
Integrated baseband·YesYesYesNoYesYes
Phase Noise @ 100 kHzdBc/Hz−80−111 +−100 +−105.4 +−84−71
TX output powerdBm13.3513−1.6511
RX maximum gaindB100876015.3 ^5664
Die sizemm24.84·91.5327.84
Current consumptionmA220160·110·344
+: The phase noise of the VCO at an offset frequency of 1 MHz. ^: LNA + Mixer. &: Chirp parameter value when the transceiver was in FMCW mode.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ko, G.-H.; Moon, S.-J.; Kim, S.-H.; Kim, J.-G.; Baek, D. Fully Integrated 24-GHz 1TX-2RX Transceiver for Compact FMCW Radar Applications. Sensors 2024, 24, 1460. https://doi.org/10.3390/s24051460

AMA Style

Ko G-H, Moon S-J, Kim S-H, Kim J-G, Baek D. Fully Integrated 24-GHz 1TX-2RX Transceiver for Compact FMCW Radar Applications. Sensors. 2024; 24(5):1460. https://doi.org/10.3390/s24051460

Chicago/Turabian Style

Ko, Goo-Han, Seung-Jin Moon, Seong-Hoon Kim, Jeong-Geun Kim, and Donghyun Baek. 2024. "Fully Integrated 24-GHz 1TX-2RX Transceiver for Compact FMCW Radar Applications" Sensors 24, no. 5: 1460. https://doi.org/10.3390/s24051460

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop