imec recently announced what it claims is the world’s first CMOS 140GHz radar-on-chip system with integrated antennas in standard 28nm technology. This kind of advanced semiconductor technology is critical in the development of radar-based sensors for so many kinds of applications, some of which are building security, remote health monitoring of car drivers, breathing and heart rate of patients, and gesture recognition for man-machine interaction.
The latest RADAR designs are finding uses in automotive, also as sensors for contactless, non-intrusive interaction in internet-of-things applications such as people detection & classification, vital signs monitoring and gesture interfacing. imec’s research on 140GHz radar technology targets will enable new areas of adoption if they can achieve higher resolution, smaller size, better power-efficiency, and lower costs.
imec’s low-power 140GHz radar solution is composed of an imec-proprietary two-antenna SISO (Single Input Single Output) radar transceiver chip and a frequency modulated continuous wave phase-locked loop (FMCW PLL), with off-the shelf ADCs and FPGA, plus a Matlab chain. The transceiver also has on-chip antennas which achieve a gain close to 3dBi.
Excellent RADAR link budgets are supported because of the transmitter Effective Isotropic Radiated Power (EIRP) in excess of 9dBm with a receiver noise figure below 6.4dB. The total power consumption for transmitter and receiver is less than 500mW and can be further reduced by employing duty cycling.
The FMCW PLL enables fast signal edges up to 500MHz/ms over a 10GHz bandwidth around 140GHz with a slope linearity error below 0.5% and has a power consumption below 50mW. The FPGA enables real-time implementation of basic RADAR processing functions such as FFTs (Fast Fourier Transforms) and filters, and is complemented by a Matlab chain for detections, CFAR (Constant False Alarm Rate), direction-of-arrival estimation and other advanced radar processing.
Let’s take a look at some other recent achievements in related developments:
Design details for 110 to 140 GHz reconfigurable RADAR front-end IC with on-chip antenna1
A 110-140 GHz, reconfigurable RADAR front-end IC with on-chip antenna for use as a millimeter wave RADAR for automotive, robot control, remote sensing, and material characterization. The process used is a SiGe technology which enables low cost and integrated antennas with applications such as a millimeter imaging system to reconstruct a target’s 2D or even 3D image. Figure 1 shows the reconfigurable transceiver IC block diagram.
The block diagram of the reconfigurable transceiver IC (Image courtesy of Reference 1)
Three Marchand baluns1 were designed into the IC for single-to-differential signal conversion seen in Figure 1.
A wideband Low-Noise Amplifier (LNA) with a two-stage cascode topology was designed into the Rx path in order to compensate for the balun and switch losses. The LNA was also well shielded with an EM-shielded microstrip configuration. See Figure 2.
The shielded microstrip line structure in a cross-section view (Image courtesy of Reference 1)
Heterojunction Bipolar Transistor (HBT) IC technology was used in the integrated switches in this IC. This technology enables THz frequency performance.
Finally, the on-chip antenna in on-chip and designed to handle 120 to 140 GHz with 1.5 to 3 dB gain loss and 6.6 dBm PA output power capability with power-to-the-antenna of 3.2 dBm.
Ku-band FMCW RADAR on-chip for healthcare monitoring2
The design of this RADAR IC was an integrated Frequency Modulated Continuous Wave (FMCW) RADAR IC for real-time, wireless monitoring of patient health conditions using this low power, high sensitivity solution. See Figure 3.
The FMCW RADAR-on-Chip (RoC) architecture This RADAR architecture encompasses the transmitter and the receiver, that enables a 1 GHz chirp bandwidth at 15 GHz center frequency with 1.2 V supply and 238 mW power (Image courtesy of Reference 2)
The IC operates at Ku-band with a 15 GHz center frequency that enables accurate and continuous monitoring using interferometry phase analysis algorithms. The implementation of an on-chip digitally-tunable mixed-signal-mode FMCW chirp synthesizer ensures phase accuracy at high power efficiency. It is possible that both the RADAR and the phase analysis algorithms could be integrated into a single ASIC for even higher performance.
Because the physiological micromovements from an induced signal are very small with a low frequency, it is very difficult to directly detect via Doppler Frequency shift or Inverse Synthetic Aperture RADAR (ISAR) imaging.
It is evident that the phase information, the integration of frequency shift in the time domain, is very sensitive and correlated to the minute movements of objects, so that the phase information may be analyzed in the time domain to recover the corresponding small but critical physiological signals. The proposed interferometry phase analysis method will be able to accurately monitor these small movements based on the phase history of the signal, making it a powerful method for the real-time radar healthcare monitoring.
Some possible home healthcare applications are real-time diagnosis and observation of many kinds of health conditions including respiratory ailments, heartbeat disorders, and falls just to name a few.
An automotive mmW RADAR application: It’s all about the package3
New mmWave systems enable ADAS (Advanced Driver-Assisted Systems) for safety in automobiles with System on Chip (SoC) designs by Texas Instruments. The design is a fully-integrated 76-81 GHz one which uses a 45nm RFCMOS process for an automotive transceiver in a 0.65 mm pitch, 161-pin, 10.4mm x 10.4mm flip-chip BGA package. This is the TI AWR1243.
In this design, mmWave frequencies in a complex process integration design, give way to electromagnetic mayhem at the system level which is made up of the Silicon, package, and the PC board. A very clever and unique package/system design was developed using a coupled circuit-to-electromagnetic codesign modeling and simulation analysis. See Figure 4.
At these very high frequencies, we know that bondwires, vias, solder bumps, and solder balls on the package can cause impedance discontinuity and make it virtually impossible to reach the insertion loss target of greater than -2dB from DC to 80GHz. But, it is possible to achieve the goal within a relatively narrow bandwidth. The method is to recognize and take advantage of the resonance phenomenon that occurs naturally between the bondwire inductance and via/ball capacitance. See Figure 5.
Controlling the resonant frequency of a FCBGA signal channel (Image courtesy of Reference 3)
Enter another package choice, the nFCBGA package. This type of package is a laminate-based Chip Scale Package (CSP) that uses bond wire as the interconnect between die and the package substrate while connections between the package substrate and the board are made with alloy balls.
For power integrity and coupling analysis, the extraction and optimization of parasitics for the critical power/ground supply network were performed. These included both the analog and digital supplies. Of significant note, the intra- and inter-coupling between the serial interfaces/clocks were modeled and optimized to achieve desired requirements.
Since the electromagnetic interactions at the interfaces will impact performance significantly, TI needed to find a comprehensive modeling/analysis co-design methodology which will be successfully applied to characterize the performance of the device early on in the design phase.
In the package technology selection, looking at QFN, nFBGA, and FCBGA, the optimization of the package and system was achieved through a coupled circuit-to-electromagnetic co-design modeling and simulation methodology, especially in High Volume Manufacturing (HVM) in these extreme high frequency designs. TI was able to perform parametric sweeping of the key performance figure of merits under manufacturing/assembly process variations.
1 110-140 GHz single-chip, reconfigurable RADAR Front-end with on-chip antenna, Shuai Yuan, Hermann Schumacher, Institute of Electron Devices and Circuits, Ulm University, IEEE 2015
2 A Ku-band FMCW Radar on Chip for Wireless Micro Physiological Signal Monitoring by Interferometry Phase Analysis, Zhongyuan Fang, Liheng Lou, Chuanshi Yang, Kai Tang, and Yuanjin Zheng, Virtus IC Design Center of Excellence, IEEE 2018
3 Package Co-Design of a Fully Integrated Multimode 76-81GHz 45nm RFCMOS FMCW Automotive Radar Transceiver, Minhong Mi, Meysam Moallem, Jie Chen, Ming Li, and Rajen Murugan, Texas Instruments, 2018 IEEE 68th Electronic Components and Technology Conference, IEEE 2018