Matched circuit components are crucial for maximum efficiency in radio frequency (RF) systems, necessitating accurate load impedance monitoring. This paper proposes a method using reference point optimization with periodic structures for precise, low-overhead impedance measurements suitable for Built-in Self-Test (BIST). The novel technique uses a linear matching network and voltage detectors to sense load variations. A periodic structure measures voltages, and a multi-step algorithm selects optimal reference points to determine the load impedance.

RF systems, including RF MIMO RADARs, are increasingly integrated with digital systems in fine-geometry processes. Due to the prevalent use of RF MIMO RADARs in automotive and other safety-critical applications, in-field testing and tuning of these systems are needed to meet performance and safety targets. The fundamental performance targets of an RF MIMO system include the signal-to-noise ratio at the end of the receiver chain, matching characteristics between different signal paths, gain, noise figure, and linearity of the RF front end. In a RADAR device, matching between signal paths affects the angular resolution of the system. The gain and noise figure of the receiver control the maximum distance and the smallest object that the system can detect. In this work, we present a global tuning algorithm for RF MIMO RADARs to meet critical system performance targets while minimizing power consumption. The efficacy of the method is demonstrated with extensive simulations and hardware experiments.

Due to increasing performance demands, RF circuits are increasingly integrated into fine-geometry processes which necessitates their post-production and in-field calibration. One of the most important parameters for RF performance is matching between various nodes in the circuit. Process, voltage, and temperature (PVT) variations and other environmental conditions such as objects in the near-field of an antenna can cause shifts in the RF matching between sources and the loads. To compensate for such shifts, generally a tunable element or a network is incorporated between the source and the load, which will enable post-production calibration. In this paper, we present a tuning algorithm that adjusts the tuning structures until a match is detected via built-in measurements. The proposed technique can be used both for post-production and for in-field calibration. It only requires three strategically placed power detectors along the transmission line connecting the source to the load and simple computations. Experimental results based on simulations and hardware measurements show that the technique can provide highly accurate matching between a source and a load component.

Due to their simplicity, low power consumption, and high-performance, direct conversion transceivers are used widely in RF-front ends. Direct conversion receivers demodulate the signal to its in-phase (I) and quadrature (Q) parts by multiplying the RF signal with two signals with a 90°phase separation. To achieve best performance, I and Q paths should be matched in terms of gain, phase, and have no DC offset. Any impairment in these parameters would result in reduced performance and higher bit error rate. In this work, we propose a BIST scheme for IQ mismatch compensation with low overhead for direct conversion receivers. We use an envelope detector to detect the combined amplitude of the I and Q signals and we use this information to iteratively correct for the receiver’s IQ imbalance. The only requirement on the power detector is a small linear dynamic range (15dB). The conversion gain of the power detector does not need to be known. The proposed method can be used in the mission mode in the background to calibrate any deviations in performance. After the receiver’s IQ imbalance is corrected, it can be used to measure and compensate for any transmitter IQ imbalance in a loopback mode. Simulations and hardware measurements confirm that the proposed technique can measure the imbalances with high accuracy.

RF sensor technology in mm-wave range has improved significantly in recent years, which led to its widespread use in mission-critical systems, such as automotive radar. With the advent of multi-antenna systems, cascaded radar designs has gained prominence to increase resolution in terms of distance, speed, and angle. However, the performance of the radar systems can be highly dependent on dynamic conditions and they may require in-field calibration, specifically in terms of magnitude and phase mismatches of gain and input reflection coefficient. While in-field measurement of gain using built-in power sensors is more or less straightforward, measurement of the reflection coefficient requires the implementation of an N-port reflectometer on the chip. In this paper, we present an analytical model for noise and gain imperfections in N-port network analyzers. Based on this model, we propose a methodology for optimizing the design of the N-port reflectometer to obtain the highest accuracy under given realistic constraints, such as coupler gain/loss, splitter loss, power detector non-idealities, and noise.