Many applications in today's market require accurate measurement capabilities of depth sensing, including industrial, consumer and automobile fields. The products include automatic guided vehicles (AGV), sweeping robots, unmanned aerial vehicles and self-driving cars, which have great market development potential. This article will show you the principle of depth sensing technology, the reference design of LiDAR introduced by Arrow Electronics, and the product characteristics of key devices from ROHM, onsemi, ADI and Murata used in this solution.
Depth sensing technology with its own advantages and disadvantages
There are many different methods in applying depth sensing, which include stereo triangulation using standard CMOS image sensor, phase detection pixel and structured light.
The distance can be measured using stereo triangulation by triangulating the received light from two different cameras. By comparing the differences of object positions between images taken by cameras, the distance between cameras and objects can be calculated. Stereo triangulation has the advantages of the passive method and the standard image sensor, but it also has some disadvantages, such as the need to use two cameras, the maximum distance depending on the distance between cameras, the high dependence on lighting conditions, and the need the computational cost.
The phase detection pixel method uses a single camera to obtain the distance of the points in the scene. At the pixel level, the image sensor calculates the depth by the phase difference of the received light from pairs of pixels with light shields at different positions, or by using multiple photodiodes under the same microlens, such as the autofocus of an iPhone camera. The phase detection pixel method has the advantages of the passive method and the standard image sensor, but it also has some disadvantages, such as poor depth resolution, high dependence on light conditions, computational costs and short distance.
Structured light uses a camera with a traditional CMOS image sensor to analyze a pattern of received infrared light and calculate the depth with the distortion in the scene. The distortion of the pattern can be used to obtain the 3D shape of the object. Structured light has the advantages of being suitable for short distances, but it also has the disadvantages of active method and sensitive to ambient light, and the depth error increases with distance, and though it is not suitable for long distance, it can be utilized for face recognition.
LiDAR has a high depth sensing capability
Because of its high depth and angular resolution, LiDAR has higher depth sensing abilities than the alternative method, and because of its active method of using infrared transmitter and receiver, it can operate under all lighting conditions. LiDAR has been widely deployed in many different markets for various applications and usage cases, including automotive, industrial, robotics and consumer augmented reality and virtual reality (AR/VR) applications.
In general, LiDAR will adopt direct time-of-flight (dToF) measurement technique, which calculates the time delay between the transmitted signal and its return echo(es), and the other method is indirect time-of-flight (iToF), both of which can be implemented by pulse or continuous modulation.
LiDAR solution for pulsed ToF system
In order to speed up the development of customers' products, the Openlab of Arrow Electronics launched the reference design of pulsed dToF LiDAR solution. Aiming at the range/distance measurement, the LiDAR ToF solution of Arrow Electronics adopts the pulsed ToF system. The signal processing part of the LiDAR ToF system adopts the time-to-digital converter (TDC) or analog-to-digital converter (ADC) distance estimation method. The TDC-based method uses high-precision clocking devices to count the start/stop events as time differences, while the ADC-based method at regular intervals measures and digitizes the return signals, and then estimates the time differences.
The range resolution of this LiDAR ToF system is inversely proportional to the combined rise time and response time of analog parts (laser diode, laser driver, low noise amplifier and photodetector). The TDC-based method can solve the resolution problem in the analog domain, while the ADC-based method can solve some problems through some complex digital return signal detection schemes, baseband systems, and software.
The LiDAR ToF solution of Arrow Electronics chooses the TDC-based method, and pays more attention to analog hardware design to achieve better rise time and response behavior, so as to reach the best ranging application in the LiDAR solution of Arrow Electronics.
When using the pulsed ToF system, the wavelength of 905nm will be used, because the 905nm system (infrared) is more suitable for the maximum optical power up to 75 W. On the other hand, 650nm laser (visible red light) can't generally realize power-on pulse operation, and the maximum optical power is about 100mW.
Narrow pulse operation can expand the operating range
In the hardware design of the LiDAR ToF solution of Arrow Electronics, the ROHM scheme is adopted to achieve the optimal performance of LiDAR in the analog domain. In order to shorten the laser trigger pulse, ROHM RLD90QZW3 pulsed laser diode is used in the LiDAR ToF solution of Arrow Electronics, which can support narrow pulse operation with a pulse width of 15 nanoseconds, while the pulse width commonly used in the traditional LiDAR solution is 30 nanoseconds. By reducing the pulse width by 50%, higher optical power can be provided under the same operating conditions, thus expanding the operating range.
By using this short pulse width, the LiDAR ToF solution of Arrow Electronics supports multi-pulse operation, which can improve the measurement accuracy and eliminate environmental noise and interference by averaging or statistical analysis of multiple measurements. This solution also uses GaN FET in the laser diode to achieve faster switching, thus further improving the transmission delay efficiency. GaN FET transistor is used to replace the traditional MOSFET transistor, which can provide 10 times faster switching behavior, thus shortening the rise time of laser transmission path.
This solution also strengthens the PCB layout to further reduce the delay time of the laser driver section. PCB layout plays an important role in the switching behavior of laser transmission path, especially in such multi-power supply systems, and it is necessary to provide 25V for laser diode and GaN FET, 5V for laser gate driver, and 3.3V for MCU system to generate a LD trigger pulse. In addition, the design of GND plate is also very important for fast switching and optimization of transmission delay by using the best signal return path.
In addition, this solution also uses onsemi’s RD-series silicon photomultiplier (SiPM) to replace the traditional avalanche photodiode (APD) to further improve the Rx response time. The FAST OUT terminal in SiPM can provide a rise time response of less than 500 picoseconds, 50% lower than the APD standard output terminal.
In the Tx and Rx path detector systems, faster comparators can be used to further improve the Rx response time. The comparator circuit is used to convert the analog Rx and Tx signals into TDC pulse start and pulse stop inputs for timing calculation, so the propagation delay of the comparator is also very critical to the measurement accuracy. Using ADI's ADPCM600 fast comparator, the delay time is only 3 nanoseconds at 30mV input signal level, thus providing the best delay time in the LiDAR receive path.
Industry-leading devices make up a complete solution
The key components of the whole LiDAR ToF solution of Arrow Electronics include ROHM RLD90QZW3 laser diode as 75W 905nm invisible light pulse laser diode, SiPM MicroRD-10035-MLP RD series of onsemi, and ADI ADCMP600 high-speed comparator, which is an extremely fast TTL/CMOS comparator with a propagation delay of 5.5 nanoseconds. ADI HMC589AST89E high-speed amplifier is an InGaP HBT gain module MMIC amplifier (DC-4GHz).
In addition, there are ADI LT®8330 DC/DC boost converter with wide input voltage, which can support 3V to 40V input voltage, 1A and 60V switching boost converters, ADI LT3082 200mA low noise low dropout regulator (LDO), NXP LPC546XX Series 32-bit ARM Contex-M4 microcontroller, and TI TDC7201 TDC and Murata WMRAG32K76CS1C00R0 32.768 kHz MEMS resonator. For development that requires extra processing power or machine learning.
ROHM laser diode RLD90QZW3, which is a 75W infrared high optical output laser diode specially designed for LiDAR, used for distance measurement and spatial recognition services in 3D ToF systems for AGV and other applications. ROHM uses its own device development technology to achieve an unprecedented emission width of 225μm under equivalent optical output, which is 22% narrower than conventional products and with improved beam characteristics.
At the same time, the uniform emission intensity and the low temperature dependence of laser wavelength ensure performance stability, which will help to achieve higher accuracy and longer distance in various LiDAR applications. In addition, it has the same power conversion efficiency of 21% as the standard products (24A forward current and 75W output) (trade-off with narrow emission width), and can be used without increasing power consumption.
ROHM has also introduced reference designs that can be used for laser diode control, including their next-generation device EcoGANTM, the GaN gate driver BD2311NVX-C, and the GaN HEMT high-speed gate driver that can help improve the characteristics (distance and resolution) of the LiDAR sensor.
onsemi SiPM is a high-gain, single-photon sensitive sensor, which is used to detect visible light to near infrared wavelength. The RB series sensor introduced by onsemi is the second release of SiPM in R-series. These sensors further improve the sensitivity in the red and near infrared (NIR) regions of the electromagnetic spectrum.
onsemi has also introduced a new SiPM array series, ArrayRDM−0112A20−QFN, which is a monolithic 1 × 12 array of SiPM pixel based on market-leading RDM process. The RDM process is specially developed to create products with high PDE (Photon Detection Efficiency) at 905/940nm NIR wavelength, which are usually used in LiDAR and 3D dToF ranging applications.
onsemi's SiPM array is packaged in a robust QFN package and can access to the 12 individual pixels. In order to meet the requirements of automotive LiDAR application, the product meets the AEC-Q102 standard.
NXP LPC546xx Arm Cortes M4 MCU family is used in the LiDAR system that provides up to 220 MHz performance, Ethernet support, a TFT LCD controller and two CAN FD modules, while striking the right balance between feature integration and power efficiency with the Cortex-M4 achieving an active mode current of 100 µA/MHz.
For development that required more advanced computing or machine learning, NXP also offered new MCX N Series with dual Arm® Cortex®-M33 cores operating up to 150 MHz. This advanced series introduces the new instantiation of the NXP designed Neural Processing Unit (NPU). The integrated NPU delivers up to 30x faster machine learning (ML) throughput compared to a CPU core alone enabling it to spend less time awake and reducing overall power consumption. The low power cache enhances system performance and the dual bank Flash and full ECC RAM support system safety providing an extra layer of protection and assurance.
The ADCMP600, ADCMP601 and ADCMP602 are very fast comparators manufactured by ADI's proprietary process XFCB2. The device provides 5 ns propagation delay and 10 mV overdrive at a typical supply current of 3 mA. It is very suitable for time-critical applications, such as TOF measurement and LiDAR applications.
ADI's LT8330 is a current-mode DC/DC converter that can generate a positive or negative output voltage using a single feedback pin. It can be configured as a boost, SEPIC or inverting converter, and the quiescent current consumption is as low as 6A. Low ripple burst mode operation can maintain high efficiency at very low output currents, while keeping the output ripple below 15mV in typical applications.
The in-vehicle chip multilayer ceramic capacitor introduced by Murata is an ideal choice for automobile power system and safety device. This product can be used to control the drive system of safety devices, such as engine ECU, airbag and ABS. Even in temperature cycle and humidity load tests, the product passed stricter test conditions than the general product (GRM series). The ceramic capacitor can be used at the temperature of 125°C and 150°C, and also provides a series of products at 150°C which can be used in the engine room. In addition, the ceramic capacitor also adopts external electrode Sn plating, which has excellent solderability.
Conclusion
With the application of depth sensing becoming wider, LiDAR has gradually dropped to an acceptable price range, which has gradually expanded the market development space of related products. The LiDAR ToF solution introduced by Arrow Electronics adopts industry-leading devices from ROHM, onsemi, ADI, Murata and so on, guarantees excellent performance, and is worthy of further understanding and adoption by manufacturers interested in developing related products.