At the beginning of the new year in 2020, the raging of the new crown epidemic has spawned 5G applications such as telemedicine, online learning, video conferencing, and online office. 5G application scenarios such as large bandwidth, low latency, and wide connectivity are gradually enriched and expanded. As the basis for 5G applications, the new 5G network infrastructure is still advancing in a tight and orderly manner despite the impact of the epidemic, both domestically and abroad. China Academy of Information and Communications Technology’s Thiel System Laboratory (hereinafter referred to as “China Academy of Information and Communications Technology’s Thier System Laboratory”) has been actively tracking and participating in the work of 3GPP, ITU and other international standardization organizations for a long time, and has led the formulation of many domestic 5G industry standards. After continuous construction and preparation in 5G standard formulation, technical personnel training, testing equipment and environmental facilities, the testing capabilities of system equipment such as 5G base stations have formed a scale.
Changes in 5G base stations relative to 4G base stations
In terms of working frequency bands, according to the 3GPP 38.101 agreement, 5G base station equipment mainly uses FR1 frequency band (450MHz-6GHz) and FR2 frequency band (24.25GHz-52.6GHz), the former is also called Sub-6GHz frequency band. Since most of the frequency bands covered by FR2 are frequencies with a wavelength of less than 10 millimeters, it is also called the “millimeter wave” frequency band. The two millimeter wave frequency bands of 24.25-27.5GHz and 37-43.5GHz are the frequency bands supported by most countries. At present, in China, in the initial stage of new infrastructure construction, 5G networks are mainly built in the Sub-6GHz frequency band. In foreign countries, some operators use the millimeter wave frequency band to build 5G networks.
In terms of base station form, 4G base stations are mainly composed of BBU, RRU and passive antennas, while 5G base stations have become BBU (CU (centralized unit) + DU (distributed unit)) and AAU (RRU + antenna feeder system, that is, active antenna). , 4 station types are defined in 3GPP: conduction type (BS type 1-C), hybrid type (BS type 1-H), and air interface type (BS type 1-O, type 2-O).
In terms of key technologies, 5G base stations have introduced massive MIMO technology and beamforming technology. Massive MIMO technology greatly improves network capacity and signal quality through more antennas. 5G base stations using massive MIMO technology can not only improve network capacity by multiplexing more wireless signal streams, but also greatly improve network coverage through beamforming. The beamforming technology adjusts the spatial distribution of the antenna gain, so that the signal energy is more concentrated to the target terminal during transmission, so as to compensate for the loss of spatial transmission after the signal is sent, and greatly improve the network coverage capability.
5G base station core detection capability of Tell System Laboratory
As the lead unit, the China Academy of Information and Communications Technology’s Taier System Laboratory led the completion of the YD/T 2583.17-2019 “Electromagnetic Compatibility Requirements and Measurement Methods for Cellular Mobile Communication Equipment Part 17: 5G Base Stations and Their Ancillary Equipment” 5G Base Station Electromagnetic Compatibility Standard This standard is the world’s first electromagnetic compatibility standard for 5G base stations and has become a reference document for my country’s network access test. In addition, the laboratory-led industry standard “Requirements and Measurement Methods for Air Interface Electromagnetic Characteristics of 5G Active Integrated Systems” has been established, which will continue to assist in the testing and certification of electromagnetic characteristics of 5G system equipment.
At the same time, the Tel System Laboratory of the China Academy of Information and Communications Technology has innovatively introduced a compact field test system to complete the OTA radio frequency test capability building of 5G base station transmitters and receivers, especially the millimeter wave radio frequency OTA test capability construction that is currently the most technically difficult. The laboratory completed the CNAS expansion work of the latest 3GPP-related standards at the fastest speed, meeting the development needs of the industry for the first time. Make the laboratory the first 5G testing laboratory in China to pass the CNAS certification.
Relying on this compact field test platform, we have completed the first batch of 5G base station network access tests for a number of major equipment manufacturers, and provided customers with professional technical support in domestic certification testing, product development testing and other aspects. According to the latest 3GPP standards, completed the overseas commissioned test of an equipment manufacturer’s mid-low frequency 5G base station and millimeter-wave base station. The test reports issued have been recognized by foreign operators, and the company participated in the operator’s 5G construction bidding and expanded overseas markets. , the rapid resumption of work and production during the epidemic has provided strong support.
In terms of electromagnetic radiation of 5G base stations, the China Academy of Information and Communications Technology’s Taier System Laboratory is entrusted by the Ministry of Industry and Information Technology, the Ministry of Ecology and Environment and other relevant ministries and commissions, and cooperates with the three major operators and domestic main equipment manufacturers to carry out 5G base station electromagnetic radiation monitoring and research, with the support of the Ministry of Ecology and Environment. my country’s first electromagnetic radiation environment monitoring standard for 5G base stations, “5G Mobile Communication Base Station Electromagnetic Radiation Environment Monitoring Method”, jointly drafted by the unit, has been completed and will be released in due course. This research proposes innovative technological breakthroughs in traditional radiation monitoring methods and brings new ideas for electromagnetic radiation environmental protection.
In addition, the Tel System Laboratory of the China Academy of Information and Communications Technology also has the ability to detect electromagnetic compatibility, electromagnetic radiation, electrical safety and network quality of 5G base station system equipment. For 5G-related upstream and downstream industries and infrastructure, conduct 5G array antenna passive radiation characteristics, circuit characteristics and environmental reliability tests. 5G base station AAU radome mechanical performance, material performance, radiation performance test. At the same time, it has the ability to detect 5G access network related equipment, including on-board lightning protection products for 5G base stations, optical cables and optical modules for 5G, smart tower detection services for 5G, 5G integrated Power Supply system, 5G meter measurement test, etc. Realize one-stop testing services for 5G industry.
5G base station RF OTA test
5G Antenna Test
5G Base Station Electromagnetic Radiation Detection and Evaluation
5G Base Station Electromagnetic Compatibility Test
5G testing business coverage of Tell System Laboratory
With the gradual expansion of the global market, while promoting the development of my country’s 5G industry, it is also necessary for Chinese manufacturing to move to overseas markets. The China Academy of Information and Communications Technology’s Taier System Laboratory launched the certification and access commissioned testing of 5G system equipment in countries along the “Belt and Road”. At present, the test reports of 5G base stations, antennas and other products issued by the laboratory have been recognized by operators in Thailand, Indonesia and other countries, helping more Chinese manufacturers to go abroad and enter the international market. China Academy of Information and Communications Technology Taier System Laboratory adheres to the concept of “customer-centric, creating value for customers, adhering to market orientation and problem orientation, and seeking breakthroughs through innovation”, and will continue to improve its own testing capabilities in the construction of new 5G infrastructure projects , pursue innovation, and strive to provide one-stop service for 5G new infrastructure.
“The most striking feature of TWS earphones is the ease of wearing them wirelessly. Compared with traditional Bluetooth headsets, TWS headsets have many advantages such as small size, good sound quality, and high stability, as well as certain waterproofness and intelligence, which quickly attracted the attention of consumers. At present, the shipment volume and overall market size of TWS earphones are constantly expanding, and it is a hot research and development field of consumer electronics.
“
The most striking feature of TWS earphones is the ease of wearing them wirelessly. Compared with traditional Bluetooth headsets, TWS headsets have many advantages such as small size, good sound quality, and high stability, as well as certain waterproofness and intelligence, which quickly attracted the attention of consumers. At present, the shipment volume and overall market size of TWS earphones are constantly expanding, and it is a hot research and development field of consumer electronics.
system structure
The 1-Wire TWS headset solution MAXREFDES1302 introduced in this article includes two parts: the charging box and the headset. The overall hardware architecture of the system is shown in Figure 1.
Figure 1. 1-Wire TWS charging case and headphone system architecture.
The charging box uses a 3.7V 1500mAh single-cell lithium battery to power the system, and the charger MAX77651 that supports the USB Type-C protocol is used to charge the lithium battery. Users only need to use a single USB Type-C data cable to charge the whole machine . In terms of power rails, the charging box uses the MAX17224 boost module to boost the system voltage of the charger to 5V. The 5V voltage is generated by the MAX38640 step-down module to generate 3.3V to supply power to the microcontroller MAX32655. At the same time, the 5V voltage is also passed through The 1-Wire control circuit is transmitted to the headphone, which acts as a charging power source for the headphone system.
In terms of power monitoring, the charging box uses the MAX17262 fuel gauge with a built-in current-sense resistor to monitor the battery. The fuel gauge combines traditional coulomb counting methods with the innovative ModelGauge™ m5 EZ algorithm, eliminating the need for battery characterization, flexible configuration, and ease of use. In terms of microcontroller, the charging box adopts the microprocessor MAX32655 with BLE 5.2 module and built-in SIMO power module. Read and write control to the DS2488 on the headphone side, providing great convenience for 1-Wire communication and charging. The SWD interface of the charging box can be connected to the MAX32625PICO downloader, which can update the firmware of the MAX32655 of the charging box and Display battery information on the computer through the virtual serial port. Information about the battery can also be displayed through the OLED screen on the charging case.
The headset uses a 3.7V 130mAh single-cell lithium battery to power the system, and uses a DS2488 bidirectional 1-Wire bridge to realize the data interaction between the headset and the charging box, and to control the 5V charging power source from the charging box. On the controller side, the headset also uses the MAX32655 as a microcontroller, which uses the UART interface to simulate 1-Wire timing to read and write to the DS2488, and also uses the SWD interface to connect to the MAX32625PICO downloader to download programs. In terms of power rails, a 3.3V LDO output of the charger MAX77734 used by the headset supplies power to the microcontroller MAX32655. At the same time, the 3.3V and the 1.8V and 1.2V power supplies generated by the built-in SIMO module of the MAX32655 together form an audio encoder. Power rail for the decoder MAX98050. In terms of power monitoring, the headset also uses the fuel gauge MAX17262 to monitor the battery.
Figure 2 is a physical image of the 1-Wire TWS charging case and earphones. The actual size of the charging box is 10.20cm × 5.80cm, and the actual size of the earphones is 10.20cm × 6.50cm. Since this design is a prototype to assist customers in design, testing and research, the actual product size can be obtained by simplifying the test points. It can be greatly compressed to meet the size requirements of the actual application of TWS earphones.
Figure 2.1-Wire TWS charging box and headphone PCBA physical map.
1-Wire data communication and energy transfer
In the TWS headset application, it is very important to realize the data communication and energy transmission between the charging box and the headset in a reliable and convenient way. The common TWS earphones currently on the market usually use 3 or more contacts to connect with the charging box to realize the functions of data communication and energy transmission. However, too many contacts usually lead to an increase in system cost, which is extremely detrimental to low-cost wearable product design. In addition, more contacts usually require more space, which goes against the small size requirement of TWS earphones. Additionally, more contacts tend to increase the likelihood of failure. This design uses ADI’s proprietary 1-Wire bidirectional bridge DS2488 designed for the TWS solution to achieve energy transmission and data interaction between the headset and the charging box. The DS2488 supports the 1-Wire bus protocol, enabling communication and charging with a single wire. Since the system requires an extra contact to connect the earphone and the ground of the charging box, the overall solution only needs to use two contacts, which can greatly improve system reliability and reduce size and cost. The block diagram of the 1-Wire communication charging circuit used in this design is shown in Figure 3.
Figure 3.1-Wire communication charging circuit block diagram.
How DS2488 works
As shown in Figure 3, DS2488 is a 1-Wire bidirectional bridge with two 1-Wire communication pins, IOA and IOB, for the microcontrollers on both sides to control. IOA is controlled by the microcontroller of the charging box, and IOB is controlled by the headset microcontroller control. The IOA supports input voltages up to 5.5V and supports different communication and charging levels on the 1-Wire bus (IOA). As a 1-Wire Device, each DS2488 device also has a unique 64-bit ROM ID for user identification and authentication. There is also an 8-byte buffer inside the DS2488, which can be read and written by the microcontroller to update the battery information on both sides of the storage in real time. In this design, the information stored in the buffer is shown in Table 1.
Table 1. Information Stored in the DS2488 Buffer
The TOKEN pin of DS2488 indicates the control status of DS2488: TOKEN is low, indicating that the microcontroller of the charging box obtains the control authority of DS2488; TOKEN is high indicating that the microcontroller on the earphone side obtains the control authority of DS2488. The CD/PIOC pin of the DS2488 controls whether the charging box charges the headset: when the voltage on the 1-Wire bus (IOA) is less than 4V, the CD/PIOC is in a high-impedance state, the transistor is turned off, and charging stops; when the 1-Wire bus (IOA) is in a high-impedance state When the voltage on (IOA) is greater than 4V, CD/PIOC is low, the transistor is turned on, and the voltage on the 1-Wire bus (IOA) is directly applied to the charger of the headset, and charging starts. The selection logic of earphone charging and communication is mainly realized by a MOSFET connected to 5V. The on-off of the MOSFET is controlled by the microcontroller of the charging box. The use of the charging box and the earphone is mainly divided into the following situations.
The headset is in the charging case and the charging case lid is open
At this time, the microcontroller of the charging box turns off the MOSFET and obtains the control authority of the DS2488, TOKEN is low, and CD/PIOC is in a high-impedance state. The charging box reads and writes the internal 8-byte buffer of the DS2488 through the IOA, reads the byte information of the earphone battery, and updates the byte information written to the charging box battery. At this time, charging stops and communication is performed.
The headset is in the charging case and the charging case lid is closed
At this time, the microcontroller of the charging box turns on the MOSFET, and 5V is directly transmitted to the earphone through the 1-Wire bus (IOA). At this time, TOKEN is high and CD/PIOC is low. The 5V voltage of the charging box is transmitted to the earphone side to charge the lithium battery of the earphone. At the same time, the microcontroller of the headset obtains the control authority of the DS2488, reads and writes the internal 8-byte buffer of the DS2488 through the IOB, updates the byte information written to the headset battery, and reads the byte information of the charging box battery. At this time, communication stops and charging is performed.
The headset is not in the charging case or the charging case battery is dead
At this point, the 1-Wire bus (IOA) is in a high-impedance state, TOKEN is high, and CD/PIOC is in a high-impedance state. At this time, the microcontroller of the headset obtains the control authority of the DS2488, reads and writes the internal 8-byte buffer of the DS2488 through the IOB, and updates the byte information written into the battery of the headset.
DS2488 1-Wire Data Communication
As mentioned above, this design uses the DS2488 as a bridge between the charging box and the microcontrollers on both sides of the headset to realize data interaction between the microcontrollers on both sides. DS2488 supports typical 1-Wire communication protocol. The sequence of the protocol is divided into reset and response sequence and read and write sequence. The read and write sequence is divided into write 0 time slot, write 1 time slot and read time slot, as shown in Figure 4 and Figure 5 shown. The detailed data of the time range of each timing high and low level phase can refer to the DS2488 data sheet.
Figure 4. DS2488 1-Wire reset and response timing.
Figure 5. DS2488 1-Wire read and write timing.
All 1-Wire devices are internally composed of state machines, and the state transition diagram is shown in Figure 6. As shown in Figure 4, when the microcontroller sends a reset signal to the DS2488 device, the 1-Wire bus will be pulled low for 48μs to 80μs, and then the bus will be pulled high and released by the pull-up resistor. If the DS2488 is connected to the bus, the DS2488 will respond to this reset signal by pulling the 1-Wire bus low again for 6 to 10 μs after the bus is released 48 μs. At this time, the microcontroller can detect the level change on the bus, that is, whether there is a DS2488 connected to the 1-Wire bus by detecting whether the bus is pulled low again.
Figure 6. State transition diagram for a 1-Wire device.
When the DS2488 responds to the reset signal, the microcontroller will send the ROM Function Command. The ROM function commands are the same for all 1-Wire devices, and some common ROM function commands are shown in Table 2. Due to the design of TWS earphones, two earphones usually need to be accommodated in the charging box, so two DS2488s are usually connected to the 1-Wire bus (IOA). This design first uses the Read ROM command (0x33) and the Match ROM command (0x55) to read the ROM IDs of the two DS2488s on the 1-Wire bus (IOA) and the DS2488 device matching the specific ROM ID respectively, to realize the identification and identification of the left and right earphones. strobe.
Table 2. Common 1-Wire ROM Function Commands
After sending the ROM function command, the microcontroller will send the Device Function Command to perform further operations on the device. Different 1-Wire devices have different device function commands. For the DS2488, some common device function commands are shown in Table 3. In this design, the Write Buffer command (0x33) and the Read Buffer command (0x44) are used to read and write the 8-byte buffer inside the DS2488 to realize the interaction between the charging box and the earphone battery information.
Table 3. Commonly used DS2488 device function commands
The two sets of GPIOs (P0.6 and P0.7, P0.18 and P0.19) of the microcontroller MAX32655 of the charging box can be configured as the OWM_IO pin and OWM_PE pin of the 1-Wire module, respectively, which are the same as those of the DS2488. communication and 5V transmission. This design connects the OWM_IO pin of the MAX32655 to the IOA pin of the DS2488 to implement 1-Wire communication between the charging box and the DS2488.
The difference is that, considering that some microcontrollers on the market do not have 1-Wire interface, for the convenience of user design, the microcontroller MAX32655 of the headset uses UART interface to simulate 1-Wire timing, and communicates with DS2488 through IOB. As shown in Figure 3. The microcontroller does this by configuring a specific UART baud rate and sending a specific pattern. Taking the reset and response sequence shown in Figure 4 as an example, when the baud rate is 115200, the time length for the UART to send and receive each bit of data is about 8.68 μs. Therefore, the time length of 1 byte (8 bits) of data is about 69.44 μs, and 0xE0 (binary: 11100000) (UART sends low-order data first) corresponds to the timing of the 1-Wire reset signal. At this time, if the microcontroller sends 0xE0 (reset signal) through TX, the DS2488 on the 1-Wire bus (IOB) will respond to this reset signal and pull the bus down for 6μs to 10μs, and the signal received on RX should be 0xC0 (binary: 11000000) or 0x80 (binary: 10000000). The microcontroller can achieve the function of simulating 1-Wire timing through the UART by sending and receiving different patterns and comparing the received and sent signals.
DS2488 1-Wire Power Delivery
As shown in Figure 3, the OWM_PE pin of the microcontroller MAX32655 of the charging box controls the on-off of the MOSFET. When the MOSFET is off, the system performs 1-Wire communication; when the MOSFET is on, the 5V voltage passes through the 1-Wire bus ( IOA) is transmitted to the earphone side. At this time, the DS2488 detects 5V, and the CD/PIOC pin changes to a low level to turn on the transistor, and the 5V voltage is transmitted to the charger to charge the lithium battery of the earphone.
Battery Management and Power Configuration
The battery management and power configuration system of the charging box consists of the USB Type-C charger MAX77751, fuel gauge MAX17262, step-up DC/DC converter MAX17224 and step-down DC/DC converter MAX38640. Typically, the end-of-charge voltage of a single-cell lithium battery is 4.2V, so the MAX77751CEFG+ is chosen as the specific charger model. The charging current of this charger is configured by the resistors connected to the IFAST pin and the ITOPOFF pin. Considering the actual needs, a fast charging current of 500mA and a termination current of 100mA are selected, and the corresponding resistances are 2.4kΩ and 8.06kΩ, respectively. The fuel gauge MAX17262 has the ModelGauge m5 EZ algorithm, which can automatically measure the battery after configuring battery parameters such as battery capacity, termination current, and charging voltage threshold, without additional battery modeling. The output voltages of the step-up DC/DC converter MAX17224 and step-down DC/DC converter MAX38640 are both configured by resistors connected to the SEL and RSEL pins, where 0Ω and 56.2kΩ are chosen to output 5V and 3.3V, respectively .
The headset’s battery management and power configuration system consists of the MAX77734 charger and the MAX17262 fuel gauge. The SIMO output of the microcontroller MAX32655 also provides both 1.8V and 1.2V power rails for the system. Since only one 3.3V LDO output is required, the specific model of the charger is MAX77734GENP+. The charger can also be configured via I2C into factory shipping, shutdown and standby modes to extend battery life. The microcontroller MAX32655 provides four SIMO outputs, each of which can be configured to output a different voltage through registers.
Firmware Design
The firmware flow chart of the charging box is shown in Figure 7. After power up, the charging box’s microcontroller will initialize the GPIOs and configure the fuel gauge MAX17262 and the OLED module. The microcontroller then polls the status of the charging case cover. If the charging case door is closed, the microcontroller will disable the 1-Wire module and apply 5V to the 1-Wire bus (IOA) to charge the headset. In this state, if the microprocessor detects that the remaining capacity of the battery in the charging case is less than 5%, charging will stop. If the charging case door is open, the microcontroller disables the 5V charging voltage and enables the 1-Wire module to read and write to the DS2488’s buffer. The battery information of the charging box and the headset is displayed through the OLED module or virtual serial port.
Figure 7. The charging case firmware flow chart.
The firmware flow chart of the headset is shown in Figure 8. After power-up, the headset’s microcontroller initializes the GPIOs, configuring the MAX17262 fuel gauge and the MAX77734 charger. The microcontroller then polls the charger for a valid input voltage. If the input voltage is valid and greater than 4V, the microcontroller enables the charger and begins charging. At this point, the microcontroller polls the status of TOKEN, and if TOKEN is low, the charging box has read and write permissions for the DS2488. If TOKEN is high, the headset has the read and write permission of DS2488, and the microcontroller writes the battery information of the headset into the buffer of DS2488 for the charging box to read.
Figure 8. Headphone firmware flowchart.
Test Results
The design requirements and test results of the power rails for the charging case and earphones are shown in Tables 4 and 5. It can be seen that this design can meet the design requirements of the system.
Table 4. Design Requirements and Test Results for Charging Box Power Rails
Table 5. Design Requirements and Test Results for Headphone Power Rails
The test results when the charging box cover is closed and when the charging box cover is open are shown in Figure 9 and Figure 10. It can be seen that this design can Display the information of the charging box and the earphone battery in real time, and read and display the ROM ID of the DS2488 on the earphone.
Figure 9. Test results with the charging case cover closed.
Figure 10. Test results with the charging case cover open.
in conclusion
Prototyping TWS headsets is often a huge challenge for engineers, balancing ease of use, low cost, portability, and stability. The DS2488 1-Wire bidirectional bridge paves the way for a low-power, high-stability, high-performance TWS headset solution in a smaller space and at a lower cost. Based on the DS2488, the MAXREFDES1302 includes hardware and firmware designs for power transfer and data communication through only two contact points, and is an easy-to-use TWS headset prototype.
The supply gap of semiconductor chips starts with automotive chips, and the shortage of automotive chips is no different from Achilles’ heel for car companies.
For example, we can see that due to the shortage of chips, many car companies on the market, including Audi, Volkswagen, and BMW, have been out of stock.
What’s even more funny is that some car owners broke the news that some Audi models currently only provide a remote control key and a mechanical key tooth, which makes people laugh and cry, and there is even a “miracle” that there is no car for sale in the whole store overseas.
BMW is even more simple and rude, directly reducing the wireless charging, WiFi hotspot, digital key and audio and other Electronic equipment of its many models, and customers do not want to pay for additional installations.
But recently, something even more worrying has happened, that is, two factories of the world’s largest automotive chip manufacturer have stopped production, that is, Infineon. Infineon will surpass NXP in 2020 and become the world’s first in the field of automotive chips.
Infineon’s factory in Texas was suspended due to the storm before, and has not fully resumed production. Recently, due to the epidemic, another factory (Malaysian factory) has been shut down.
This has led to a significant reduction in Infineon’s chip production, affecting its ability to deliver chips to core manufacturers. In Infineon’s words, the current chip inventory is at a historically low level.
And Infineon’s chip supply can’t keep up. Naturally, the first impact is on German cars, then American cars, and then Japanese cars, Korean cars, Chinese brand cars, etc. After all, Infineon’s market is too large and chips There are too many types, and most car manufacturers are inseparable from its chips.
However, relatively speaking, German and American cars have the greatest impact. It is said that in at least these two quarters, about 2.5 million cars will not be able to be produced. Some media even claim that these are German cars and American cars in the past 30 years. The worst supply shortage.
Next, we will see when Infineon’s chip production can resume. If the recovery is not timely, it may not be the worst in 30 years, but it may become 40 years, or 50 years, or even 100 years. What do you think?
On May 25, researchers at the Hong Kong University of Science and Technology have designed the world’s first 3D artificial eye, which may be clearer than the human eye.
If all goes well, it is expected to bring sight to millions of people within five years.
The artificial eye creates images through a variety of tiny sensors that mimic the light-detecting photoreceptors of the human eye, the report said. The sensor is packaged in aluminum and tungsten films, forming a hemisphere with a diameter of over 2 cm, mimicking the human retina.
What is the function of this 3D artificial eyeball?
According to Prof. Zhiyong Fan from the Hong Kong University of Science and Technology, the size of the bionic eye is comparable to that of the human eye, and the structure of the bionic eye is also highly similar to that of the human eye. When a single nanowire is electrically addressed, it has the potential to achieve high imaging resolution. The image is transformed by a multitude of tiny sensors housed in a hemispherical membrane made of aluminum and tungsten that mimics the human retina, which could theoretically exceed the high-resolution imaging of the human eye.
Not only that, this artificial retina is sensitive to all frequencies of light in the visible spectrum, and it responds in as little as 19.2 milliseconds after receiving a light stimulus, and then returns to an inactive state in 23.9 milliseconds, 40 times longer than the photoreceptor cells in the human retina. -150ms response and recovery time is much shorter.
Experts say the technology could be used in a wide range of applications. In addition to helping individuals improve their eyesight, other biomimetic light-sensitive devices can be made. Animal and clinical trials are currently being planned, and it is expected to be ready for use within five years.
It is reported that in addition to helping people improve vision, this bionic eye technology can also be used to make other biomimetic photosensitive devices, and animal and clinical trials are currently being planned.
Structure and function of the eyeball
The human eye is an approximate spherical body, with an anterior and posterior diameter of about 23-24 mm and a lateral diameter of about 20 mm, which usually becomes the eyeball. The eyeball is composed of two parts: the refractive system and the photoreceptor system.
The wall of the eyeball consists of three membranes with different textures:
(1) Cornea and sclera. The outermost layers of the eyeball wall are the cornea and sclera. The cornea is in front of the eyeball, accounting for about 1/6 of the entire eyeball wall area. It is a transparent film with a thickness of about 1mm and a refractive index of 1.336. The role of the cornea is to focus the light entering the eye, that is, to refract and concentrate the light entering the eye.
The sclera is the outermost white and tough membrane in the middle and rear, accounting for about 5/6 of the entire eyeball wall area, with a thickness of about 0.4-1.1mm, which is our “white of the eye”, and its role is to protect the eyeball.
(2) The iris and choroid. The iris, choroid, and ciliary body make up the middle layer of the eyeball wall. The iris is the annular membrane layer behind the cornea that divides the space between the cornea and the lens into two parts, the anterior chamber and the posterior chamber. The inner edge of the iris, called the pupil, acts like the aperture on a camera lens, automatically controlling the amount of incoming light.
The iris can contract and stretch, so that the pupil dilates when the light is weak and shrinks when the light is strong, and the diameter can vary from 2 to 8 mm.
The ciliary body is formed by the thickening of the choroid behind the junction of the sclera and the cornea. It contains smooth muscle that supports the position of the lens and regulates the convexity (curvature) of the lens. The choroid has the widest range, close to the inner surface of the sclera, about 0.4mm thick, and rich in melanocytes. It is like a camera camera obscura, which can absorb stray light in the eyeball and ensure that the light only enters the eye from the pupil to form a clear image.
(3) Retina. This is the innermost transparent film of the eyeball wall, attached to the inner surface of the choroid, with a thickness of about 0.1 to 0.5 mm. There are a large number of visual photoreceptor cells, cone cells and rod cells on the retina, which is the photoreceptor part of the eye, and its function is like the photoreceptor material in the camera.
In the central part of the back of the eyeball, there is a particularly dense area of cells on the retina, which is yellow in color, called the macula, about 2 to 3 mm in diameter, and has a small fossa in the center of the macula, called the fovea, where the most vision is sharp place.
The macula is about 4mm away from the nasal side. There is a disc-shaped optic nerve head. Because it has no photoreceptor cells, it has no photoreceptive ability, so it is called a blind spot. Light signals from external objects form an image on the retina, where the inner segment of the optic nerve transmits information to the brain.
Since the beginning of this year, the wafer foundry capacity has been extremely short. In response to the huge demand from customers, wafer foundries have successively expanded their investment and production capacity. The mature process capacity will be opened one after another in 2022, and will reach its peak in 2023. The tight production capacity is expected to be alleviated; however, as the production capacity is fully launched, the oversupply situation that the industry may face in the future is still a potential concern.
TSMC is committed to chasing Moore’s Law and promoting the development of advanced processes; however, under the severe shortage of global mature process capacity, in order to support customer demand, TSMC also rarely expands mature process capacity, and will expand the 28-nanometer process with a monthly production capacity of 20,000 pieces at the Nanjing plant .
TSMC believes that the shortage of capacity in the semiconductor industry will continue into next year, and the mature process is more likely to be lacking until 2022, while the new capacity of TSMC’s mature process will be opened in 2023.
UMC will work with a number of customers to expand the production capacity in the Nanke Fab12AP6 factory. It will adopt a 28-nanometer process and have a monthly production capacity of 27,500 pieces. Customers will pay a deposit in advance at the negotiated price. The expanded production capacity is expected to be put into production in the second quarter of 2023.
UMC believes that considering factors such as delivery time and geopolitics, and the market will not see significant capacity increase in the past 1-2 years, it is expected that the tight production capacity of mature processes will not ease before 2023.
The world’s advanced has also purchased AUO’s L3B plant and factory facilities in Zhuke, which can accommodate an 8-inch monthly production capacity of 40,000 pieces. Mass production is not expected until the end of 2022, or 2023.
NSMC will build a new 12-inch wafer factory in Tongluo, Miaoli, with a total production capacity of 100,000 wafers per month. It will be put into production in phases from 2023, and the annual output value at full load will exceed NT$60 billion.
SMIC has joined hands with the Shenzhen Municipal Government to expand the production capacity of the 28nm (and above) process, with a monthly production capacity of 40,000 12-inch wafers. It is expected to start production in 2022.
In addition to wafer foundries, many semiconductor factories have also successively expanded their investment. For example, Intel will return to the foundry industry, and the memory giant Hynix will expand its investment in the foundry business.
With the rapid rise of new demand for chips for automobiles, 5G, and AIoT, it has caused structural changes to the global industry, resulting in an explosion in the demand for mature process chips, and the shortage of supply will become more serious in the future.
Although major manufacturers are scrambling to expand production, in addition to hoping to relieve the pressure of shortages in the market, they are bound to be full of confidence in the prospects of the semiconductor industry. However, the economic reversal that may be faced in the future is still a potential risk.
On the other hand, as countries have poured huge sums of money to subsidize the autonomous production of local semiconductor chips, some experts have warned that with too much chip production capacity, when the turbulent wave of demand recedes, the industry may lead to a situation of oversupply in the future.
To ease the global chip supply crisis, the European Union may ease restrictions on state aid to the semiconductor industry.
The European Commission is likely to agree on a new state aid policy at a meeting on Wednesday that would allow the government to subsidize Europe’s cutting-edge chip factories, according to a draft policy document.
Specifically, given the importance and difficulty of securing chip supply, the European Commission may “approve public support to fill possible funding gaps in the semiconductor ecosystem, especially in building Europe’s own pioneering facilities.”
European Commission President Ursula von der Leyen also visited the Dutch ASML company on Monday, and stated that Europe needs to improve chip design and research technology, increase production capacity, and establish closer cooperation with the industry. “Increased chip production in Europe is good for Europe, which means less reliance on East Asia,” she said.
It is worth mentioning that the EU will also announce the so-called “European Chip Act” in the first half of next year, which aims to promote its semiconductor production strategy, one of which is to occupy 20% of the global market share by 2030.
voices against government intervention
However, even if the global chip supply situation is severe, the EU’s policy has not been recognized by all member states.
The calls for more state intervention are mainly from the larger member states, led by France and Germany. Those wanting to stick to the EU’s free market have opposed this, arguing that it would give an unfair advantage to those with stronger economies, and warned that loosening state aid rules could weaken Europe’s position at the WTO. position in.
For example, last week, six EU member states, including the Netherlands, Denmark, and Ireland, sent a letter to the European Commission opposing the use of government funds for large-scale production or commercial activities, saying that key industries were “excessively and untargeted in strategic investment.” funding” will lead to a subsidy race and unfair competition within the EU.
Margrethe Vestager, head of the European Union’s antitrust bureau, is also one of the supporters against government intervention. In his view, European chip “self-sufficiency is an illusion.”
In addition, France is vigorously calling for the development of Europe’s own cutting-edge chip industry, but there are also many opponents who insist that funds need to be invested in less cutting-edge chips, because the shortage of these chips is the main factor affecting European manufacturing.
C114 (Reuters) – The Trump administration has notified Huawei suppliers, including chipmaker Intel, that it will withdraw their permission to sell to the Chinese company, people familiar with the matter said. of certain licenses and intends to reject dozens of other applications to supply Huawei.
Reuters said it could be the last action against Huawei under U.S. Republican President Donald Trump and the latest in a long-running effort to undercut the world’s largest telecommunications equipment maker.
The notices came amid a flurry of U.S. actions against China in the final days before the Trump administration’s curtain call. Democrat Joe Biden will be sworn in as U.S. president on Wednesday.
An Intel spokesman had no immediate comment, and a Commerce Department spokesman did not immediately respond to a request for comment.
In an email documenting the actions seen by Reuters, the U.S. semiconductor Industry Association said on Friday that the Commerce Department “intentionally rejects numerous license applications for exports to Huawei and withdraws at least one previous license application. License issued”. A person familiar with the matter, who asked not to be named, said more than one license was to be revoked. Eight licenses from the four companies will be revoked, one of the sources said.
Japanese flash memory chip maker Kioxia has had at least one license revoked, two sources said. The company’s predecessor was Toshiba Storage Corporation.
The American Semiconductor Industry Association said in an email that the actions involved a “wide range” of products in the semiconductor industry and asked if the companies involved had been notified.
The email noted that companies had been waiting “months” for a licensing decision, and that with less than a week left in the Trump administration, how to deal with a “denial” was a challenge.
Companies that receive notices of “Intent to Deny” will have 20 days to respond, while the Commerce Department has 45 days to notify those companies of any changes in the decision, or the decision will become final. The companies will then have 45 days to appeal.
In May 2019, the U.S. placed Huawei on the U.S. Department of Commerce’s “entity list,” restricting suppliers from selling U.S. products and technology to the company.
But while the U.S. has tightened restrictions on Huawei, some sales to Huawei have been allowed and others have been denied, including a requirement that sales of semiconductors manufactured abroad using U.S. technology must be approved by the U.S. government. license.
The latest U.S. move comes after about 150 licenses to sell goods and technology worth $120 billion had not been approved, a person familiar with the matter said, as U.S. agencies failed to agree on whether the licenses should be authorized. Permissions have been on hold.
Another $280 billion in licenses to sell goods and technology to Huawei have yet to be processed, but now face a higher chance of being rejected, the sources said.
A rule last August said 5G-capable products could be rejected, but sales of less advanced technology would be decided on a case-by-case basis.
The U.S. government’s latest decision follows six meetings between senior officials from the Commerce, State, Defense and Energy departments, starting on January 4, the sources said. Officials developed detailed guidance on which technologies would support 5G and then adopted the standard, the person said.
In doing so, officials rejected the vast majority of about 150 contested applications and revoked eight licenses, the sources said.
The U.S. action came under pressure from Corey Stewart, Trump’s recently appointed Commerce Department official. Corey Stewart spent two months at the U.S. Commerce Department before the end of the Trump administration, where he wants to pursue a tough China policy.
“There are many ways to sense motor position today, and optical encoders are favored by motor control system designers because of their height and standardized “ABI” outputs that are easily controlled by a microcontroller.
“
There are many ways to sense motor position today, and optical encoders are favored by motor control system designers because of their height and standardized “ABI” outputs that are easily controlled by a microcontroller.
But contactless magnetic position sensors are now the better choice for a number of reasons. Due to their smaller size and resistance to contaminants such as dust, grease, and moisture, magnetic position sensors can be used in applications with higher size and/or reliability requirements.
In the past, there has been a trend against magnetic position sensors: new brushless DC (BLDC) motors have overall high efficiency goals to reduce power consumption. At the same time, designers were tasked with adding torque to the new motor, enabling low-speed operation of the motor to support direct-drive systems. Eventually, the transmission will no longer be a necessity, which greatly reduces bill of materials.
To maximize torque and efficiency, a brushless DC motor must have a very high degree of engine rotation angle data at high speeds—which is difficult to obtain with traditional magnetic sensors. Now, a new generation of products has achieved a major breakthrough in sensor design, they can almost completely measure the rotation angle at high speed.
How to implement angle measurement
A brushless DC motor consists of a permanent magnet motor (rotor) and three or more equidistant fixed coils (stator). A magnetic field of any direction and size can be formed by controlling the current in the fixed coil. Moment is the attraction and repulsion between the rotor running on the rotating shaft and the stationary coil.
The torque reaches a value when the stationary coil magnetic field and the rotor magnetic field are perpendicular to each other. The measured rotor angle is fed back to a system that controls current through fixed coils (see Figure 1), creating a vertical magnetic field.
Figure 1: A brushless DC motor control system requires either a magnetic position sensor (usually used in the automotive field) or an optical position
In most high-end applications, brushless DC motors are being replaced by permanent magnet synchronous motors (PMSM). Permanent magnet synchronous motors replace the modular commutation scheme affected by torque ripple in brushless DC motors, and can switch freely between coils, reducing vibration and achieving higher efficiency.
Of course, while industrial and automotive electric motor designs must often be optimized for efficiency and reliability, many other electric motors, especially those in consumer products, are cost-conscious. For a simple motor, the Hall switch array provides the proper position measurement and also generates the proper torque for smooth operation.
However, the degree and accuracy of the Hall switch array often cannot meet the torque and utilization requirements of high-performance engines. Conversely, a magnetic encoder (a semiconductor that integrates a Hall sensor into a silicon chip) produces high-resolution, high-resolution position data. It enables measurements on shafts at rest or at low rotational speeds. Unlike optical encoders commonly used in industrial applications, magnetic position sensors are immune to contamination and have a small footprint.
On the other hand, most Hall sensor chips suffer from two major drawbacks: dynamic angle errors at high rotational speeds caused by propagation delays; and shielding measures required in stray magnetic field environments.
These defects increase system cost and impair system performance. Dynamic angle error compensation requires a lot of processing power, and additional protection of the IC from stray magnetic fields also increases the hardware bill of materials.
Causes of Dynamic Angle Errors
The Hall sensor chip continuously samples the magnetic field strength of the magnet on the rotating shaft. The chip is mounted in a fixed position with its surface parallel to the surface of the rotating magnet, typically with a 1 to 2 mm gap between the chip and the magnet.
The chip contains a signal conditioning and processing circuit that converts the measured magnetic field strength to the angular position of the rotor (in degrees). The time required for this transition is the chip’s fixed propagation delay (see Figure 2). The duration of the delay varies from chip to chip, but chip propagation delays on the market today are typically between 10µs and 400µs.
Figure 2: Signal processing in magnetic position sensors causes propagation delays
The problem of propagation delay causes dynamic angle errors as the rotor turns. The dynamic angle error increases linearly with speed; the higher the propagation delay and speed, the larger the dynamic angle error. (See Figure 3).
Figure 3 shows the increase in dynamic angle error. Suppose the chip reads the magnetic field strength when the rotor is at the red line position, and the propagation delay of the chip is 100 μs while the rotor is rotating. When the chip converts the magnetic field strength to an angle, the rotor uses 100? s time goes to the blue line – but the chip shows the ECU or MCU that the rotor is still on the red line.
Figure 3: Linear relationship between dynamic angle error and rotational speed
In the absence of error compensation, the current in the adjustment scheme will go to the start coil at the red line position instead of the blue position. As a result, the system cannot be torqued, thus wasting energy and reducing system efficiency.
If the propagation delay of the chip is 100 μs and the speed of the engine is 1000 rpm, then the dynamic angle error is 1.2 degrees. If the rotational speed of the rotor is increased to 10,000 revolutions per second, the dynamic angle error increases to 12 degrees.
Figure 4: How Propagation Delay Increases Dynamic Angle Error
Propagation delay is characteristic of all magnetic position sensors, so system design engineers try to apply compensation algorithms to reduce dynamic angle errors. Unfortunately, the compensation of several thousand data samples per second places a severe burden on the host ECU, even requiring an additional custom MCU dedicated to error compensation.
The design team didn’t want to inherently increase the bill of materials, nor did they want to spend too much time developing, testing, and revising their compensation algorithms.
New sensor reduces dynamic angle error
As just mentioned, the propagation delay of the magnetic position sensor is fixed, and the value of the dynamic angle error depends on the time and rotation speed of the propagation delay.
Now, Austria Microelectronics has developed a new compensation scheme applied to the magnetic sensor, the scheme is pending. This new internal compensation technique, called DAEC (Dynamic Angle Error Compensation), was first tested on the 47 series of magnetic sensors. DAEC can effectively reduce the propagation delay error of the automotive position sensor AS5147 to only 1.9μs. This means that the dynamic angle error of the AS5147 at 14,500 rpm is only 0.17 degrees, which is almost negligible.
Figure 5: Sensor output with integrated compensation scheme (left) and without integrated compensation scheme (right)
Figure 6 shows the difference between the measurement output of the AS5147 (left) and a conventional magnetic position sensor (right), with some optical encoder output as a reference. The graph to the right shows that the sensor output is affected by a 200µs propagation delay, resulting in a dynamic angular error of 18 degrees at 14,500 rpm.
In contrast, the error of the AS5147 is almost negligible, which means that its signal can be used directly to tune the controller without external compensation. In fact, the dynamic angle error produced by internal compensation with DAEC technology may be smaller than external compensation because sampling errors are often present in ECUs and MCUs.
Of course, the internal compensation of the sensor can also reduce the cost of the system, because there is no additional MCU, or because a lower power ECU can be used.
Protection against stray magnetic fields
Another disadvantage of many magnetic sensors is their susceptibility to stray magnetic fields. Magnetic field disturbances other than the rotor magnets can corrupt the chip’s angle measurement at any time, and such random errors cannot be remedied by the host ECU or MCU. Therefore, the user has to take shielding measures for the chip, which increases the cost of materials and assembly; and may also violate the structural design of space-demanding applications.
According to the ISO 26262 automotive functional safety standard, immunity from stray magnetic fields has become a mandatory requirement for engine systems.
“Differential Sensing” technology is used in all Austrian Microelectronics magnetic position sensors, including the 47 series, making the sensor immune to stray magnetic fields up to a value of 25,000A/m. Below this threshold, no shielding is required.
in conclusion
The introduction of DAEC technology from Austria Microelectronics means that manufacturers of brushless DC motors and permanent magnet synchronous motors can use extreme position data to maximize torque in high-speed applications, while reducing motor size and improving reliability through magnetic position sensors .
DAEC technology is now available in AS5147* single-layer wafers) and AS5247 (double-layer redundant wafers) automotive magnetic position sensors (AEC-Q100 Phase 0 automotive applications), supporting brushless DC motors in automotive applications such as electronics Power steering (EPS), transmission (gearbox, actuators), pumps and brakes.
In industrial applications, the AS5047D with DAEC technology is also in use, providing a decimal ABI output, ideal for replacing optical encoders.
On July 9, 2020, at the opening ceremony of the 2020 World Artificial Intelligence Conference (AIC” target=”_blank”>WAIC) Cloud Summit, Qualcomm President Cristiano Amon delivered a speech at the Industrial Development Plenary Session with the theme of ” 5G+AI opens the future of intelligent interconnection” speech. The following will share the content of Anmon’s speech as follows.
These are exciting times for the industry as we enter a new era of intelligent cloud connectivity powered by AI and 5G. Everything will be connected to the cloud, interconnected in a reliable way. Now, terminals and experiences are inseparable from content, data, computing power and storage in the cloud. 5G’s fiber-like connection speeds and low latency, coupled with advanced processing capabilities, provide the foundation for all of this to make our edge devices smarter. The combination of 5G and AI that will impact every aspect of our lives and numerous industries is the next big technological change. To achieve AI at scale, we must implement distributed intelligence across the network. Today, intelligence is widely distributed in the cloud and gradually migrated to the terminal side. In the future, massive data from billions of edge terminals will require the creation of a new edge cloud. In the new edge cloud, the processing of content and control will move to where the traffic is generated. Now that edge cloud and 5G can reduce latency and enable new applications and services, we will see improved performance, security and privacy.
Edge cloud combined with terminal-side AI will also facilitate a wider range of new use cases. For us, life will be richer because of the realization of more intuitive, immersive and contextual experiences – from interactive content to unbounded virtual reality to real-time voice translation. In more industries, 5G and AI combined with edge cloud will also promote enterprise innovation in a positive and disruptive way. You will see how people can take advantage of the vast amounts of data generated by intelligent machines and objects. People can conduct on-premises, real-time network analysis, driving business insights, taking productivity to a whole new level, and bringing better levels of automation and control. This trend will also change the way we look at endpoints, as the cloud will play a bigger role relative to operating systems and app stores. Imagine (the terminal) being always online and getting almost unlimited storage and processing power from the cloud. The realization of this scenario will promote the emergence of more powerful and potential applications. When we see applications such as WeChat and Tencent QQ today, we realize that the evolution of applications is so fast, and the functions of future applications will be more powerful and independent of the operating system. When people can connect devices and the cloud in a reliable way, there is no need to develop separate applications for multiple operating systems. Regardless of the ecosystem, applications and experiences will be the main focus, which will accelerate innovation and the development of new applications, further expanding the usability of applications.
5G and AI will transform many segments and industries.
First, let’s talk about education. Using technologies such as 5G and AI, intelligent online education can become richer and more inclusive. It will be able to support differentiated and personalized learning. Technologies such as real-time translation will allow students to access content from global education databases and learn in their own language. Furthermore, it is clear that network connectivity will make the education system more flexible. As of April 2020, just over two months ago, a total of 1.7 billion students worldwide had adopted distance learning, an experience that accentuated the digital divide caused by the broadband connectivity gap, but 5G could connect them. Once the capabilities of 5G and edge cloud are available, a new learning experience, high-quality interactive content, and a more productive and intelligent society can be realized anytime, anywhere. Another important area is healthcare, especially in the current environment. We can see how 5G and AI can transform healthcare, enabling more accurate early diagnosis, health monitoring, and drug and vaccine development. The key to making it all work is connectivity. In the US alone, it is expected that by 2030, 50% of medical care will be performed online. With technology, you can enjoy high-quality video, massive data-backed images and research, and real-time AI interactions. In just one small step of technological innovation in this area, we are already starting to see first responders now in South Korea able to use 5G-connected 360-degree cameras to transmit data in real-time from the emergency room to a team of doctors. For the healthcare industry, which affects nearly 10% of global GDP, technology and AI will help expand coverage, improve experience, and reduce costs. Now, let’s talk about the future of productivity and productivity. This is another important transformational area that is undergoing change, with the migration of enterprise workflows to the cloud to enable smarter and more optimized data management. Now these systems must support anytime, anywhere work, and network connectivity helps improve today’s productivity. Think of all the benefits that 5G can bring, now that data is tightly linked to the cloud and AI. For example, in recent months, 84% of organizations indicated that partial work from home or telecommuting will likely continue. Now, many companies are even starting to rethink how to merge the digital space and the physical world, with technology and AI complemented by extended reality (XR), augmented reality and holographic remoteness to support new ways of working. The transformation of the business has only just begun. Retail is another area that will be significantly impacted. 5G and AI can transform the retail industry from end-to-end, covering everything from intelligent supply chain, logistics management, and new in-store experiences. In the 5G environment, Boundless XR (Boundless XR), separate rendering and AI processing technologies can bring a personalized shopping experience. You can see AI-tailored product recommendations in the window, and even see a photo-realistic image of yourself in the Display. High-throughput, low-latency 5G XR glasses will also use cloud and terminal-side intelligence to map virtual objects to the real world. There are great opportunities in this. When we think about industry, we have to talk about manufacturing. The manufacturing industry is currently undergoing major changes, and we can see the changes brought about by intelligent interconnection. 5G and AI will enable a new generation of reconfigurable, flexible manufacturing. The enterprise private network will provide secure industrial data management through highly reliable 5G, and data processing will be carried out separately between the terminal and edge cloud or local server as required. The ultra-reliable and low-latency features of 5G will enable industrial-grade remote control of machines; intelligent operations with end-to-end tracking and control functions will raise productivity to a higher level. We will even see innovations never imagined. Finally, 5G and AI will open up a bright future for the transportation industry. Wireless technology is already transforming the transportation industry as we connect our cars to the cloud. We expect 5G and AI to make cars smarter: cars will have smart cockpits, autonomous driving and data analysis capabilities and new services. From a safety perspective, 5G C-V2X promises to make travel safer, more efficient and more enjoyable as we connect cars to cars, cars to infrastructure, and cars to pedestrians. Roadside infrastructure will utilize AI-assisted cameras to sense and implement traffic flow control. Combining sensors and AI in computing platforms will make intersections safer. 5G commercial use can also make full use of the combination of roadside units and public network small base stations to create efficient collaboration between smart cities and operators. This is a potentially dramatic change in the transportation industry. All of the above mentioned must be achieved through cooperation. A smart, connected future requires multiple companies to work together. Qualcomm has always been committed to cooperation within a global ecosystem, innovation and collaboration to create new technologies. We are proud to partner with many leading companies around the world and in China to leverage 5G and AI to drive innovation and growth across multiple industries. Unleashing the full potential of AI in the cloud, at the edge, and at the endpoint is incredibly exciting. 5G will be a platform for connecting all things with intelligence, which means huge opportunities for all players in the mobile industry.
Create continued growth with a portfolio that generates $8.2 billion in revenue, expanding the scale and diversity of the business.
Increase domain expertise and expand engineering capabilities to develop more complete solutions to our clients’ complex problems and challenges.
It is expected to be accretive to free cash flow at closing, boost adjusted earnings per share (EPS) within 18 months of closing, and realize cost synergies of $275 million by the end of the second year.
Analog Devices, Inc. and Maxim Integrated Products, Inc. announced on July 13 that they have reached a definitive agreement for Analog Devices to acquire Maxim in an all-stock transaction with a combined company market value of more than $68 billion. The boards of directors of both companies have unanimously approved the transaction. The transaction will strengthen ADI’s leadership in analog semiconductors by expanding the breadth and scale of its business in multiple attractive end markets.
Under the terms of the agreement, upon closing of the transaction, shareholders of Maxim common stock will receive 0.630 shares of ADI common stock for each share. After the transaction closes, ADI’s current shareholders will own approximately 69% of the combined company, while Maxim shareholders will own approximately 31%. The transaction is intended to qualify for a tax-exempt restructuring under U.S. federal income tax law.
ADI President and CEO Vincent Roche“Today, we are making an exciting announcement with Maxim, explaining the next step in ADI’s vision to bridge the real and digital worlds. Both ADI and Maxim are committed to solving the complex problems of our customers, and together we will further Expand the breadth and depth of technology and talent to enable the development of more complete leading-edge solutions. Maxim is a prestigious signal processing and power management company with a proven technology portfolio and an impressive history of design innovation. Our strong Join forces to work together to deliver the next wave of growth in the semiconductor industry while creating a healthier, safer and more sustainable future for all.”
Tunç Doluca, President and CEO of Maxim Integrated“For more than three decades, we’ve held on to our belief that we continue to innovate and develop high-performance semiconductor products to help our customers invent. In the future, I’m very excited to work with Analog Devices to continue to push the boundaries of technology and go beyond what’s possible.” Both of our companies have deep engineering expertise and a strong culture of innovation. Together, we will build a stronger industry leader, delivering exceptional value for our customers, employees and shareholders.”
Following the transaction, two Maxim directors will join ADI’s board of directors, including Maxim President and CEO Tunç Doluca.
Compelling strategic and financial case
An industry leader in expanding global business: The merger strengthens ADI’s leadership in the analog semiconductor market and, on the pro forma, is expected to reach $8.2 billion in revenue and $2.7 billion in free cash flow. Maxim’s strength in the automotive and data center markets complements ADI’s strength across a broad range of industrial, communications and digital healthcare markets and is set to drive key long-term growth trends. In power management, Maxim’s application-focused product portfolio complements ADI’s broad-market product portfolio.
Extensive domain knowledge and technical capabilities: The integration of best-in-class technologies will further strengthen ADI’s domain expertise and engineering capabilities, spanning from DC to 100 Ghz, from nanowatts to kilowatts, and from sensors to the cloud, covering more than 50,000 products. The combined company can provide more complete solutions, serve more than 125,000 customers, and capture opportunities in a target market totaling $60 billion3.
A common philosophy of innovation-led growth: Both parties have similar corporate cultures that value talent, innovation and engineering excellence, with a total of more than 10,000 engineers and an annual R&D investment of nearly $1.5 billion. The combined company will continue to attract top engineering talent in various fields.
Revenue growth and cost savings: Adjusted EPS is expected to increase gradually over the 18 months following closing due to lower operating expenses and COGS, with cost synergies reaching $275 million by the end of the second year. By the end of the third year after the deal closes, manufacturing process optimization is expected to bring additional cost synergies.
Strong financial strength and cash flow capability: ADI expects the combined company to have a stronger balance sheet, with an estimated net leverage close to 1.2×4. The transaction is also expected to bolster free cash flow at closing, generating more returns for shareholders.
timing and approval
The transaction is expected to close in summer 2021, subject to the satisfaction of closing conditions, including regulatory approvals in the U.S. and outside the U.S. and approval by shareholders of both companies.
About Analog Devices
Analog Devices, Inc. is the world’s leading high-performance analog technology company dedicated to solving the toughest engineering design challenges. With outstanding detection, measurement, power, connection and interpretation technology, build intelligent bridges between the real and digital worlds, thereby helping customers to re-understand the world around them.
About Maxim Integrated
Maxim Integrated develops innovative analog and mixed-signal products and technologies that make systems smaller and smarter while enhancing their safety and energy efficiency. We help our customers with innovative designs in automotive, industrial, health, mobile consumer and cloud data centers, providing industry-leading solutions that make the world a better place.
