Talking about automotive chips: Serdes chips and testing
The new issue of "Core Technology"
Nice to see you all again!
In the previous issue, Technical Interpretation of the "Guidelines for the Construction of Automotive Chip Standard System" and Overview of Power Chip Measurement , we introduced the interpretation of key chips and test areas involved in the "Guidelines for the Construction of Automotive Chip Standard System" issued by the Ministry of Industry and Information Technology. This issue continues to extend the interpretation of MIPI A-PHY and automotive Serdes chip technology and testing.
There are more technical information downloads and lucky draws at the end of the article!
The importance of automotive Serdes chips
In the era of intelligent driving, especially with the popularization of assisted driving/autonomous driving, environmental perception has become an emerging key technology field in the automotive industry. The number and resolution of on-board cameras are increasing rapidly. A large amount of data requires high-speed, high-bandwidth transmission, making high-performance on-board Serdes chips one of the core components. The following figure is a block diagram of the autonomous driving architecture. Serdes/MIPI A-PHY is responsible for the high-speed data interconnection between the autonomous driving ADAS chip and the camera.
Figure: Schematic diagram of ADAS system and camera interconnection
At present, from on-board cameras to on-board displays, ADAS domain controllers, cockpit domain controllers, etc., on-board communication Serdes chips are used to realize video signal and real-time data transmission.
Overview of mainstream technical standards for vehicle-mounted Serdes
Currently, there are many standards for automotive SerDes, and private protocols include TI's FPD-Link standard, ADI's GMSL, Inova Semiconductors' APIX, and Rohm's Clockless Link.
There are currently three main public standards: MIPI A-PHY and ASA, as well as China's HSMT standard.
Participants in the MIPI A-PHY camp include BMW, Toyota, Bosch, Denso, ZF, Intel, Microsoft, Panasonic, Qualcomm, etc. In 2018, the MIPI AWG (Automotive Working Group) was established to pave the way for the introduction of a common communication framework called MASS (MIPI Automotive SERDES Standard). In 2021, the MIPI Alliance recently released MIPI A-PHY v1.0, the first automotive long-distance serializer/deserializer (Serdes) physical layer interface specification.
Currently, the latest standard version of MIPI A-PHY is MIPI A-PHY® v1.1.1 (May 2023).
Interpretation of MIPI A-PHY CTS V1.0 Standard
The primary goal of advanced driver assistance systems (ADAS) is to ensure the safety of occupants in the car and pedestrians around. The subsystem can only work properly when the receiver operates at a qualified bit error rate. It is worth noting that receiver testing has been included in the compliance test specification (CTS) of various related specifications involved in MIPI A-PHY. When testing Serdes physical layer (PHY) signals, it is crucial to understand how it sends messages and then how it receives messages at the other end. Based on the latest CTS v1.0, we can cover the entire physical layer test conformance test, starting with signal integrity testing, which aims to eliminate distortion, reflection, attenuation and/or coupling noise from other sources. Secondly, to verify the functionality of the device under test (DUT) and the signal transmission between PHY and ECU at a higher level, understand whether the signal transmission is prioritized correctly. Finally, you need to put the DUT into a series of stress environments and identify at which stage it fails, which is critical to the testing of the entire system.
Serdes/A-PHY Test Overview and Key Points
The testing of Serdes chips can generally be summarized into three categories: transmitter testing, receiver testing and link testing. Below we will give a detailed introduction to several types of tests and key test points.
Serdes/A-PHY transmitter test
The transmitter Tx test of Serdes/A-PHY is mainly completed using an oscilloscope with Keysight's proprietary software. The following figure is a typical differential SMA connected transmitter test block diagram.
Figure: Differential connection transmitter Tx test block diagram
The UXR series oscilloscope is used as the signal receiver to obtain excellent noise and jitter performance. With the Keysight AE2010T automotive Serdes transmitter test application software, the oscilloscope can be automatically configured during the test and detailed test results, including margin and statistical analysis results, are provided. The general test items are as follows:
Figure: MIPI A-PHY 1.0 CTS Version 0.9 Test Items
Figure: ASA 1.1 CTS version 0.2 test items
Figure: GMSL2 test items
The key test points of transmitter testing include:
Can perform verification tests based on Serdes protocols (including MIPI A-PHY and ASA), as well as the following test items
• Frequency
• distortion
• Jitter
• Pressure drop
• MDI return loss
For detailed information, please refer to the download content at the end of the article, transmitter test section.
Serdes/A-PHY Receiver Test
The receiver Rx test of Serdes/A-PHY is relatively complex and requires the use of a high-speed arbitrary waveform generator. There are also relatively more test items. The following is a typical Rx test block diagram. With the help of the M8195A arbitrary waveform generator, a stressed signal is generated, and noise is added through the AE2090B fixture. The control software AE2010R runs alone on the PC.
Figure: Receiver Rx test block diagram
Key test points include:
The receiver is responsible for decoding the data sent over the link and then passing it to the ECU or display device for further processing. Bit errors in the receiver can result in lost or corrupted data from safety-critical sensors such as cameras, radars, and lidars.
Traditional receiver capabilities are increasingly ill-suited for complex modulation schemes such as PAM-4, especially when the signal is transmitted over long channels and is interfered by multiple noise sources simultaneously. To characterize the performance of a receiver, engineers must measure its bit error level when interfered by multiple noise sources, including:
• Narrowband interference
• High current injection
• Online transients
• Alien Crosstalk
The measurement system can include noise sources, amplifiers, and coupling circuits to inject precise noise levels into the active SerDes link. The receiver must be able to correctly interpret the symbols in the presence of noise interference. The focus of receiver testing is to ensure that the receiver can maintain a low bit error rate (BER) when exposed to interference.
The following is a video of the receiver test.
For your reference and study.