Signal monitoring board for single event upset effect test
By designing a signal monitoring board, the microprocessor is monitored online using an ARM processor and scan chain test code point files. This solves the problems of detection delay and incompleteness in existing technologies, and enables timely detection and accurate analysis of single-particle flips.
Patent Information
- Application Number
- CN202511833124.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-08
- Publication Date
- 2026-03-06
AI Technical Summary
Existing technologies suffer from communication delays and incomplete detection when detecting single-event upsets, failing to capture them in a timely manner and accurately reflect the microprocessor's single-event immunity.
A signal monitoring board is used to read the scan chain test code point file through the ARM processor, configure the GPIO output level, read the GPIO input level and compare it with the code point file, so as to realize online monitoring of the microprocessor input and output registers, and timely detect and locate the register position.
It enables timely detection and accurate monitoring of single-event upset effects, improves detection efficiency and data reliability, reduces costs, and is suitable for universal detection using different microprocessors.
Smart Images
Figure CN121613792A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of electronic technology, specifically a signal monitoring board for single-event upset effect experiments. Background Technology
[0002] With the development of electronic technology, control systems are widely used in the field of irradiation testing. The single-event upset (SET) effect test signal monitoring board of this invention is applied to microprocessor-based SET tests and is currently in normal use. The signal monitoring board has advantages such as small size, high component density, flexible control methods, and reliable performance.
[0003] The commonly used method for detecting single-event upsets is to use a host computer + slave computer. The host computer sends a detection command, which is transmitted to the slave computer through a communication port and transmission line or wireless transmission. The slave computer detects different functional units of the microprocessor according to different commands. When a single-event upset is detected, the upset status is reported to the host computer through the communication port. The host computer parses the data packets transmitted from the slave computer and displays the upset status on the UI interface.
[0004] The currently used processing methods have two major drawbacks:
[0005] First, there is a communication time between the host computer and the slave computer. The host computer decoding data packets, the slave computer encoding data packets, the communication transmission time, and waiting for the other party's response all consume a significant amount of CPU cycle time. If a single-event upset occurs during this time, the host-slave detection method cannot detect it in time. This invention completely eliminates the host-slave detection method. By not using this method, there is no communication time between the host and slave computers, maximizing the time available for detecting the single-event upset effect and displaying the upset status promptly.
[0006] Second, the current method of detecting functional units has limitations. It can only detect whether the output of a functional unit has flipped, but not whether the input has flipped. Therefore, it cannot fully reflect the microprocessor's single-event immunity. This invention, by detecting scan chain code points, can analyze the single-event immunity of all registers in a microprocessor, including input and output registers. Once a flip occurs, it can be located within the scan chain code points to pinpoint the exact location of the flipped register. Based on this, design and manufacturing defects in the microprocessor can be identified and improved—something the functional unit detection method cannot do. Summary of the Invention
[0007] The purpose of this invention is to provide a control board with stable performance that can monitor the signals of single-event upset experiments online, ensuring the accuracy of the acquired experimental data.
[0008] The technical solution adopted by the present invention to achieve the above objectives is as follows:
[0009] A signal monitoring board for single-event upset (SET) testing includes a PCB printed circuit board and an ARM processor, a USB-type-A dual-layer female connector interface, an RJ45 Ethernet connector interface, a DB9 interface, and a microprocessor for performing SET testing, all mounted on the circuit board.
[0010] A method for implementing a signal monitoring board for single-event upset experiments includes the following steps:
[0011] 1) Initialize the thread class object of the open-source Linux operating system;
[0012] 2) Select and read the scan chain test code point file stored internally by the ARM processor;
[0013] 3) Configure the output level of the specific function pin GPIO sent to the microprocessor based on the code points in the scan chain test code point file;
[0014] 4) Read the input level of a specific function pin (GPIO) of another part of the microprocessor;
[0015] 5) Compare the read input level with the corresponding part of the scan chain test code point file, and determine whether a flip has occurred. If a flip has occurred, display the single particle flip status through the UI interface and record it in the test file in text form. Otherwise, return to step 3) to continue reading the next line of scan chain test code points.
[0016] 6) Repeat steps 3) to 5) until all the code point files are complete.
[0017] Step 1) specifically refers to:
[0018] The system globally maps the GPIO address, modifies the status word of the GPIO control register, specifies the GPIO transmission direction, initializes the thread class object, and starts the object into the running state.
[0019] Step 2) specifically refers to:
[0020] The microprocessor DUT for single-event upset (SWE) testing is selected and configured using the scan chain test code point file. The length of each line in the scan chain test code point file is a fixed value, representing the number of specific pins that need to be connected to the DUT. Each column represents whether a specific input pin of the DUT should be configured to a high or low level, and whether the output pin should be read as a high or low level in the corresponding register if no SWE occurs.
[0021] Step 5) specifically involves:
[0022] The input level read is compared with the corresponding part of the scan chain test code point file. If they do not match, a single-event flip is determined to have occurred.
[0023] The present invention has the following beneficial effects and advantages:
[0024] 1. The innovation of this invention lies in realizing a control board with stable performance that can monitor the experimental data of single-event upset effect online, ensuring the accuracy of the acquired experimental data.
[0025] 2. This invention features stable performance, high acquisition accuracy, low cost, and high reliability.
[0026] 3. For different types of microprocessors, only the scan chain test code point file needs to be changed to enable observation and comparative analysis of single-event upset experiments of microprocessors. Attached Figure Description
[0027] Figure 1 This is a schematic diagram of the hardware principle of the present invention;
[0028] Figure 2 This is a flowchart of the software structure of the present invention. Detailed Implementation
[0029] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.
[0030] A signal monitoring board for single-event upset effect testing includes a PCB printed circuit board, an ARM processor, a USB-type-A dual-layer female connector interface, an RJ45 Ethernet connector interface, a DB9 interface, and a microprocessor for performing single-event upset testing.
[0031] The ARM processor reads the scan chain test code point file stored in the ARM's internal memory, line by line according to the file's format. For each line of code point text (0 or 1), the corresponding GPIO output pin is converted to a low or high level. These GPIO pins are connected to specific input pins of the microprocessor. After several machine cycles, the low or high levels of another set of GPIO input pins (connected to specific output pins of the microprocessor) are read and converted to text 0 or 1. This data is then compared item by item with the contents of the scan chain test code point file to determine the single-event flip.
[0032] The USB-type-A dual-layer female connector interface allows you to insert a USB flash drive to store the scan chain test code point file into the ARM's internal memory. Alternatively, you can use this interface to read out the newly generated code point comparison file for each test and store it on the USB flash drive inserted into the interface.
[0033] The RJ45 Ethernet connector allows you to plug a network cable into the interface and use mainstream remote observation software tools such as Windows Remote Desktop and VNCviewer to observe the single-event flip of the microprocessor in real time.
[0034] The DB9 interface allows connection to a programmable power supply, enabling power to be supplied to ARM processors and microprocessors.
[0035] like Figure 1 As shown, this invention provides a stable signal monitoring board capable of online processing of single-event upset signals. It consists of a PCB printed circuit board and, on the board, an ARM processor, a USB-type-A dual-layer female connector interface, an RJ45 Ethernet connector interface, a DB9 interface, and a microprocessor for conducting single-event upset tests.
[0036] The ARM processor of the signal monitoring board of this invention runs the open-source Linux operating system;
[0037] The Ethernet driver, USB driver, and other hardware drivers of the signal monitoring board of this invention are all based on the open-source Linux operating system.
[0038] like Figure 2 As shown, a method for implementing a signal monitoring board for single-event upset experiments includes the following steps:
[0039] 1. Before the irradiation source is turned on, the software program starts running from the "Start" node until it stabilizes and enters a loop state. When the irradiation source begins to output single particles, the program is already in a loop state;
[0040] 2. Initialize the thread class object, globally map the GPIO addresses, and modify the status word of the GPIO control register. Specify that the transmission direction of some GPIOs is input and that of others is output. Initialize the thread class object and start the object into the running state;
[0041] 3. Select and read the scan chain test code point file. Each line in the file has a fixed length, representing the specific number of pins of the microprocessor (DUT) that need to be connected. Each column represents whether a specific input pin of the microprocessor (DUT) should be configured to a high or low level, and whether the output pin should be read as high or low level without a single-event flip in the corresponding register.
[0042] 4. Based on the code points, configure the output levels of several GPIOs (sent to specific function pins of the microprocessor). Read the input levels of several GPIOs (reading specific function pins of another part of the microprocessor), with high level corresponding to 1 and low level corresponding to 0. Compare these levels with the corresponding points in the scan chain test code point file. If the levels read from the microprocessor (DUT) do not match the code point file, a single-event flip (SIF) is determined to have occurred. Immediately display the SIF status on the UI interface and record it in text form to the test file. Repeat the sending, reading, and comparison process until all contents of the code point file are completed.
Claims
1. A signal monitoring board for single event upset effect test, characterized in that, The PCB includes a printed circuit board and an ARM processor, a USB-type-A double female connector interface, an RJ45 Ethernet connector interface, a DB9 interface and a microprocessor for single event upset test carried on the circuit board, wherein the USB-type-A double female connector interface, the RJ45 Ethernet connector interface, the DB9 interface and the microprocessor for single event upset test are connected with the ARM processor.
2. An implementation method of a signal monitoring board for single event upset effect test, characterized in that, The method comprises the following steps: 1) initializing a thread class object of an open-source Linux operating system; 2) selecting and reading a scan chain test code point file stored in the ARM processor; 3) configuring the output level of a specific function pin GPIO sent to the microprocessor according to the code point in the scan chain test code point file; 4) reading the input level of another specific function pin GPIO of the microprocessor; 5) comparing the read input level with the corresponding part of the scan chain test code point file, and judging whether a flip has occurred, if a flip has occurred, displaying the single event upset situation through a UI interface and recording it in a test file in text form, otherwise, returning to step 3) to continue reading the next line of scan chain test code point; 6) repeating steps 3) to 5) until all contents of the code point file are completed.
3. The implementation method of claim 1, wherein, The step 1) is specifically: mapping the address of the global GPIO, modifying the state word of the GPIO control register, specifying the transmission direction of the GPIO, initializing the thread class object, and starting the class object to enter the running state.
4. The implementation method of claim 1, wherein, The step 2) is specifically: selecting and configuring the microprocessor DUT for single event upset test through the scan chain test code point file, wherein the length of each line in the scan chain test code point file is a fixed value, representing the number of specific pins to be connected to the DUT, and each column represents whether the specific input pin of the DUT should be configured as high level or low level, and whether the output pin should be read as high level or low level in the case of no single event upset of the corresponding register.
5. The implementation method of a signal monitoring board for single event upset effect test according to claim 1, wherein, The step 5) is specifically: comparing the read input level with the corresponding part of the scan chain test code point file, if they are inconsistent, it is determined that a single event upset has occurred.