Abstract : Based on the principle of aero-engine temperature sensors, a smart temperature sensor based on DSP and CAN is proposed. A power-on self-test circuit, a thermocouple signal processing circuit, a DSP and CAN bus interface circuit, and a power supply circuit are designed. This sensor system has high integration and fast measurement and processing speed. Experiments show that the temperature sensor has a measurement error of only 1℃, high measurement accuracy, and good real-time performance. It can be applied to the full authority digital electronic control (FADEC) system of aero-engines and has significant practical value. 0. Introduction Full authority digital electronic control (FADEC) of aero-engines is the development trend of integrated control of modern fighter jet flight/propulsion systems. It leverages the powerful and fast digital computing and logical judgment capabilities of computers to achieve a more advanced, complex, and reliable control method than mechanical hydraulic control systems, improving the control quality of the integrated flight/propulsion system. However, if analog signals from traditional sensors are used as outputs, the entire flight/propulsion integrated system has more than 30 sensor input signals. The central processing unit will consume 50% to 70% of its resources on analog signal data processing, redundancy management, and fault diagnosis, greatly weakening the advantages of the digital control system. The emergence of intelligent sensors has opened up broad prospects for solving this problem. Intelligent sensors used in aviation control systems, in addition to transmitting/receiving digital signals, also perform tasks such as signal acquisition and processing, fault self-diagnosis, fault isolation, and fault tolerance. They offload the heavy, low-level tasks of the FADEC system, freeing up significant CPU resources to implement complex and precise control algorithms and monitoring management, thereby improving the dynamic characteristics and overall performance of the flight/propulsion integrated system. Sensors, as crucial tools for acquiring information, are located at the forefront of information systems. Their characteristics and the reliability of their output information are critical to the overall system quality. Compared to traditional sensors, intelligent sensors replace the original voltage or current standard signals with digital signals, thus improving the reliability of signal transmission and anti-interference capabilities. Furthermore, intelligent sensors use a unified bus standard, making the system more open and versatile. Intelligent sensors represent the future direction of sensor development; these sensors have a standard digital bus interface and are capable of self-management. They transform and process the detected signals, then communicate and transmit them to the central processing unit in digital form via a fieldbus. This paper proposes an intelligent temperature sensor for aero-engines based on distributed control, constructed using AD595 and TMS320C2407A DSP. It mainly realizes cold junction temperature compensation, fault alarm, and nonlinear correction during thermocouple operation, possessing significant practical value. 1. Principle of the Temperature Sensor The turbine afterburner gas temperature T₄ is an important state parameter of the main fuel control system of a certain type of turbofan engine. When the engine control plan is in combat or training-combat state (including maximum and afterburner) and the engine inlet temperature is 288K... At 373K, the electronic controller adjusts the fuel supply mf to cause T4 to increase linearly by 15℃ as T1 increases. Therefore, accurate measurement and control of the turbine afterburner gas temperature T4 is crucial for improving the operational stability and reliability of the entire aero-engine distributed control system. The sensitive temperature measuring component for the turbine afterburner gas temperature T4 in aero-turbofan engines is a type K thermocouple. The principle of a thermocouple is based on the Seneca effect in physics, which is used to create a temperature-sensitive sensor. When two different conductors A and B form a closed circuit, if the junction temperatures at both ends are different (T0 and T, respectively), a current is generated in the circuit, and the corresponding potential is called thermoelectric potential. It consists of two parts: contact potential and temperature difference potential. Its magnitude is related to the temperature difference between the two ends and the material properties. The material is required to have stable thermal properties, low resistivity, high conductivity, strong thermoelectric effect, and good reproducibility. 2. Introduction to TMS320LF2407A: The TMS320LF2407A is a low-power DSP chip manufactured by TI, operating at 3.3V and using a 16-bit fixed-point architecture. It features on-chip Flash program memory; the "A" indicates the chip's encryption level. Its program memory (DARAM, SARAM, ROM, and Flash) and data memory (three DARAM blocks) have independent units and bus structures (Harvard architecture), allowing simultaneous access to program instructions and data, with data read and write operations possible within a single clock cycle. The CPU instruction cycle is 25ns, and it includes a 32-bit arithmetic logic unit, a 32-bit accumulator, a 16-bit × 16-bit multiplier, an 8 × 16-bit auxiliary register, and two status registers. The TMS320LF2407A's on-chip peripherals include many modules: a watchdog timer module (WDM), a digital input/output module (I/OM), two event management modules (EVM), a 10-bit analog-to-digital converter module (ADCM), a high-speed synchronous serial peripheral module (SPIM), a programmable serial communication module (SCIM), and a local area network control module (CANM), facilitating the development of different types of control and communication methods. Among these, the general-purpose timer (GPT) in the event manager (EV) module is used for pulse counting, and its core is a 16-bit readable/writable timer/counter TxCNT, which is the frequency counting unit for the intelligent speed sensor. [b]3. Intelligent Speed ​​Sensor Design[/b] 3.1 Hardware Circuit Design The circuit design principle of the intelligent speed sensor is shown in Figure 3. The intelligent speed sensor mainly consists of a power-on self-test circuit, a thermocouple signal processing circuit, a DSP and CAN bus interface circuit, and a power supply circuit. The intelligent temperature sensor features a power-on self-test function. The electronic analog switch uses the MAX319 chip. When the DSP's general-purpose I/O pin IOPA5 outputs a high level, the power-on self-test circuit is activated; when the DSP's general-purpose I/O pin IOPA5 outputs a low level, the thermocouple temperature measurement signal is introduced into the signal conditioning circuit. The thermocouple signal conditioning circuit mainly consists of an AD595 chip and an operational amplifier. The AD595 is a signal amplification chip specifically designed for K-type thermocouples. After laser trimming, it works with K-type thermocouples and features automatic cold junction temperature compensation, real-time alarm for thermocouple breakage, high output linearity, low power consumption, and a wide supply voltage range. When a thermocouple breaks or overheats, the alarm signal is connected to the external interrupt pin XINT1 of the DSP 2407A via an optical isolation device, triggering an external interrupt. The interrupt routine then sends the alarm signal to the central processing unit via the CAN bus. The thermocouple signal is amplified and conditioned by the AD595 chip, then compensated for by the cold junction temperature of the thermocouple by the variable potentiometer, and finally sent to the analog input channel ADC0 of the DSP 2407A via an operational amplifier to complete the sampling of the thermocouple signal. The thermocouple temperature measurement signal is processed internally by the DSP 2407A (including signal filtering, linear fitting of the thermocouple, and nonlinear correction), converted into a digital signal, and sent to the CAN controller in the 2407A. It is then transmitted to the CAN bus via a transmitter mailbox to achieve data exchange with the central processing unit. [align=center][img=500,203]http://www.e-works.net.cn/images/127982049468281250.GIF[/img] Figure 1 Schematic diagram of the intelligent temperature sensor circuit[/align][align=left] 3.2 Measurement Software Design The software mainly performs signal filtering, linear fitting of the thermocouple, and nonlinear correction on the sensor signal, and saves the results. When the sensor system receives a data receive command from the FADEC central processing unit, the DSP begins to transmit the current temperature value; when it receives an interrupt or mask command from the central processing unit, the DSP latches the current temperature value and does not transmit it. It is worth noting that the engine temperature sensor is dual-redundant, so the larger of the two measured temperature values ​​is used as the effective frequency value in the calculation. If the difference between the two values ​​exceeds 10% of the larger value, the effective value is selected while a fault signal for the other sensor is output. [align=center][img=203,335]http://www.e-works.net.cn/images/127982049829218750.GIF[/img] Figure 2 Flowchart of the main program for the intelligent temperature sensor[/align] Due to space limitations, only a section of the program is given here: .title “Sensor _CP” ; filename.bss GPR0,1 ; strobe register.include 2407A.h ; call register header file.copy “2407Avector.h” ; middle vector header file.def d_into ; define breakpoint D1 .equ 7000h ; define temporary storage unit D1 .text d_into:LDP #0h SETC INTM ; disable interrupt SPLK SXM SPLK #0000h,IMR ; mask all middle LAC CFR ; read interrupt flag SACL IFR Clear interrupt flag LDP #00E0h; Set DP=E0h SPLK #006Fh, WDCR; Disable watchdog SPLK #0000h, T1CNT; Clear counter 1 LOOP: SPLK #0F42h, T1CON; Enable timer counting, input clock is 1/128 of CPU clock SPLK #2711h, T1PR; Set period register value to 10001 LACL T1CNT; Load counter 1 value into accumulator SFR; Shift accumulator right by one bit SACL T1CNT, D1; Store result in D1 SPLK #0000h, T1CNT; Clear counter 1 SFR; Shift accumulator right by one bit SACL T1CNT, D1; Store result in D1 SPLK #0000h, T1CNT; Clear counter 1 LACL D1; Store result in D1 Load the value into the accumulator SACL 2407A.h; store the result into register LOOP. 4. Experimental Results To verify the measurement effect of the intelligent temperature sensor, the operation of a real speed sensor was simulated based on its design principle, and its input-output characteristic curves were obtained. The results are shown in Figure 3. [align=center][img=281,208]http://www.e-works.net.cn/images/127982050061250000.GIF[/img] Figure 3 Experimental Results[/align] Its least squares fitting formula and error results are shown in Table 1 below: Y = A + B ＊ X [align=center] Table 1 [img=318,70]http://www.e-works.net.cn/images/127982050261250000.gif[/img][/align] It can be seen that the intelligent sensor has high accuracy. In practical applications, the errors from intelligent sensors to the central processing unit mainly originate from two aspects: measurement errors inherent to the sensors themselves and errors generated under harsh flight conditions or strong electromagnetic interference. How to further eliminate these errors is a problem that requires further research. 5. Conclusion In summary, this intelligent temperature sensor signal replaces the FADEC system in performing simple functions such as signal excitation, digital filtering, A/D conversion, time averaging, measurement, linearization, and temperature compensation, while also enabling data transmission and status information reception. On the one hand, it reduces the weight of the entire control system, especially the cabling; on the other hand, the intelligent sensor design frees the FADEC system from performing low-level functions, allowing it to better implement complex control algorithms, thereby improving the engine's dynamic characteristics and overall performance.