Analysis of RF transceivers used in automotive

Transcription

1 Scientific Bulletin of Politehnica University Timisoara TRANSACTIONS on ELECTRONICS and COMMUNICATIONS Volume 60(74), Issue, 0 Analysis of RF transceivers used in automotive Camelia Loredana Ţeicu Abstract The paper presents the redesign of a radio frequency (RF) front-end transceiver used in automotive as part of the remote keyless entry (RKE) and tirepressure monitoring (TPMS) system. The redesign must keep the same RF performances or improve them in some cases. A practical impedance analysis is done to choose the best design in terms of performance versus costs for reception and transmission. Comparative measurements for the main characteristics, on the RF transceiver board, before and after the redesign, were realized to confirm the obtained performances. Keywords: RKE, TPMS, transceiver, SAW filter external power amplifier and a wiring connector for communication interface and power supply. For RKE messages, a two-frequency agility function will be used, which means that the first part of a transmission is at one frequency f (868.0MHz) and the other within MHz at f (868.MHz). Another receive path with matching to a second frequency band f 3 (433.9MHz) used for TPMS messages will be achieved by HW-switches controlled by the µc. I. INTRODUCTION Short-range radio systems designed to work in unlicensed industrial, scientific, and medical (ISM) bands between 300MHz and 98MHz use as key components the low-cost FSK and/or ASK transceiver ICs. Applications for these short-range devices (SRDs) include remote keyless entry and tire-pressure monitoring systems. Remote keyless entry is very popular in the new vehicles, because beside the advantages for comfort this technology minimizes the risk of theft. The new generation of RKE systems is capable to inform the users what is the status of the vehicle: if the doors are not locked or if needs more gas. [] The tire-pressure monitoring system (TPMS) is more and more requested by the automotive producers as well, because is increasing the safety of the driver and the efficiency of the driving. There are two types of TPM: indirect or direct. The direct TPM is more precise and easy to use than the indirect TPM, which is using the speed sensors from the anti-lock braking system (ABS). [] A. RF transceiver In this paper, the RF transceiver is operating in two ISM bands used in Europe, 434MHz for TPM and 868MHz for RKE. The transceiver unit (Fig..) consist of an internal antenna, an RFIC, matching for the transmit frequency for communication to remote unit (key), matching for RKE and TPMS receive frequencies, a micro controller (µc), a voltage regulator, a temperature sensor for a dynamic frequency compensation, a ISO94/LIN transceiver, an Fig.. Block diagram for RF transceiver The transceiver unit receives a radio signal from the corresponding transmitting unit and demodulates it. Then will analyze if the received signal is containing a valid RF data. B. RF transceiver characteristics Sensitivity is defined as the minimal signal level which can be properly received and demodulated by the receiver at the required bit error rate. [3] Sensitivity bandwidth is defined as the frequency spread between the half-power (3dB) points on the receiver response curve. Co-channel rejection measures the receiver s ability to receive the wanted signal in the presence of an unwanted signal at the same frequency. The limit for this RF transceiver is at +db S/N ratio. Adjacent channel rejection measures the receiver s ability to receive the wanted signal in the presence of interference at an adjacent channel. Each channel has two adjacent channels, lower and upper, with certain channel spacing. This rejection must be bigger than 30dB. Faculty of Electronics and Telecommunications, Communications Dept. Bd. V. Parvan, 3003 Timisoara, Romania, Camelia.Teicu@continental-corporation.com 8

2 Image frequency suppression is the ratio between the sensitivity for a signal at the image frequency to the sensitivity in the wanted channel. The RF transceiver must have a ratio bigger than 40dB. Dynamic range is an important performance characteristic. The dynamic range sets the maximum and minimum limits of the signal level which can be properly processed by the transceiver. The minimum limit is defined by the sensitivity level, whereas the maximum limit will be determined by the receiver s capability of processing the high signal level linearity.[3] The in-band selectivity (within f channel ±MHz) is a measure of the receiver s capability to detect a certain modulated signal without exceeding a given degradation due to the presence of an unwanted (un-modulated) signal within the reception band. Output power is the transmitting power level at a certain frequency, required for a good reception and decoding of the signal at the receiving unit. Harmonics are component frequencies of the signal that is an integer multiple of the fundamental frequency. These must be reduces in order to be in line with the radio regulations. Because the transmit channels are in the 868MHz ISM band, only the first two harmonics are measured. These are the main characteristics, but there are others, amongst: desensitization out-of band, spurious response in-band/out-of-band, intermodulation rejection 3 rd order, LO leakage, spurious emissions, adjacent channel power and occupied bandwidth. Best way to optimize the matching is to look in impedance at S and S parameters, because this two need to be close and with the same shape in a frequency span. Thus, the part to part distribution can be reduced. The measurement setup consists of a vector network analyzer (VNA) and cables soldered at specific test points on the board. The vector network analyzer must be initially calibrated and an offset for the cables must be introduced. The impedance analysis implies the practical measurements of the matching for the two SAW filters, each in five configurations:. the old SAW filter with the corresponding matching;. the new SAW filter with the old matching; 3. the new SAW filter with the matching specified by the filter producer in the datasheet; 4. the new SAW filter with a 0Ω optimized matching;. the new SAW filter with the final matching, which will be implemented with the redesign. The matching schematics were done in Microwave Office, which is a software used for RF design and simulation. The first configuration has the matching in a π topology for RKE band (Fig..) and an L topology for TPMS band (Fig. 3.): II. REDESIGN OF THE RF TRANSCEIVER Due to evolution of technology and the increasing market, components suppliers tend to improve their products by releasing other models for the same component. This improvement can be the quality of the product or just to reduce the costs for production. Therefore there is a constant need in automotive to redesign specific modules to be compliant with other components. In the present paper, the RF transceiver has two surface acoustic wave (SAW) filters, for reception, which can select between the RKE and TPM bands. These SAW filters, namely B376 and B3780, have 8 pins and will no longer be produced, but replaced by B3744 and B3936 filters. This requires the transceiver to be redesigned. Redesigning includes also a printed circuit board (PCB) supplier replacement and a layout change, because the new SAW filters have 6 pins with different configuration than the previous versions. A. Analysis in impedance for reception Fig.. Old SAW filter with the actual matching for RKE band Fig. 3. SAW filter with the actual matching for TPMS band Prior to redesign, the transceiver has good performances for reception in both bands. The configurations from Fig. and Fig. 3 have the impedance at channel frequencies shown below (Fig. 4-7). The markers are at 868,0MHz and 868,MHz for RKE band, and at 433,9MHz for TPMS band. First step in designing a matching network for SAW filters is to measure on the board the insertion loss and the selectivity of the filters, but also the S and S parameters to see the reflection at port or at port. 9

3 CH S U : j.694 CH S U :.3 j MHz MHz : 34. j9.6 : 0.4 -j MHz MHz k 0k CENTER 868 MHz SPAN 0 MHz CENTER MHz SPAN 0 MHz Fig. 4. S parameter for Fig. configuration Fig. 7. S parameter for Fig. 3 configuration CH S U 0. : 4.9 j MHz : 83.6 j MHz Fig. 8. New SAW filter for RKE band, B3744 with the old matching - 0k CENTER 868 MHz SPAN 0 MHz Fig.. S parameter for Fig. configuration CH S U 0. : 46.6 j MHz : 4.63 j MHz Fig. 9. New SAW filter for RKE band, B3936 with the old matching k CENTER MHz SPAN 0 MHz Fig. 6. S parameter for Fig. 3 configuration Fig. 0. S parameter for Fig. 8. configuration 0

4 Fig.. S parameter for Fig. 8. Configuration The second configuration analyzed (Fig. 8-9) shows good results for the RKE SAW filter, but for the TPMS SAW filter the performances are worse than the old ones. There can be noticed (Fig. 0-3) the PCB and layout influence and the new SAW filters characteristics. The third and the fourth configurations were done in order to see if the filter s producer recommendations are reproducible on every type of PCB and how much improvement the matching brings. Finally, for RKE SAW filter it was decided to keep the second configuration, because the sensitivity at channels frequencies is still compliant with the requirements and there s no need to change the components, resulting in a cost reduction for this redesign. For the TPMS SAW filter, the final matching is presented in fifth configuration, that is the filter producer matching, but without one capacitor, to keep the same number of components in the bill of materials (BOM). In this case, a compromise was made, because the old way of doing the matching of the filter is not relevant due to the fact that the whole reception part has new impedances along the traces and at the test points there is no more a 0Ω impedance. So the solution to decide which is the best matching configuration is to measure the reception sensitivity. An overview based on the measurements done for the five matching configurations is presented below (Table ). We can see also the measurements done for the selectivity of the SAW filters. Table. Matching configurations results Fig.. S parameter for Fig. 9. configuration B. Analysis in impedance for transmission Fig. 3. S parameter for Fig. 9. configuration Because this is a redesign, the target is to keep at least the same performances. A solution is to compare the load impedance that the integrated power amplifier needs for best operation. In the transmission part, only the PCB is replaced. The expectations are that the influences of the PCB stack-up to be small since the new stack-up has the same configuration with the old one. The load impedance at the PA is measured on the transceiver with a VNA, when the RFIC is disconnected, the active components are on and the

5 HW-switch is controlled for transmission, which leads to a 0Ω termination instead of the antenna. Before the redesign, the load impedance measured at the integrated PA is 8,3 + j9,3ω at 868.0MHz and 8,9 + j9,6ω at 868.MHz (Fig. 4). a design validation, more samples must be measured to see the part to part distribution, due to components tolerances and other small influences. A. Setup configuration M CH U 0. : 8.3 j MHz : 8.9 j MHz The setup (Fig. 6) must simulate the corresponding RKE remote unit and the TPMS sensor and also to provide the interference signal for the tests. CAI k START 680 MHz STOP. GHz Fig. 4. Load impedance at the integrated PA before redesign After the redesign, the load impedance measured at the integrated PA is j7,ω at 868.0MHz and 0,79 + j7,66ω at 868.MHz (Fig. ). The difference is small, only Ω in the real part. In this case, nothing else is modified in the matching. The transmitter characteristics will remain the same and the output power is already calibrated on the production line, by increasing the gain and power steps of the PA. M CH U 0. : 0.74 j MHz : 0.79 j MHz Fig. 6. Test setup configuration The measurements involve the application of the RF signal at the 0Ω reference by disconnecting the antenna matching and inserting a coaxial cable. The transceiver is inside the shielded box, where it is powered with V and the response on the ISO94 is monitored with the PC. The PC also commands the trigger box and all the generators. The RF parameters corresponding to every type of telegram are introduced with a specific application on the PC. The test setup is calibrated on each frequency, so the attenuation inserted by the cables can be correlated in the results. For output power and harmonics measurement a spectrum analyzer and a 0dB attenuator were included in the setup. CAI B. Comparative results k The following main characteristics were measured with the setup presented above: sensitivity at center frequency for each channel, sensitivity bandwidth, image frequency suppression, adjacent channel rejection, co-channel rejection, in band selectivity, dynamic range, output power and harmonics. A summarized table is presented below (Table ): - START 680 MHz STOP. GHz Table. Comparison of measured characteristics before and after redesign Fig.. Load impedance at the integrated PA after redesign III. MAIN CHARACTERISTICS MEASUREMENTS In this section, the main characteristics of the RF transceiver are measured to see if the RF performance is still compliant with the design before the redesign. For these measurements two boards were used to compare the results before and after redesign. But for

6 The measuring method is very important as it can have significant influence on the results. A short description of every measurement is explained below: Sensitivity at center frequency: the power level is adjusted until the sensitivity limit is reached. Sensitivity bandwidth: the frequency will be sweep in a 300 khz span, with khz step. For each step the power level is adjusted until the sensitivity limit is reached. The frequency and power level are recorded. Image frequency suppression: the same signal is applied at center frequency and at image frequency. The sensitivity is measured for each frequency and recorded. The difference between sensitivities is the image frequency suppression. Adjacent channel rejection: the signal is set at center frequency f c and the FM interference is set first at f c -channel spacing and then at f c +channel spacing. The S/N is recorded. Co-channel rejection: the signal is set at center frequency f c and the FM interference is set at same frequency. The S/N is recorded. Selectivity in band f c ±MHz: the signal center frequency is set at channel frequency and the continuous wave (CW) interference is swept in a MHz span, with khz step. S/N is recorded. Dynamic range: the signal set at center frequency will be increased from -0dBm to 0dBM with 0dB steps. The errors are recorded at each step.[4] Output power and harmonics: the transceiver is put in transmitting mode, and the power level is measured with the spectrum analyzer. Fig. 9. Selectivity in band for TPMS channel The results after the redesign are close to the old design, only the sensitivity is improved with around db for the RKE channels (Fig. 7-8). This proves that the RF performance is still compliant with the transceiver requirements. IV. CONCLUSIONS The paper presents a redesign of an RF transceiver front-end used in automotive, dual band and dual channel system. It is important to have at least same performance as initial even if the technology is evolving, and the main components must be changed. In this case, the SAW filters used for reception and the PCB supplier changed. The introduction of these new components is analyzed in impedance for reception and transmission to see the effects and compensate them. Also the main characteristics and the way they are tested are presented to prove the performance conformity with the old design. It can be observed that the solution for redesign is better in some cases than the old one, but only because the cumulative effect of the filter s matching and the new PCB layout helps. The conclusion is that after each change in the RF area, the design need to be adjusted. There is no one by one replacement for any component used in the front-end of the transceiver. Impedance can be easy shifted if any of the elements between antenna and RFIC is modified. V. REFERENCES Fig. 7. Selectivity in band for first RKE channel [] *** AN389 Power Amplifier Theory for High-Efficiency Low- Cost ISM-Band TransmittersApplication Note, Maxim integrated, 00. [] *** TPMSWP Freescale Single-Package Tire Pressure Monitoring System (TPMS) White Paper, Freescale Semiconductor, 007. [3] *** ATA743 Integrating an External LNA Application Note, Atmel Corporation, 006. [4] C.L. Teicu, Analiza Transceiverelor de RF Utilizate in Industria Automotive Master Thesis, Politehnica University Timisoara June 04. Fig. 8. Selectivity in band for second RKE channel 3