Printer Friendly

Prediction Method for Automobile EMI Test Result in AM Frequency Band.

1. INTRODUCTION

In the international standard, CISPR 25, measuring methods of radiated electromagnetic noise for automobiles and components are defined. Regarding the automobile test, the radiated electromagnetic noise from the component installed in the automobile is measured by the antenna of the automobile. On the other hand on the component test, the radiated electromagnetic noise from the component is measured by the mono-pole antenna set forward of the component. The EMI of automobiles and the components are often evaluated referring to CISPR 25. However, the components sometimes fail the automobile test even if they passed the component test due to the difference of the method. In this case, the component has to be designed again. For example, power electronics components have some high power switching devices that the switching frequency is near the radio frequency band, called AM frequency band in Japan, from 526.5kHz to 1606.5kHz. The electromagnetic noise by switching will interfere with an AM radio receiver . The interference is sometimes found by the automobile test after the component test has been finished. Therefore, the prediction method for the automobile test result is required. In References [2], [3] and [4], the radiated electromagnetic noise propagation mechanism of the component test was investigated. And in References [4] and [5], radiated electromagnetic noise propagation mechanism of the automobile test was investigated. However, the difference of the mechanism between the automobile test and the component test was not discussed sufficiently. Moreover, the prediction method of the automobile test result was not developed. In this paper, we tried to modify the standard component test configuration to predict the automobile test result. In this study, a fuel pump system whose component test results did not agree with the automobile test results was used. The fuel pump system consisted of a fuel pump, wire harnesses, and a controller.

This paper is organized as follows. In Section 2, main investigation of this paper is shown. In Section 3, the investigative study results for the radiated electromagnetic noise propagation in the automobile test system is discussed. In Section 4, the investigative study results for the radiated electromagnetic noise propagation in the component test system is discussed. In Section 5, the radiated electromagnetic noise propagation in the component test system is compared with that in the automobile test system. In Section 6, the standard component test is modified to predict the automobile test result.

2. INVESTIGATION OF RADIATED ELECTROMAGNETIC NOISE PROPAGATION PATHS IN COMPONENT TEST SYSTEM AND AUTOMOBILE TEST SYSTEM

Reference [5] is an example of an earlier study on radiated electromagnetic noise propagation paths. This paper reported that the parasitic capacitance between the automobile body and the earth forms one of the radiated electromagnetic noise propagation paths. The radiated electromagnetic noise from the fuel pump generates the voltage in the parasitic capacitance between the automobile body and the earth as discussed in Reference [5]. This voltage is divided by the parasitic capacitance between the antenna and the automobile body and the parasitic capacitance between the antenna and the earth. The voltage between the antenna and the automobile body interferes with the AM radio receiver. To further explore the above study results, the radiated electromagnetic noise propagation path was investigated with a focus on the difference of the location of parasitic capacitance between the component test system and the automobile test system.

3. RADIATED ELECTROMAGNETIC NOISE PROPAGATION IN AUTOMOBILE TEST SYSTEM

This section discusses investigative study results for the radiated electromagnetic noise propagation in the automobile test system. Radiated electromagnetic noise from radiation sources propagate to the antenna.

3.1. Radiation Sources

Radiation sources are described first. The fuel pump system mounted on an automobile is schematically illustrated in Figure 1. The fuel pump was installed in a plastic fuel tank mounted outside the automobile and soaked in gasoline. The controller was installed inside the automobile and connected to the fuel pump through wire harnesses. This means that wire harnesses were laid inside the automobile. The controller generated PWM signal to control the fuel pump. The harmonic of the PWM signal overlapped the AM radio frequency band and interfered with the AM radio receiver. In this system, three radiation sources, [V.sub.wb_a], [V.sub.pb_a] and [V.sub.cb_a], existed as shown in Figure 2, which was a simplified form of Figure1. Three radiation sources are explained below.

The harmonic of the PWM signal transmitted from the controller partly leaked into wire harnesses as the noise current and propagated into the automobile body through the parasitic capacitance, [C.sub.wb_a], between wire harnesses and the automobile body. In this process, the voltage, [V.sub.wb_a], was generated in [C.sub.wb_a] and acted as a radiation source. The remaining noise current reached the fuel pump through wire harnesses propagated into the automobile body through the parasitic capacitance, [C.sub.pb_a], between the fuel pump enclosure and the automobile body. In this process, the voltage, [V.sub.pb_a], was generated in [C.sub.pb_a], and also acted as a radiation source. The noise current that had propagated into the automobile body returned to the controller through the parasitic capacitance, [C.sub.cb_a], between the controller and the automobile body. In this process, the voltage, [V.sub.cb_a], was generated in [C.sub.cb_a] and also acted as a radiation source.

3.2. Radiated Electromagnetic Noise Propagation from the Fuel Pump

Next, the radiated electromagnetic noise propagation from radiation sources to the antenna are discussed. The radiated electromagnetic noise propagation path from the component to the parasitic capacitance between the automobile body and the earth, which is not discussed in Reference [5], was investigated. Based on the results, radiated electromagnetic noise propagation paths in the automobile test system are discussed. As an example, the radiation of the electromagnetic noise from the fuel pump, [V.sub.pb_a], was used.

As shown in Figure 3, parasitic capacitances, [C.sub.pe_a] and [C.sub.be_a], existed between the fuel pump and the earth, and between the automobile body and the earth, respectively. When taking the above fact into account, the radiated electromagnetic noise propagation from [V.sub.pb_a] to [C.sub.be_a] could be represented by the equivalent circuit shown in Figure 4. In this Figure, [C.sub.pb_a] is omitted due to no relation with the radiated electromagnetic noise propagation. [C.sub.pe_a] and [C.sub.be_a] were considered to act as radiated electromagnetic noise propagation paths. [V.sub.pb_a] was divided by [C.sub.pe_a] and [C.sub.be_a] so that the voltage, [V.sub.be_a], was generated in [C.sub.be_a]. Therefore, applying Ohm's law, [V.sub.be_a] was calculated by Equation (1).

[mathematical expression not reproducible] (1)

Where, units of [V.sub.be_a] and [V.sub.pb_a] are dBuV. Moreover, considering that the equivalent circuit of Figure 5 in Reference [5], the electromagnetic noise propagation from the fuel pump to the antenna in the automobile test system could be represented by the equivalent circuit shown in Figure 6. Therefore, [V.sub.ab_a], the antenna inductive voltage as a measure to evaluate EMI of components in accordance with CISPR 25, was calculated by Equation (2).

[mathematical expression not reproducible] (2)

Where, a unit of [V.sub.ab_a] is dBuV, and [C.sub.ae_a] and [C.sub.ab_a] are parasitic capacitances between the antenna and the earth, and between the antenna and the automobile body, respectively. The second term and the third term are propagation losses of the radiated electromagnetic noise. To check the validity of the equivalent circuit, [V.sub.ab_a] calculated by Equation (2) was compared with [V.sub.ab_a] actually measured. [C.sub.pe_a], [C.sub.be_a], [C.sub.ae_a] and [C.sub.ab_a] were determined from [S.sub.11], impedance, that was measured by a network analyzer. [V.sub.pb_a] was measured by a high frequency probe and a spectrum analyzer. [V.sub.ab_a] was measured by the spectrum analyzer as the output of the antenna. The comparison result is shown in Figure 7. In this figure, the solid line represents the measured values, while the broken line represents the values calculated by Equation (2). The calculated values agreed with the measured values within a tolerance of approximately 4dB, verifying the validity of the equivalent circuit shown in Figure 6.

3.3. Radiated Electromagnetic Noise Propagation from Wire Harnesses and the Controller

With regard to the electromagnetic noise radiation from wire harnesses and the controller, parasitic capacitances, [C.sub.wa_a] and [C.sub.ca_a], also existed between wire harnesses and the antenna, and between the controller and the antenna as shown in Figure 8, respectively. The above-described [V.sub.wb_a] and [V.sub.cb_a] were generated at the positions shown in Figure 8. The electric circuit equivalent to Figure 8 is shown in Figure 9. This figure represents the radiated electromagnetic noise propagation in the entire test system. In this figure, [C.sub.wb_a], [C.sub.pb_a] and [C.sub.cb_a] are omitted due to no relation with the radiated electromagnetic noise propagation.

4. RADIATED ELECTROMAGNETIC NOISE PROPAGATION IN COMPONENT TEST SYSTEM

This section discusses investigative study results for the radiated electromagnetic noise propagation in the component test system. Radiated electromagnetic noise from radiation sources propagate to the antenna in the same manner as in the automobile test system.

4.1. Radiation Sources

Radiation source are described first. According to the component test method specified in CISPR 25, DUT, device under test, should be placed on a ground plane, copper plate that is grounded to the floor on a table, as shown in Figure 10. The mono-pole antenna should be used as the test antenna. The fuel pump soaked in a pseudo-fluid was controlled by the controller that was connected to the fuel pump through wire harnesses. In the component test system, the harmonic of the PWM signal generated by the controller interfered with the AM radio receiver as is the case in the automobile test system. In this system, three radiation sources, [V.sub.wg_c], [V.sub.pg_c] and [V.sub.cg_c], existed as shown in Figure 10. Three radiation sources are explained below. The harmonic of the PWM signal transmitted from the controller partly leaked into wire harnesses as the noise current and propagated to the ground plane through the parasitic capacitance, [C.sub.wg_c], between wire harnesses and the ground plane. In this process, the voltage, [V.sub.wg_c], was generated in [C.sub.wg_c] and acted as a radiation source. The remaining noise current reached the fuel pump through wire harnesses propagated to the ground plane through the parasitic capacitance, [C.sub.pg_c], between the fuel pump enclosure and the ground plane. In this process, the voltage, [V.sub.pg_c], was generated in [C.sub.pg_c], and also acted as a radiation source. The noise current that had propagated to the ground plane returned to the controller through the parasitic capacitance, [C.sub.cg_c], between the controller and the ground plane. In this process, the voltage, [V.sub.cg_c], was generated in [C.sub.cg_c] and also acted as a radiation source.

4.2. Radiated Electromagnetic Noise Propagation from the Controller

Next, the radiated electromagnetic noise propagation from radiation sources to the antenna are discussed. As an example, the radiation of the electromagnetic noise from the controller, [V.sub.cg_c], was used. As shown in Figure 11, parasitic capacitances, [C.sub.ca_c] and [C.sub.ag_c], existed between the controller and the antenna, and between the antenna and the ground plane, respectively. When taking the above fact into account, the radiated electromagnetic noise propagation from [V.sub.cg_c] to the antenna could be represented by the equivalent circuit shown in Figure 12. [C.sub.ca_c] and [C.sub.ag_c] were considered to act as radiated electromagnetic noise propagation paths. [V.sub.cg_c] was divided by [C.sub.ca_c] and [C.sub.ag_c]. Therefore, applying Ohm's law, [V.sub.ag_c], the induced antenna voltage as a measure to evaluate EMI of components in accordance with CISPR 25, was calculated by Equation (3).

[mathematical expression not reproducible] (3)

Where, units of [V.sub.ag_c] and [V.sub.cg_c] are dBuV, and the second term is the propagation loss of the radiated electromagnetic noise. To check the validity of the equivalent circuit, [V.sub.ag_c] calculated by Equation (3) was compared with [V.sub.ag_c] actually measured. [C.sub.ag_c] and [C.sub.ca_a] were determined from [S.sub.11], impedance, that was measured by the network analyzer. [V.sub.cg_c] was measured by the high frequency probe and the spectrum analyzer. [V.sub.ag_c] was measured by the spectrum analyzer as the output of the antenna. The comparison result is shown in Figure 13. In this figure, the solid line represents the measured values, while the broken line represents the values calculated by Equation (3). The calculated values agreed with the measured values within a tolerance of approximately 3dB, verifying the validity of the equivalent circuit shown in Figure 12.

4.3. Radiated Electromagnetic Noise Propagation from Wire Harnesses and the Fuel Pump

With regard to the electromagnetic noise radiation from wire harnesses and the fuel pump, parasitic capacitances, [C.sub.wa_c] and [C.sub.pa_c], also existed between wire harnesses and the antenna, and between the fuel pump and the antenna as shown in Figure 14, respectively. The above-described [V.sub.wg_c] and [V.sub.pg_c] were generated at the positions shown in Figure 14. The electric circuit equivalent to Figure 14 is shown in Figure 15. This figure represents the radiated electromagnetic noise propagation in the entire test system. In this figure, [C.sub.wg_c], [C.sub.pg_c] and [C.sub.cg_c] are omitted due to no relation with the radiated electromagnetic noise propagation.

5. COMPARISON OF RADIATED ELECTROMAGNETIC NOISE PROPAGATION IN AUTOMOBILE TEST SYSTEM AND COMPONENT TEST SYSTEM

Using results presented in Section 3.3 and 4.3, the radiated electromagnetic noise propagation in the automobile test system was compared with that in the component test system. From the comparison results shown in Figure 16, the radiated electromagnetic noise propagation differs between the two test systems. The most significant difference was in the radiated electromagnetic noise propagation path from the fuel pump. Four parasitic capacitances, [C.sub.pe_a], [C.sub.be_a], [C.sub.cb_a], and [C.sub.ae_a], affected the radiated electromagnetic noise propagation from the fuel pump in the automobile test system, whereas two parasitic capacitances, [C.sub.pa_c] and [C.sub.ag_c], affected the radiated electromagnetic noise propagation from the fuel pump in the component test system. From the above results, the propagation loss of radiated electromagnetic noise would differ between the two test systems. The each parasitic capacitance value would also differ between the two test systems since the fuel pump system was set differently in these test systems. In addition, the voltage was generated when noise current flowed through a parasitic capacitance and this voltage acted as a radiation source. Since the each parasitic capacitance value differed between the two test systems, the strength of noise radiation sources also differed between the two systems.

6. MODIFICATION AND DISCUSSION OF TEST SYSTEM THAT CAN YIELD TEST RESULTS HIGHLY CORRELATED WITH AUTOMOBILE TEST RESULTS

In this section, we tried to modify the component test system that can yield test results highly correlated with automobile test results. To achieve the purpose, the following three procedures should be followed.

1. Change the radiated electromagnetic noise propagation in the component test system, which is represented by the equivalent circuit shown as the component test system in Figure 16, to the equivalent circuit shown as the automobile test system in the same figure.

2. Since the propagation loss of radiated electromagnetic noise depends on the parasitic capacitance value, adjust the parasitic capacitance value so that the propagation loss of radiated electromagnetic noise in the automobile test system can be realized.

3. Equalize the strength of electromagnetic noise radiation sources by adjusting parasitic capacitance values between the fuel pump system and the ground the plane equal to parasitic capacitance values between the fuel pump system and the automobile body.

To simulate the equivalent circuit for the automobile test system according to Procedure (1), the ground plane is disconnected from the floor as shown in Figure 17. Following the above, the fuel pump was relocated to a space between the ground plane and the floor. In this setup, the radiated electromagnetic noise propagation can be represented as shown by the equivalent circuit in Figure 18. It is same configuration as the Figure 9, the radiated electromagnetic noise propagation in automobile test system.

Next, as in Procedure (2), the parasitic capacitance value was adjusted. First, the distance between the objects in the component test system were adjusted. In other words, when the parasitic capacitance value increase, the objects in the test system are brought closer, while when the parasitic capacitance value decrease, the objects are separated. When the above method was ineffective to obtain a desired parasitic capacitance value, a capacitor was additionally connected in parallel with the parasitic capacitance.

Propagation losses of radiated electromagnetic noise in the test system modified and adjusted according to Procedure (1) and (2), were measured and compared with that in the automobile test system, to check whether the test system simulated propagation losses of radiated electromagnetic noise in the automobile test system. Propagation losses of radiated electromagnetic noise were measured by a network analyzer, NA, as [S.sub.21]. The port 1 of the NA was connected to an electromagnetic noise radiation source and the port 2 of the NA was connected to the antenna. The comparison result example is shown in Figure 19. In this figure, the broken line and the solid line represent the propagation loss of radiated electromagnetic noise in the automobile test system and in the modified component test system, respectively. As this figure shows, the modified component test system simulated the propagation loss of radiated electromagnetic noise in an automobile test system within a tolerance of 2dB.

Subsequent to the above procedure, the strength of electromagnetic noise radiation sources were equalized according to Procedure (3) by adjusting the distance of the objects and adding capacitors as needed as with the case of Procedure (2). Moreover, the parasitic capacitance inside the fuel pump, [C.sub.ne], was a noise current path as shown in Figure 20. Thus, [C.sub.ne] must have been a factor to determine a strength of electromagnetic noise radiation source and was filled with the fluid that the fuel pump was soaked. Considering above, the fluid that dielectric constant is close to gasoline was used to soak the fuel pump. In this test system, the radiation source voltages were measured by the high frequency probe and the spectrum analyzer to check whether the voltages were same as that of the automobile test system. The result example is shown in Figure 21. In this figure, the solid line represents the [V.sub.cg_m], shown in Figure 17, while the broken line represents the [V.sub.cb_a] shown in Figure 8. The test data obtained from the two test systems agreed with each other within a tolerance of approximately 4dB, which confirmed the modified component test method can simulate the electromagnetic noise source in the automobile test system.

To evaluate the modified component test system, the induced antenna voltage, [V.sub.ag_m], was measured and compered with [V.sub.ab_a] and [V.sub.ag_c]. Because [V.sub.ag_m], [V.sub.ab_a] and [V.sub.ag_c] are measures for evaluating EMI of components. As shown in Figure 22, [V.sub.ag_m] was about 30dB closer to [V.sub.ab_a] than [V.sub.ag_c]. In addition, the error between [V.sub.ag_m] and [V.sub.ab_a] was approximately 6dB, meaning automobile test result could be predicted by the modified component test system.

From these test results, we confirmed that, in both component test system and automobile test system, the radiated electromagnetic noise propagation in the AM frequency band is dominantly affected by the parasitic capacitances surrounding the antenna and the DUT.

Therefore, to develop a component test system that can commonly be used for sedans, minivans, and all other types of automobiles, to determine parasitic capacitances existing in the automobile test system is needed. Parasitic capacitances in the automobile test system depend on the shape, positional relation, and other three-dimensional features of objects in the automobile test system. Since an approximate three-dimensional CAD model of each object enables the determination of the approximate the parasitic capacitance value in every object of automobiles having various external shapes using Q3D or other simulator, the modified component test system can be constructed for all types of automobiles.

7. CONCLUSION

In this paper, the radiated electromagnetic noise propagation from the fuel pump system in the automobile test system and the component test system were investigated. In particular, the parasitic capacitances in the two test systems were focused on, the radiated electromagnetic noise propagations were expressed with equivalent circuits which consisted of the parasitic capacitances, and these circuits were compared. As a result, configurations of the circuits differed from each other and this difference caused a difference in the propagation loss of radiated electromagnetic noise between the two test systems. Based on the above result, we tried to modify a component test system that can yield test results highly correlated with the automobile test results. In practice, the ground plane was disconnected from the floor, the fuel pump was relocated to a space between the ground plane and the floor, and the parasitic capacitance value in each objects was adjusted in the component test system. As a result, the modified component test system could yield test results highly correlated with automobile test results. The parasitic capacitance value, which greatly affects the propagation loss of radiated electromagnetic noise, can be calculated by Q3D or another simulator, if a three-dimensional CAD model of each object is provided. This means that the component test system can be constructed to be useful for all automobiles with different external shapes and yields test results highly correlated with the automobile test results. Our future task is to confirm that the method to modify the component test presented in this paper can be commonly used for all automobiles with different external shapes.

REFERENCES

1. IEC INTERNATIONAL STANDARD, "Vehicles, boats and internal combustion engines - Radio disturbance characteristics -Limits and methods of measurement for the protection of onboard receivers," CISPR 25, Edition 3.0, Mar. 2008.

2. Jin, Jia., Denis, Rinas., Stephan, Frei., "Prediction of Radiated Fields from Cable Bundles based on Current Distribution Measurements, "EMC Europe, Sept.2012.

3. Jin, Jia., Denis, Rinas., Stephan, Frei., "An Alternative Method for Measurement of Radiated Emissions According to CISPR 25, "EMC Europe, Sept.2013.

4. Stephan, Frei., "Where we Stand Today with Automotive EMC Simulation," EMC Europe, Sept.2013.

5. Kawai, K., Tsuda, T., Uno, T., Taki, H., Ichikawa, K., "Radio Noise in AM-band by Common Voltage which Occurs in the Electric Capacity between Vehicles and the Ground" JSAE Annual Spring Congress, May.2012.

CONTACT INFORMATION

Takashi Nomura

takasi_nomura@denso.co.jp

Takashi Nomura and Kazuma Kawai

DENSO Corporation

doi:10.4271/2017-01-0014
COPYRIGHT 2017 SAE International
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:electromagnetic interference; amplitude modulation
Author:Nomura, Takashi; Kawai, Kazuma
Publication:SAE International Journal of Passenger Cars - Electronic and Electrical Systems
Article Type:Technical report
Date:May 1, 2017
Words:3957
Previous Article:Centralized Lifecycle Management of Model-Based Software Artifacts.
Next Article:Calculation of Failure Detection Probability on Safety Mechanisms of Correlated Sensor Signals According to ISO 26262.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters