Dither edge softening

Abstract

A method of generating a pulse width modulated (PWM) waveform for providing power to an electric motor for use in a vehicle which reduces an amount of electromagnetic interference is disclosed. The method includes generating a PWM waveform, applying frequency dithering to the PWM waveform, and providing the PWM waveform to the electric motor.

Description

DITHER EDGE SOFTENINGCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application 63 / 719,371, filed on November 12, 2024, the disclosure of which is hereby incorporated by reference in its entirety.FIELD OF THE INVENTION

[0002] The present invention relates to a method of providing a pulse width modulation (PWM) signal to an electric motor used in an automotive vehicle. More specifically, the present invention relates to a method of reducing the amount of electromagnetic interference (EMI) produced when generating the PWM signal for controlling the rotational speed an electric motor.DESCRIPTION OF RELATED ART

[0003] Many vehicles today have an electric motor which is configured to rotate at variable speeds in response to a voltage amplitude provided to the electric motor. Typically, the rotational speed of the electric motor decreases as the voltage amplitude decreases. One method of varying the voltage amplitude is using pulse width modulation (PWM) to produce a digital signal. PWM adjusts a magnitude of an average voltage provided to the electric motor by adjusting the relative number of on / off pulses in the digital signal, described as a duty cycle. For example, a 50% duty cycle includes a series of pulses with 50% on time and 50% off time. A 75% duty cycle includes a series of pulses with 75% on time and 25 % off time. The amount of power provided by the digital signal to the electric motor is proportional to the duty cycle generated by PWM.

[0004] However, a digital signal generated using PWM may produce electromagnetic interference (EMI) which is conducted through the electrical supply and may affect the voltage and / or the current. Further, electric parts connected to the electric circuit, including the electric motor, may radiate electromagnetic noise. With no suppression mechanisms, this electromagnetic noise can generate unwanted artifacts which can negatively impact other electronic devices. For this reason, automotive OEM’s and other regulatory agencies typically specify a maximum amount of allowable EMI generated by electrical circuits and other electrical components. Typically,electrical circuits which use PWM to control an electric motor include hardware-based countermeasures to limit the amount of EMI produced.

[0005] It is desirable to reduce the amount of EMI generated when using PWM by using softwarebased implementations using a microprocessor. Further, it is desirable to combine software-based methods of reducing EMI with hardware-based countermeasures.SUMMARY OF THE INVENTION

[0006] According to one embodiment, there is a provided a method of generating a pulse width modulated (PWM) waveform for providing power to an electric motor for use in a vehicle. The method includes generating a PWM waveform, applying frequency dithering to the PWM waveform, and providing the PWM waveform to the electric motor.

[0007] According to a second embodiment, there is provided a method of generating a pulse width modulated (PWM) waveform for providing power to an electric motor for use in a vehicle. The method includes generating a PWM waveform, applying transition softening to the PWM waveform, and providing the PWM waveform to the electric motor.BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Advantages of the present invention will be readily appreciated as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings wherein:

[0009] Figure 1 is a diagram showing a controller electrically connected to an electric motor, according to one embodiment of the present disclosure;

[0010] Figure 2 is a graph of a first pulse width modulated (PWM) waveform which is sinusoidal in a time domain with a fixed frequency of 10 hertz, according to a first embodiment of the present invention;

[0011] Figure 3 is a graph of a first power spectrum of a first frequency waveform generated by performing a fast Fourier transform on the first PWM waveform of Figure 2;

[0012] Figure 4 is a graph of a second power spectrum of a second frequency waveform generated by performing a fast Fourier transform on a second PWM waveform generated by dithering the first PWM waveform of Figure 2, according to a second embodiment of the present invention;

[0013] Figure 5 is an enlarged view of the graph of the second power spectrum of Figure 4, showing a spectral content in a frequency range of 9.9 Hz to 10.1 Hz;

[0014] Figure 6 is graph of a third PWM waveform, which has a square waveform in the time domain with a static frequency of 10 hertz, according to a third embodiment of the present invention;

[0015] Figure 7 is a graph of a third power spectrum of a third frequency waveform generated by performing a fast Fourier transform on the third PWM waveform of Figure 6;

[0016] Figure 8 is a graph of a fourth power spectrum of a fourth frequency waveform generated by performing a fast Fourier transform on a fourth PWM waveform, which was generated by dithering the third PWM waveform of Figure 6, according to a fourth embodiment of the present invention;

[0017] Figure 9 is a graph of a fifth PWM waveform, which was generated by applying transition softening to the third PWM waveform of Figure 6, according to a fifth embodiment of the present invention;

[0018] Figure 10 is a graph of a fifth power spectrum of the fifth PWM waveform of Figure 9;

[0019] Figure 11 is a graph of a sixth PWM waveform, which was formed by dithering and applying transition softening to the third PWM waveform of Figure 6, according to a sixth embodiment of the present invention;

[0020] Figure 12 is a table showing the attenuation in the frequency domain of the third PWM waveform of Figure 6 and the fourth PWM waveform of Figure 8;

[0021] Figure 13 is a table showing the attenuation in the frequency domain of the third PWM waveform of Figure 6 and the fifth PWM waveform of Figure 10; and

[0022] Figure 14 is a table showing the attenuation in the frequency domain of the fourth PWM waveform of Figure 8 and the sixth PWM waveform of Figure 11.DETAILED DESCRIPTION OF THE INVENTION

[0023] Figures 1-14 illustrate waveforms 10 produced by pulse width modulation (PWM) for providing power to an electric motor 12 for use in an automotive vehicle, according to embodiments described herein. Directional references employed or shown in the description, figures or claims, such as top, bottom, upper, lower, upward, downward, lengthwise, widthwise, left, right, and the like, are relative terms employed for ease of description and are not intended to limit the scope of the invention in any respect. Referring to the Figures, like numerals indicate like or corresponding parts throughout the several views.

[0024] Depicted in Figure 1, an exemplary electric motor 12 is electrically connected to a controller 14 by a wiring harness 16. The controller 14 provides electrical power 18 to the electric motor 12 through the wiring harness 16. The electric motor 12 is configured to rotate at variable speeds (arrow 20) in response to a voltage amplitude of the power 18 provided to the motor 12, as is commonly known in the art. Typically, the rotational speed 20 of the electric motor 12 decreases as the voltage amplitude of the power 18 decreases.

[0025] One method of varying the voltage amplitude of the power 18 is using pulse width modulation (PWM) to produce a digital signal 22. The controller 14 includes a PWM module 24, which generates the PWM digital signal 22 for providing power 18 to the electric motor 12, as is commonly known in the art. The PWM module 24 adjusts a magnitude of an average voltage provided to the electric motor 12 by adjusting the relative number of on / off pulses in the digital signal 22, described as a duty cycle. The amount of power 18 provided by the digital signal 22 to the electric motor 12 is proportional to the duty cycle generated by the PWM module 24.

[0026] The PWM digital signal 22 generated by the PWM module 24 may produce electromagnetic interference (EMI), which is conducted through the electrical supply and may affect the voltage and / or the current. Further, electric parts connected to the electric circuit, including the electric motor 12, may radiate electromagnetic noise.

[0027] The controller 14 includes a microprocessor 26, which uses software-based implementations to reduce the amount of EMI generated when using PWM, according to embodiments of the present invention. The controller 14 also includes one or more hardware-based countermeasures 28, which are configured to reduce the EMI generated when using PWM, as is commonly known in the art. Further, the controller 14 combines the software-based implementations of the present invention and the hardware-based countermeasures 28 to further reduce the EMI generated when using the PWM.

[0028] Software-based implementations are described below in reference to Figures 2-14, according to embodiments of the present invention. Figures 2-11 show PWM waveforms 10, 30, 32 which were generated using a MATLAB script at very low frequencies (around 10 Hertz) doing pulse width modulation (PWM). The PWM waveforms 10, 30, 32, represent variations of PWM digital signals 22 described above for providing electrical power 18 to the motor 12. Typically, electrical circuits used to generate PWM digital signals 22 operate in the order of 10k Hz. However, using a frequency of about 10 Hz minimized the number of data points while clearly indicating the effects of the software-based implementations of the present invention. It will be appreciated that the frequencies shown in the embodiments disclosed below can be scaled up to frequencies in the 1 MHz range and higher without altering the scope of the present invention.

[0029] Figures 2 and 3 illustrate a first embodiment, wherein a first PWM waveform 10 used for switching the PWM is a sinusoidal waveform. Figure 2 shows the sinusoidal PWM waveform 10, which is sinusoidal in the time domain and has a fixed frequency of 10 Hertz. Further, Figure 2 shows i second of data to illustrate the general shape of the sinusoidal PWM waveform 10 before a fast Fourier transform is performed. The simulated data in the Figures was generated using a sample rate two orders of magnitude higher than the Nyquist frequency (2048 samples / second) to minimize aliasing of the data.

[0030] Figure 3 shows a first power spectrum 34 in the frequency domain of a first frequency waveform 36 generated by performing a fast Fourier transform of the first PWM waveform 10 of Figure 2. In Figure 3, the y-axis 38 of the first frequency waveform 36 is in decibels of gain. Theformula for decibels of gain is G[dB] = 10 log, Every change of 10 dB on the y-axis 38 inindicates the output is a factor of 10 greater for positive decibels or a factor or 0.1 smaller fornegative decibels, since the formula for decibels of gain is a logarithmic relationship. In Figure 3, all of the energy in the first frequency waveform 36 is at the fundamental frequency of 10 Hz (shown as a main peak 40) with no energy anywhere else since the first PWM waveform 10 in Figure 2 is sinusoidal and does not have any harmonics. Further, the first power spectrum 34 shown in Figure 3 has a maximum power of about 77.1 dB at the main peak 40.

[0031] Figures 4 and 5 illustrate a second embodiment, wherein the first PWM waveform 10 is dithered to produce a second PWM waveform (not shown). Frequency dithering is a process which selectively adds noise to the first PWM waveform 10, as is commonly known in the art. The first PWM waveform 10 was dithered using a dithering frequency of 0.001 Hz every 50 periods starting at 9.9 Hz and continuing through 10.1 Hz in 0.001 Hz increments. The frequency was changed every 50 periods of the first PWM waveform 10, which gives the second PWM waveform (not shown) time to settle and improve controllability. The frequency range and the frequency increment used when dithering are parameters which can be adjusted as software calibrations.

[0032] Next, the second PWM waveform (not shown) was transformed into the power spectrum using a fast Fourier transform, which generated a second frequency waveform 42 in the frequency domain. Figure 4 shows a second power spectrum 44 of the second frequency waveform 42, which has a primary peak 46. Comparing Figures 3 and 4, the energy that was centered at 10 Hz (main peak 40) in Figure 3 has been slightly spread across the dithering frequencies (9.9 Hz through 10.1 Hz) (primary peak 46) in Figure 4. In addition, the second power spectrum 44 of Figure 4 has a maximum power of about 62.5 dB at the primary peak 46, which is less than the maximum power of about 77.1 dB at the main peak 40 shown in Figure 3. The peaks 40, 46 in the power spectrum 34, 44 are also described as a lobe 40, 46, as commonly known in the art. As such, dithering the first PWM waveform 10 reduced the overall amplitude of the power spectrum 34, 44 and reduced the lobe 40, 46 from about 77.1 dB to about 62.5 dB when the PWM waveform 10 is sinusoidal. Further, the second PWM waveform (not shown) has an attenuation of about l / 29th of the undithered PWM waveform 10 when the PWM waveform 10 is sinusoidal.

[0033] Figure 5 shows an enlarged view of the spectral content 44A of the second frequency waveform 42 of Figure 4 in the frequency range of 9.9 Hz to 10.1 Hz. As shown in Figure 5, dithering the first PWM waveform 10 spreads the frequency out and reduces the lobe 46 in thesecond power spectrum 44. However, MOSFETs are not reliably controlled using the first PWM waveform 10 which is sinusoidal, since this can result in uncertainty in turn-on and turn-off thresholds in each MOSFET in the circuit. Further, using the first PWM waveform 10 which is sinusoidal can also result in conduction losses due to the MOSFETs not being purely switched on or off but would include intermediate states.

[0034] Figures 6 and 7 illustrate a third embodiment, wherein a third PWM waveform 30 is used for switching the PWM. Depicted in Figure 6, the third PWM waveform 30 has a square waveform in the time domain generated using a static frequency of 10 Hertz. Next, the PWM waveform 30 is transformed into the power spectrum using a fast Fourier transform, which generates a third frequency waveform 48 in the frequency domain. Figure 7 shows a third power spectrum 50 of the third frequency waveform 48 generated after performing a fast Fourier transform of the third PWM waveform 30 of Figure 6. Depicted in Figure 7, the spectral content 50A of the third power spectrum 50 in the frequency domain includes a lobe 52, 54, 56 at every odd harmonic (10 Hz, 30 Hz, 50 Hz, etc.), as illustrated by doing a Fourier series expansion of the square wave. The 4amplitude of the lobes 52, 54, 56 is defined to be, Hn= — , wherein n = 1, 3, 5, 7, •••.

[0035] Figure 8 illustrates a fourth embodiment, wherein the third PWM waveform 30 of Figure 6 is dithered to produce a fourth PWM waveform (not shown). The third PWM waveform 30 was dithered using a dithering frequency of 0.001 Hz every 50 periods starting at 9.9 Hz and continuing through 10.1 Hz in 0.001 Hz increments with the frequency changed every 50 periods of the third PWM waveform 30. Next, the fourth PWM waveform (not shown) was transformed into the power spectrum using a fast Fourier transform, which generated a fourth frequency waveform 58 in the frequency domain. Figure 8 shows a fourth power spectrum 60 of the fourth frequency waveform 58 generated from the fourth PWM waveform (not shown). The spectral content 60 A of the fourth frequency waveform 58 includes a lobe 62, 64, 66 at every odd harmonic (10 Hz, 30 Hz, 50 Hz, etc.), as illustrated by doing a Fourier series expansion of the square wave.

[0036] Comparing Figures 7 and 8 illustrates the effect of dithering the third PWM waveform 30 which has a square waveform shape. The lobes 62, 64, 66 of the fourth power spectrum 60 in Figure 8 are reduced in magnitude and the bandwidth increased in comparison to the respective lobes 52, 54, 56 of the third power spectrum 50 shown in Figure 7. The attenuation of theharmonics may be further optimized as necessary by adjusting the software calibrations for dithering of the third PWM waveform 30. A comparison of the attenuation in the frequency domain of the dithered fourth PWM waveform (Figure 8) and the undithered third PWM waveform 30 (Figure 7) is shown in Figure 12.

[0037] Figures 9 and 10 illustrate a fifth embodiment, which is based on the third PWM waveform 30 of Figure 6 having the square waveform at a fixed frequency in the time domain. However, in the fifth embodiment, dithering is not applied to the third PWM waveform 30. Instead, the fundamental switching frequency is unaltered and the transition edges are softened to generate a fifth PWM waveform 32 shown in Figure 9 (i.e., “softened PWM waveform). Applying transition softening to the third PWM waveform 30 adjusts how fast the softened PWM waveform 32 transitions from fully off 70 to fully on 72. The transition softening is based on a calibration parameter which controls the amount of transition softening. In Figure 9, the calibration parameter for the transition softening was set to be 15% of the waveform period. It will be appreciated that the calibration parameter can be adjusted as a software calibration without altering the scope of the present invention. In more detail, Figure 9 shows the softened PWM waveform 32 in the time domain with the transition edge 74 (from the off state 70 to the on state 72) softened by rate limiting this transition. Next, a fifth frequency waveform 76 is generated by performing a fast Fourier transform on the softened PWM waveform 32 of Figure 9. Figure 10 shows a fifth power spectrum 78 of the fifth frequency waveform 76. The fifth power spectrum 78 includes lobes 80, 82, 84 in the spectral content 78A, which correspond to the first, third, and fifth harmonic.

[0038] The effect of transition softening on the fifth power spectrum 78 is shown by comparing the third power spectrum 50 of the unsoftened third PWM waveform 30 in Figure 7 and the fifth power spectrum 78 in Figure 10. Referring to Figures 7 and 10, applying transition softening to the third PWM waveform 30 reduces the lobes 80, 82, 84 of the harmonics in the spectral content 78A in the frequency spectrum in comparison to the lobes 52, 54, 56 of the harmonics in the spectral content 50A of the unsoftened PWM waveform 30. Figure 13 shows the impact of transition softening of the third PWM waveform30 has in the frequency spectrum. The transition softening of the third PWM waveform 30 causes the attenuation effect to continue to grow and spread as the harmonics increase.

[0039] Figure 11 illustrates a sixth embodiment which is based on the third PWM waveform 30 of Figure 6 having the square waveform. In the sixth embodiment, dithering and transition softening are applied to the third PWM waveform 30 to generate a sixth PWM waveform (not shown). The third PWM waveform 30 was dithered using a dithering frequency of 0.001 Hz every 50 periods starting at 9.9 Hz and continuing through 10.1 Hz in 0.001 Hz increments with the frequency changed every 50 periods of the third PWM waveform 30. In addition, transition softening was applied to the third PWM waveform 30 using a calibration parameter of 15.00% of the waveform period. Next, a fast Fourier transform is performed on the sixth PWM waveform (not shown) to generate a sixth frequency waveform 86.

[0040] Figure 11 shows a sixth power spectrum 88 of the sixth frequency waveform 86.). The spectral content 88A of the sixth frequency waveform 86 includes lobes 90, 92, 94, which correspond to the first, third, and fifth harmonic. Figure 13 shows a table comparing the spectral content 88A of Figure 11 of the dithered and softened sixth PWM waveform (not shown) with the spectral content 60A of Figure 8 of the dithered fourth PWM waveform (not shown). The effectiveness of the transition softening is additive to the effectiveness of the dithering when comparing the effect of frequency dithering and transition softening in the spectral content.

[0041] As described above, dithering the frequency of the PWM waveform 10, 30, 32 attenuates the harmonics in the frequency domain, which can reduce the amount of generated electromagnetic interference (EMI). In addition, transition softening of the PWM waveform 10, 30, 32 also attenuates the harmonics in the frequency domain, which can also reduce the amount of generated EMI. Further, frequency dithering and transition softening can be applied to the PWM waveform 10, 30, 32 to further reduce the amount of generated EMI. The controller 14 includes softwarebased implementations comprising frequency dithering and / or transition softening the PWM waveform 10, 30, 32 prior to providing the PWM waveform 10, 30, 32 to the electric motor 12, which in turn reduces the amount of EMI generated by use of the PWM waveform 10, 30, 32 and reduces the amount of electromagnetic noise radiated by the electric motor 12. Further, the frequency dithering and / or transition softening applied by the controller 14 can be combined with conventional hardware-based countermeasures to further reduce EMI and radiated electromagnetic noise.

[0042] The invention has been described in an illustrative manner, and it is to be understood that the terminology, which has been used, is intended to be in the nature of words of description rather than of limitation. Many modifications and variations of the present invention are possible in light of the above teachings. It is, therefore, to be understood that within the scope of the appended claims, the invention may be practiced other than as specifically described.

Claims

What is claimed is:

1. A method of generating a pulse width modulated (PWM) waveform for providing power to an electric motor for use in a vehicle, the method comprising:generating a PWM waveform;applying frequency dithering to the PWM waveform; andproviding the PWM waveform to the electric motor.

2. The method as set forth in claim 1, wherein:the frequency dithering is applied to the PWM waveform using dithering frequency of 0.001 Hz every 50 periods of the PWM waveform.

3. The method as set forth in claim 1 or claim 2, wherein:the frequency dithering is applied to the PWM waveform between 9.9 Hz through 10.1 Hz in 0.0001 Hz increments.

4. The method as set forth in any one of claims 1 to 3, wherein:the frequency is changed every 50 periods of the PWM waveform.

5. The method as set forth in any one of claims 1 to 4, further comprising:applying transition softening to the PWM waveform.

6. The method as set forth in claim 5, wherein:the PWM waveform includes a waveform period; andan amount of transition softening applied to the PWM waveform is 15% of the waveform period.

7. The method as set forth in any one of claims 1 to 6, wherein:the PWM waveform is square waveform.

8. The method as set forth in any one of claims 1 to 5 wherein:the PWM waveform is a sinusoidal waveform.

9. A method of generating a pulse width modulated (PWM) waveform for providing power to an electric motor for use in a vehicle, the method comprising:generating a PWM waveform;applying transition softening to the PWM waveform; andproviding the PWM waveform to the electric motor.

10. The method as set forth in claim 9, wherein:the PWM waveform includes a waveform period; andan amount of transition softening applied to the PWM waveform is 15% of the waveform period.

11. The method as set forth in claim 9 or claim 10; further comprising:applying frequency dithering to the PWM waveform.

12. The method as set forth in claim 11, wherein:the frequency dithering is applied to the PWM waveform using dithering frequency of 0.001 Hz every 50 periods of the PWM waveform.

13. The method as set forth in claim 11 or claim 12, wherein:the frequency dithering is applied to the PWM waveform between 9.9 Hz through 10.1 Hz in 0.0001 Hz increments.

14. The method as set forth in any one of claims 11 to 13, wherein:the frequency is changed every 50 periods of the PWM waveform.

15. The method as set forth in any one of claims 9 to 14, wherein:the PWM waveform is square waveform.

Images