TransLogistics

Transport and Logistics Network

High-performance photon-driven DC motor system

Photon-driven DC motor system structure

The proposed photon-driven DC motor system is shown in Fig. 1b. The DC motor speed is regulated by the photonic converter, which is controlled by the controller via a HPLD driver. The optical fiber communication is used to transmit the feedback signals from sensors, such as motor speed, to the controller.

In the proposed photon-driven DC motor system, the photonic converter is mainly comprised of HPLD, fiber, and PPC. The HPLD is utilized to continuously supply optical power to the PPC via a fiber. The PPC is utilized to convert the optical power into suitable electrical power for driving DC motor.

Supplementary Fig. 1 provides a comparison of the energy conversion process between the proposed system and the conventional system.

Theoretical analysis of HPLD-PPC-Motor model

A HPLD-PPC-Motor model based on the theoretical equivalent circuit model, as shown in Fig. 3a and b, is proposed to derive the theoretical performance of the proposed system. This model involves the photoelectric characteristics of the HPLD and PPC, as well as the electrical and mechanical characteristics of the DC motor. The HPLD model is the electro-optic equivalent circuit35,36, and the HPLD requires to operate in its linear interval (the gray area in Supplementary Fig. 2). In this linear interval, the injection current of the HPLD IHPLD is linear to the output optical power Plight_HPLD of the HPLD, i.e., Plight_HPLD ? IHPLD. Thus, the Plight_HPLD can be modulated by the IHPLD (see section "Method" for the details).

Fig.

3: Schematic of the operation principle of the photonic converter. figure 3

a The equivalent circuit of the HPLD-PPC-Motor model. The output current (Ippc) and output voltage (Vppc) of the PPC are equal to the armature current (Ia) and the terminal voltage (Vppc) of the brushed DC motor, respectively. Iph denotes photogenerated current of the PPC, D1 denotes p-n junction diode, Rsh denotes shunt resistance, Rs denotes series resistance, C denotes shunt capacitance, Ic denotes capacitance current, Ra denotes armature winding resistance, La denotes armature winding inductance, e denotes back electromotive force (EMF), Te and TL denote the electromagnetic torque and load torque, respectively. ?rm is the rotating angular speed of the rotor, J is the inertia coefficient, and B is the viscous damping coefficient. MPP, maximum power point. b The relationship between different models within the HPLD-PPC motor model. c Current-voltage (Ippc-Vppc) curves for the PPC at the input optical power of 1.5, 3, 4.5, and 6 W with current-voltage (Ia-Vppc) curves for the brushed DC motor at load torque of 20, 30, and 40 mNm.

OMP operational matching point. d Power-voltage (Pppc-Vppc) curves for the PPC at the optical power of 1.5, 3, 4.5, and 6 W with power-voltage (Pa-Vppc) curves for the brushed DC motor at the load torque of 20, 30, and 40 mNm. e Photocurrent-input optical power (Iph-Plight_PPC) curve for the PPC. f Photocurrent-injection current (Iph-IHPLD) curves for the photonic converter under fiber transmission distance from 10 m to 10 km. The used parameters are listed in Supplementary Table 1 and Supplementary Table 2.

Full size image[1]

The fiber model shows the output optical power Plight_HPLD of the HPLD is linear to the input optical power Plight_PPC of the PPC, i.e., Plight_PPC ? Plight_HPLD in the optical energy transmission from HPLD to PPC via the fiber. The model of the fiber is based on the Bouguer-Beer-Lambert law37,38, which shows that the power of light in optical energy transmission by the fiber is exponential attenuation over distance. This attenuation, caused primarily by absorption and scattering in the fiber, is quantified as attenuation coefficient ? in decibels per kilometer (dB/km). The ? depends on two factors, including the distance of light propagation and the light's wavelength.

For a wavelength of 808 nm, the ? is around 2.0 dB/km. For the wavelength of 975 nm, the ? is around 0.8 dB/km39. The ? for SiO2 fiber is shown in Supplementary Fig. 3 under different wavelengths and distances from 0 m to 1 km. For a specific fiber, the distance of light propagation and wavelength are constant, that is, the ? for this fiber is also constant.

Therefore, Plight_PPC ? Plight_HPLD (see section "Method" for the details). Moreover, HPLD and PPC efficiencies vary by wavelength: 808 nm HPLDs have an over 45% efficiency25,40, while 975 nm reach over 50%41. PPCs operate at over 55% efficiency42 for 800-850 nm and over 24%20,43 for 900-980 nm. This study selects the 808 nm wavelength for high PPC efficiency. The PPC model is the single diode equivalent circuit44,45(see section "Method" for details), which consists of a series resistor, a parallel resistor, a diode, and a current source, as shown in Fig. 3a. The PPC receives input optical power Plight_PPC, and generates photogenerated current Iph of the PPC, which determines the output current Ippc of the PPC and the output voltage Vppc through the load characteristics. Figure 3c, d show the current-voltage (Ippc-Vppc) and power-voltage (Pppc-Vppc) curves of the PPC employed in this study with Plight_PPC from 1.5 W to 6 W.

Figure 3e shows the curves between the Iph and the Plight_PPC, where Iph ? Plight_PPC. Here, Pppc = Vppc.Ippc. In Fig. 3c, d, the maximum power points (MPP) are the maximum electrical power output corresponding to various Plight_PPC, which are located at the "knee" regions of the Ippc - Vppc curve.

To the left of the MPP, the Ippc is approximately equal to the Iph, i.e., Ippc ? Iph. To the right of the MPP, Ippc drops sharply as the Vppc increases. The brushed DC motor model1 is shown in Fig. 3a, which is composed of armature winding inductance La, armature winding resistance Ra, and the back electromotive force (EMF) e (see section "Method" for details). The DC motor receives armature current Ia, and generates electromagnetic torque Te.

The motor speed ?rm is determined when the Te balances with the load torque TL. An increase in the Ia results in greater Te, which leads to an increase in ?rm under the same TL. Figure 3c, d show the Ia-Vppc curves and the Pa-Vppc curves of the DC motor used in this work, respectively, under different TL from 20-40 mNm, where the input power Pa = Vppc.Ia.

The HPLD-PPC-Motor model of the proposed photon-driven DC motor system can be obtained by combing models of HPLD, fiber, PPC and DC motor. The IHPLD determines the Plight_HPLD, Plight_PPC, and Iph (Fig. 3f shows the Iph-IHPLD curves with fiber distances from 0.01 km to 10 km), which determines the Ippc. Ic is capacitor current and Ippc = Ia + Ic. Given the TL, the Ia determines the Te, which further determines the ?rm.

Therefore, the motor speed ?rm as a function of IHPLD and TL can be written as ££{\omega }_{{rm}}= \frac{{K}_{t}}{B}\left[{R}_{{HPLD}}{\eta }_{{trans}}{R}_{{PPC}}\left({I}_{{HPLD}}-{I}_{{th}}\right)\right]-\frac{{K}_{t}{I}_{o}}{B}\left[\exp \left(\frac{{A}_{1}}{n{V}_{T}}\right)-1\right] \\ -\frac{{K}_{t}{A}_{1}}{B{R}_{{sh}}}-\frac{{K}_{t}{I}_{c}}{B}-\frac{{T}_{L}}{B}-\frac{J}{B}\frac{d{\omega }_{{rm}}}{{dt}}££ (1)

with ££{A}_{1}= \frac{B{\omega }_{{rm}}+{T}_{L}+J\frac{d{\omega }_{{rm}}}{{dt}}}{{K}_{t}}{R}_{a}+{K}_{e}{\omega }_{{rm}}+\left(\frac{B{\omega }_{{rm}}+{T}_{L}+J\frac{d{\omega }_{{rm}}}{{dt}}}{{K}_{t}}+{I}_{c}\right){R}_{s} \\ +{L}_{a}\frac{d\left(B{\omega }_{{rm}}+{T}_{L}+J\frac{d{\omega }_{{rm}}}{{dt}}\right)}{{dt}}££ (2)

where Kt is torque constant, B is viscous damping coefficient, RHPLD is differential slope efficiency, ?trans is power efficiency, RPPC is differential responsivity, Ith is threshold current, Io is diode reverse saturation current, n is diode ideality constant, VT is junction thermal voltage, Rsh is shunt resistance of the PPC, and Rs is series resistance of the PPC, J is inertia coefficient (see Method for the function derivation). In the HPLD-PPC-Motor model as shown in Fig. 4a, given the load torque TL, the ?rm can be controlled by the Plight_PPC, which is regulated by the IHPLD. Figure 4a shows the ?rm as a function of IHPLD and TL based on the PPC and DC motor parameters used in this work, where a distinct red line highlights a startup condition of the motor.

Figure 4b shows the Vppc as a function of IHPLD and TL. Figure 4c shows the Ippc as a function of IHPLD and TL. The same trend between ?rm and Vppc can be observed within Fig. 4a, b.

Here, Vppc cannot surpass the open-circuit voltage of the PPC and the Ippc cannot exceed Iph, as shown in Fig. 4b, c, thereby limiting the increase of Vppc and ?rm. Supplementary Fig. 4 presents a comparison between the theoretical model and experimental data, including ?rm, Vppc, and Ippc. It is worth mentioning that the photon-driven DC motor system has an anti-disturbance capability to the fluctuations of load torque TL, which is not available in switching electricity converter-driven DC motor system (see Supplementary Note 1 for the principle of anti-disturbance capability).

Fig.

4: Theoretical model performance of the proposed photonic converter. figure 4

a The speed-load torque-current of the HPLD(?rm-TL-IHPLD) relationship of the photonic converter for DC motor. The depicted red line signifies the startup conditions, where the electromagnetic torque surpasses the TL, thereby the motor starts rotating. b The voltage of the PPC-load torque-current of the HPLD (Vppc-TL-IHPLD) relationship of the photonic converter for DC motor. c The current of the PPC-load torque-current of the HPLD (Ippc-TL-IHPLD) relationship of the photonic converter for DC motor. The used parameters are listed in Supplementary Table 1 and Supplementary Table 2.

Full size image[2]

Operation principle of photon-driven DC motor

The operation principle of the proposed photon-driven DC motor is proposed based on aforementioned HPLD-PPC-Motor model, where the photon-driven DC motor works at the corresponding operational matching point (OMP) in the steady state under different Plight_PPC (i.e., different IHPLD) and different TL, as shown in Fig. 3c, d.

The OMPs are the intersections (red triangles) of the PPC's Ippc-Vppc curves and the DC motor's Ia-Vppc curves, where there is Ippc = Ia in the OMP in the steady state. For an example, suppose that the photon-driven DC motor works at OMP A (Plight_PPC = 3 W, i.e., IHPLD ?

1.48 A under 10 m fiber, and TL = 20 mNm), an increase of the Plight_PPC to 4.5 W, i.e., IHPLD ?

2.01 A under 10 m fiber, would lead to change in the PPC's voltage and current curves (approximate curve shifts upward), which leads to a rise in current (Ippc and Ia), voltage Vppc, and speed ?rm, and finally works at OMP B steadily. Figure 3c shows that as the TL increases, to remain same ?rm, the Plight_PPC needs to be increased.

If the TL remains unchanged, to increase the ?rm, the Plight_PPC needs to be increased.

Power modulation-based speed control

An innovative power modulation (PM)-based speed control for the proposed photon-driven DC motor system is proposed based on aforementioned HPLD-PPC-Motor model and proposed operation principle, as shown in Fig. 5a. The PM-based speed control is composed of the controller, the photonic converter, optical fiber communication, and a suite of sensors. In the PM-based speed control, the control structure includes the system model (Fig. 5b) and the control scheme (Fig. 5c).

The proportional-integral (PI) controller is used in this paper, which is renowned for efficacy in error correction and system stability enhancement. The operational flow begins with acquiring real-time feedback signals, including the ?rm and Vppc obtained from these sensors. These signals are then sent to the controller via optical fiber communication, marking the initiation of the speed control process.

The control scheme is achieved by computing the speed error, which is the difference between the speed reference ?ref and the ?rm. Based on the speed error, the PI controller control the reference of IHPLD, which is subsequently sent to the HPLD driver, and then control the HPLD's output optical power Plight_HPLD and the ?rm. The control design for the proposed system is shown in the Supplementary Note 2.

Since specific scenarios require the motor to have forward and reverse rotation functions, a topology that can realize forward and reverse rotation functions based on the photonic converter is proposed to meet this requirement, see Supplementary Note 3 for details.

Fig.

5: Schematic of the proposed PM-based speed control for the photon-driven DC motor system. figure 5

a The control structure of the PM-based speed control in the photonic converter and experimental setup. b The system model of the photonic converter. c Control scheme of the proposed system. The speed reference is the input variable and the injection current reference of the HPLD is the output variable.

Full size image[3]

Experimental verification of PM-based speed control

A prototype was constructed to verify the PM-based speed control of the photon-driven DC motor system, and its performance was tested using commercially available off-the-shelf components including an HPLD driver, HPLD, PPC, and a brushed DC motor. The experimental setup is shown in Fig. 5a, and its photos are shown in Fig. 6a (see section "Methods" for the construction of photonic converter prototype, Supplementary Fig. 5 for experimental results in the oscilloscope display diagram, and Supplementary Fig. 6 for the relationship between the brake injection current Itorque and output load torque TL).

To prove that the proposed system can achieve speed control, three test situations are designed to verify the efficiency and feasibility of the motor system. The first test situation is that the DC motor brakes and starts under constant load torque. The second test situation is that the DC motor follows a sinusoidally varying speed reference command with constant load torque.

The third test situation is that the DC motor tracks the constant speed reference command with load disturbance.

Fig.

6: Experimental results of speed control for the photon-driven DC motor system. figure 6

a Photograph of the experimental setup. b,c The system's performance during both halt and initiation phases with rejection of ripple. Initially, the brushed DC motor maintains a constant speed of 347 rpm under a load torque of 40 mNm. At about 6 s, the speed command is adjusted to 0 rpm, and then back to 347 rpm at about 14 s; global picture (b) and local picture (c).

Full size image[4]

Test situation 1: braking and starting under constant load

Figure 6b, c shows the experimental result of the first test situation, where the dynamic response of a DC motor under variable control commands is observed.

Initially, the motor maintains a steady operation at 347 rpm under a load torque of 40 mNm. During this time, the PPC current ranges from 370 mA to 800 mA, and the voltage ranges from 5 V to 6.1 V, and the HPLD current ranges from 2.37 A to 2.6 A. At around 6 seconds, a command is issued to reset the speed reference to 0 rpm.

This causes a rapid drop in both current and voltage, with the current falling to about 460 mA and the voltage dropping to around 0.65 V within 2 seconds, leading to the motor coming to a stop. At around 14 s, a command is given to restore the speed reference to 347 rpm, prompting the motor to restart. The current quickly rises to around 630 mA, fluctuates between 570 mA and 680 mA, and then stabilizes in the range of 465-760 mA.

Similarly, the voltage eventually stabilizes around 6 V. The motor takes about 2 seconds to reach the desired speed after starting up. See Supplementary Movie 1 for the details.

It is worth mentioning that both current and voltage exhibit minimal ripple when the motor stops, demonstrating the rejection of output ripple in the photonic converter. The current and voltage ripples observed when the motor rotates, are caused by the commutation of the DC motor's commutator4.

Test situation 2: following a sinusoidal speed reference with constant load

Figure 7a, b presents the experimental results for the second test situation, focusing on the response to a sinusoidally varying speed reference command. In this experiment, the speed reference command exhibits a sinusoidal waveform with a frequency of 0.05 Hz and an amplitude of 30 rpm, which is superimposed on the constant speed of 347 rpm, under a load torque of 40 mNm. A more detailed view of the sinusoidal variation is provided in Fig. 7a, b, showing one cycle of the waveform.

Within one cycle, the range of PPCs current changes from a range of 0.395-0.760 A to a range of 0.440-0.760 A. The voltage demonstrates a variation range from a range of 4.6-5.6 V to a range of 5.4-6.4 V. The experimental results indicate that the speed control system effectively tracks the sinusoidally varying speed reference command.

See Supplementary Movie 2 for the details.

Fig.

7: Experimental results of speed control for the photon-driven DC motor system. figure 7

The system's capability to track a sinusoidal speed setpoint with a 0.05 Hz frequency under a constant load torque of 40 mNm; global picture (a) and local picture (b). The system's response to variations in load torque; starting from a base speed of 347 rpm with a 40 mNm load, the torque is reduced to 0 mNm at about 15 s, and subsequently restored to 40 mNm at about 35 s; global picture (c) and local picture (d).

Full size image[5]

Test situation 3: tracking constant speed with load disturbance

Figure 7c, d presents the experimental results for the third test condition, which explores the response of the proposed system under varying load torque conditions. Initially set at 347 rpm with a 40 mNm load, the motor experiences a sudden shift to no-load conditions at 15 s.

The abrupt change causes an excess of electromagnetic torque, which leads to an unintended speed increase. To counteract the situation, the system dynamically adjusts the PPC current, reducing electromagnetic torque and moderating the current from a range of 0.38-0.8 A down to a range of 0.03-0.32 A. This action mitigates the speed increase, stabilizing it back to 347 rpm at about 22 s.

The reapplication of the 40 mNm load torque at about 35 s further tests the system's resilience. This increase in load torque causes an immediate rise in current and a drop in voltage, reflecting the motor's struggle against the increase in load torque. Then, the speed control promptly elevates the PPC current, boosting electromagnetic torque to counter the impact of increasing load torque.

The voltage, after initially falling, starts to climb, showing a recovery in motor speed. By about 40 s, motor operation normalizes, returning the initial parameters with a voltage range of 4.8-6 V and maintaining the set speed of 347 rpm. This experiment shows that the speed control is effective in keeping the motor speed steady, even when the load torque changes.

It demonstrates the system's robustness in real-time adjustments. See Supplementary Movie 3 for the details.

Experimental verification of speed limit and output ripple

Figure 8a illustrates an experiment conducted to confirm the presence of a speed limit as shown in the model. The motor, operating under open-loop control with a 20 mNm load torque, shows variations in speed, voltage, and current with increasing HPLD injection current from 0 A to 2 A, starting at 6.3 s and continuing until 25.6 s.

Initially, the low PPC voltage indicated the motor is stationary. Once rotation starts, the PPC voltage surges. At about 22.4 s, the motor's speed stabilizes at approximately 500 rpm, and the PPC voltage stabilizes between the range from 6.77 V to 7.09 V, slightly below the PPC's open-circuit voltage of 7.2 V, despite further increases in HPLD current.

This experimental result confirms the speed limit of the above discussion, demonstrating the motor's speed stabilization despite increasing HPLD current.

Fig.

8: Model validation experiment and comparison of output ripple between switching electricity converter and photonic converter. figure 8

a Under a 20 mNm load torque, motor speed, voltage, and current responses to increasing HPLD injection current from 0 A to 2 A starting at 6.3 s and continuing until 25.6 s, over 19.3 s. Notably, at 22.4 s, the motor reached its speed limit at ~500 rpm, while PPC voltage peaked between 6.77 V and 7.09 V, demonstrating a speed limit. b Output ripple of the switching electricity converter with a load resistance of 20 ?. c Output ripple of the photonic converter with the load resistance of 20 ?.

Full size image[6]

To compare the output ripple between the photonic converter and the switching electricity converter, experimental comparisons were conducted. Both converters carry a load resistance of 20 ? and have an output voltage of 5 V and an output current of 0.25 A.

Figure 8b, c show the output voltage and output current of the switching electricity converter and the output voltage and output current of the photonic converter, respectively. The output ripple in the switching electricity converter can be observed due to its on-off switching actions. In the photonic converter, the output ripple is almost non-existent, and the voltage and current are almost a straight line.

Experimental verification of intrinsic EMI of proposed system

The intrinsic EMI of the proposed system and conventional system is measured by assessing the intensity of electromagnetic field emission from both.

A YOKOGAWA DLM5038 oscilloscope (2.5 GS/s) is utilized for monitoring voltage and current variations, and an H field probe is used to monitor the intensity of electromagnetic field emission. Figure 9 shows the EMI measurement experimental results of photon-driven DC motor system with load torque (40 mNm) and sinusoidal speed reference (superimpose a sine speed command with an amplitude of 30 rpm on a speed of 347 rpm). Figure 9a shows the close-up photo of setup.

The H field probe is positioned approximately 2 cm directly above the photonic converter, as shown in Fig. 9a. Figure 9b, c show the oscilloscope wave of experiment results, including the HPLD voltage, HPLD current, PPCs voltage, PPCs current, speed, reference speed, torque-current, and the output voltage of the H field probe. Figure 9d shows the voltage ripple measured from the H field probe.

Based on Fig. 9d, Fig. 9e shows the intensity of the intrinsic EMI. In Fig. 9e, the red line represents the intensity of electromagnetic field emissions when the proposed system is powered off, and the blue line represents the intensity of electromagnetic field emissions when the proposed system is powered on. From Fig. 9e, it can be observed that the blue line is very close to the red line with little difference, which indicates that the proposed system emits little electromagnetic field emissions.

Fig.

9: EMI measurement experiment results of proposed system with motor under load torque and sinusoidal speed reference. figure 9

a Close-up photo of the photonic converter and motor under test. Oscilloscope waveform of experiment results; global picture (b) and local picture (c). d Output voltage waveforms from the H field probe. e Intensity of the intrinsic EMI, measured when proposed system is powered on (in blue) and off (in red).

Full size image[7]

Figure 10 shows the EMI measurement experimental results of conventional switching electricity converter-driven DC motor system with load torque (40 mNm) and sinusoidal speed reference (superimpose a sine speed command with an amplitude of 30 rpm on a speed of 347 rpm). Figure 10a[8] shows the close-up photo of setup.

The H field probe is positioned ~2 cm directly above the switching power converter, as shown in Fig. 10a[9]. Figure 10b[10], c show the oscilloscope wave of experiment results, including motor voltage, motor current, speed, reference speed, torque current, and the output voltage of the H field probe. Based on Fig. 10d[11], e shows the intensity of the intrinsic EMI.

In Fig. 10e[12], the red line represents the intensity of electromagnetic field emissions when the conventional system is powered off, and the blue line represents the intensity of electromagnetic field emissions when the conventional system is powered on. It can be seen that the blue line is obviously greater than the red line, which indicates that conventional systems emit much electromagnetic field emissions. From Fig. 9e, 10e[13], it can be observed that the proposed system reduces EMI compared to the conventional system.

Experimental verification of intrinsic EMI of photonic converter is shown in the Supplementary Note 4.

See Supplementary Movie 4 for the details.

Fig.

10: EMI measurement experiment results of conventional system with motor under load torque and sinusoidal speed reference. figure 10

a Close-up photo of the switching power converter and motor under test.

Oscilloscope waveform of experiment results; global picture (b) and local picture (c). d Output voltage waveforms from the H field probe. e Intensity of the intrinsic EMI, measured when conventional system is powered on (in blue) and off (in red).

Full size image[14]

References

  1. ^ Full size image (www.nature.com)
  2. ^ Full size image (www.nature.com)
  3. ^ Full size image (www.nature.com)
  4. ^ Full size image (www.nature.com)
  5. ^ Full size image (www.nature.com)
  6. ^ Full size image (www.nature.com)
  7. ^ Full size image (www.nature.com)
  8. ^ 10a (www.nature.com)
  9. ^ 10a (www.nature.com)
  10. ^ 10b (www.nature.com)
  11. ^ 10d (www.nature.com)
  12. ^ 10e (www.nature.com)
  13. ^ 10e (www.nature.com)
  14. ^ Full size image (www.nature.com)