High-performance photon-driven DC motor system

Brushed DC motors are widely used due to their simplicity, availability, easy control, and reliability, but conventional drive systems rely on switching electricity converters whose rapid on–off actions generate high-frequency electromagnetic emissions and output voltage ripple. These effects induce torque ripple, complicate control, and make systems susceptible to environmental EMI that corrupts feedback and control signals. Power over fiber (PoF) offers EMI immunity because optical fibers are non-conductive, also providing galvanic isolation, weight reduction, and eliminating fire or shock risks compared with metal wires. This work proposes a photon-driven DC motor system that integrates PoF and DC motor control. A photonic converter converts electrical power to optical via a high-power laser diode (HPLD), transmits it through fiber, and converts it back to electrical via photovoltaic power converters (PPCs) to drive the motor. Because optical power is continuously controlled (rather than switched), the approach eliminates intrinsic EMI and output ripple from switching converters and is immune to environmental EMI. The paper develops a HPLD–fiber–PPC–motor model, derives the operation principle, and proposes a power-modulation-based speed control to address incompatibility with conventional switching-based control.

The study builds on Power over Fiber (PoF) technologies that transmit optical energy over fiber for EMI-immune power delivery and communication, with demonstrated power levels up to tens to hundreds of watts using appropriate fibers and wavelengths. Prior work on photonic power conversion has explored high-efficiency HPLDs (e.g., >45% at 808 nm and >50% at 975 nm) and PPCs (e.g., >55% efficiency in 800–850 nm bands; lower at 900–980 nm), as well as ripple-free photonic power supplies. Photovoltaic laser power converters with advanced vertical epitaxial structures have achieved high photovoltage and efficiency. EMI issues in conventional PWM motor drives and modeling/mitigation techniques are well documented, motivating EMI-immune alternatives. This paper selects 808 nm to leverage higher PPC efficiency and integrates these elements specifically for DC motor drive and control with closed-loop speed regulation over optical links.

System architecture: A photonic converter drives a brushed DC motor. A HPLD converts electrical to optical power; optical power is transmitted via multimode fiber (MMF 105/125 µm, NA 0.22) and a 1×2 50:50 fiber coupler to PPCs that reconvert to DC to supply the motor and parallel capacitors (total 660 µF). Feedback (speed, voltage, current) is sent to the controller via optical fiber communication for EMI-immune sensing. A PI controller implements power-modulation (PM)-based speed control by regulating HPLD injection current to track speed references. Modeling: A unified HPLD–fiber–PPC–motor model is developed. (1) HPLD electro-optic equivalent model: in the linear region, output optical power is proportional to injection current above threshold (P_light,HPLD = R_HPLD·(I_HPLD−I_th)). (2) Fiber model: Bouguer–Beer–Lambert law describes attenuation (α in dB/km; ~2.0 dB/km at 808 nm; ~0.8 dB/km at 975 nm), giving transmission efficiency η_trans and P_light,PPC = η_trans·P_light,HPLD. (3) PPC single-diode model with series/shunt resistances and junction diode; photocurrent I_ph = R_PPC·P_light,PPC. PPC I–V and P–V characteristics and maximum power point (MPP) are characterized; parameter extraction uses I_sc, V_oc, slopes near short/open-circuit, and MPP (Broadcom AFBR-POC306A1). (4) Brushed DC motor electrical and mechanical model: V_ppc = R_a I_a + L_a dI_a/dt + K_e ω_rm; T_e = K_t I_a; J dω_rm/dt + B ω_rm + T_L = T_e. Operation principle: In steady state, the operating matching point (OMP) is the intersection of PPC I–V and motor I–V curves for a given optical power and load torque. Increasing optical power (via I_HPLD) shifts PPC curves, increasing V_ppc, I_ppc (≈ I_a), torque, and speed; higher load torque requires higher optical power to maintain speed. The model indicates speed, PPC voltage/current surfaces versus I_HPLD and T_L and predicts a speed limit tied to PPC V_oc and I_ph. Control: Power-modulation-based speed control uses a PI controller to regulate I_HPLD from the speed error (ω_ref − ω_rm); the inverse system mapping ensures appropriate HPLD current commands. Feedback includes ω_rm and V_ppc over fiber communication. A topology for forward/reverse rotation with photonic conversion is provided (Supplementary Note 3). Experimental setup: Off-the-shelf components include HPLD driver (Yexian LDM1101), DC source (Itech IT60060-500-40), HPLD (BOX BLD-F808-15-22ST, 808 nm, up to 15 W), 1×2 MMF coupler (MC Fiber Optics MMFBTC026, 50:50), PPCs (Broadcom AFBR-POC306A1, max optical input 6 W, rated electrical output 3 W), parallel ceramic capacitors (660 µF), brushed DC motor (stator bipolar, 3-slot rotor) with 1:13 gearbox, torque brake (HAIBOHUA HB-02M-0.2Nm) controlled by HAIBOHUA CSM-500, current transducer (LEM CASR 6-NP), encoder in torque meter for speed, DSP controller with RS485 and ADC channels, oscilloscope YOKOGAWA DLM5038. EMI measurements used an H-field probe positioned ~2 cm above the device under test and recorded with the oscilloscope. Test protocols: Three closed-loop tests validate speed control: (1) braking and starting under constant load torque (40 mNm), (2) tracking a sinusoidal speed reference (0.05 Hz, ±30 rpm around 347 rpm) at constant load (40 mNm), (3) tracking constant speed (347 rpm) with load disturbance steps (40 mNm → 0 mNm → 40 mNm). Additional open-loop sweep of I_HPLD validates predicted speed limit. A comparative ripple test contrasts a switching electricity converter vs photonic converter at 5 V, 0.25 A across 20 Ω, and EMI emission spectra are compared for both systems under similar load and reference commands.

• Speed control accuracy and dynamics: Under 40 mNm load, the motor tracks commands, stopping from 347 rpm and restarting to 347 rpm in ~2 s. During stop, both PPC current and voltage exhibit minimal ripple (near ripple-free output). During rotation, observed ripples are attributed to commutation of the brushed motor.
• Sinusoidal tracking: With ω_ref = 347 rpm ± 30 rpm at 0.05 Hz and 40 mNm load, the system tracks the sinusoid. Within cycles, PPC current varies approximately from 0.395–0.760 A to 0.440–0.760 A; PPC voltage varies from about 4.6–5.6 V to 5.4–6.4 V, reflecting modulation to follow speed.
• Load disturbance rejection: For a step in load torque from 40 mNm to 0 mNm at ~15 s (base speed 347 rpm), the controller reduces PPC current from ~0.38–0.8 A to ~0.03–0.32 A to counter overspeed, restoring 347 rpm by ~22 s. Reapplying 40 mNm at ~35 s causes current rise and voltage drop; the controller increases PPC current, recovers voltage, and returns to 347 rpm by ~40 s, demonstrating robustness.
• Speed limit verification: In open-loop with 20 mNm load and I_HPLD ramped 0→2 A (6.3–25.6 s), motor speed saturates at ~500 rpm around 22.4 s; PPC voltage stabilizes at 6.77–7.09 V, slightly below PPC V_oc ~7.2 V, confirming model-predicted speed/voltage limits.
• Output ripple comparison: At 5 V, 0.25 A with 20 Ω load, the switching electricity converter exhibits observable voltage/current ripple due to switching, while the photonic converter’s output is nearly ripple-free (voltage and current nearly straight lines).
• EMI emissions: With the photonic converter and motor under load (40 mNm) and sinusoidal speed reference (347 rpm ± 30 rpm), the intrinsic EMI measured by an H-field probe shows the powered-on emission spectrum closely matches the powered-off baseline, indicating negligible additional emissions. In contrast, the conventional switching-converter-driven system shows a powered-on spectrum clearly above the powered-off baseline, evidencing significant EMI. Overall, the proposed system markedly reduces intrinsic EMI compared to the conventional system.
• Modeling validation: Measured speed, PPC voltage and current agree with the HPLD–fiber–PPC–motor model trends (Supplementary Fig. 4), including OMP behavior, dependence on load torque, and saturation near PPC V_oc/I_ph limits.

The findings demonstrate that replacing the switching electricity converter with a continuously controlled photonic converter eliminates intrinsic EMI and significantly suppresses output ripple while preserving closed-loop speed control performance. The PM-based control effectively regulates motor speed through optical power, handling command changes and load disturbances. The verified speed limit aligns with the PPC open-circuit voltage and photocurrent constraints, informing design margins for desired speed/torque envelopes. EMI measurements confirm that the non-conductive optical power path and absence of high-frequency switching actions yield emissions at environmental levels, addressing a key barrier in EMI-sensitive environments. The approach is particularly promising for applications like tethered UAVs, where EMI immunity and reduced tether weight (optical fiber) can extend operational range and reliability, and auxiliary motor drives in electromagnetic catapult systems. The integrated model provides a basis for controller design and system sizing, capturing couplings among HPLD drive, fiber losses, PPC characteristics, and motor electromechanics.

This work introduces a photon-driven DC motor system that integrates an HPLD, optical fiber transmission, and PPCs to drive a brushed DC motor with continuous photonic power control. Contributions include: (1) development and validation of a unified HPLD–fiber–PPC–motor model, (2) an operation principle centered on operating matching points between PPC and motor characteristics, (3) a power-modulation-based PI speed control architecture over optical communication, and (4) experimental verification showing accurate speed tracking, strong disturbance rejection, near-zero output ripple, and drastically reduced intrinsic EMI compared to a switching electricity converter. The concept enables EMI-immune motor drives suitable for harsh electromagnetic environments and weight-sensitive platforms such as tethered UAVs. Future work should focus on improving end-to-end energy conversion efficiency (HPLD and PPC), optimizing system integration for broader operating ranges (speed/torque), and expanding applicability in diverse environments. Scaling PPC adoption and reducing PPC cost will further enhance system practicality.

• Energy conversion efficiency: The overall electrical–optical–electrical conversion efficiency requires improvement to maximize usable mechanical output.
• PPC cost and availability: PPCs are not yet widely used in industry and are relatively expensive, which may limit near-term adoption.
• Inherent speed/voltage limits: The motor speed is bounded by PPC open-circuit voltage and available photocurrent; achieving higher speeds/torques requires appropriately rated PPCs and optical power.
• Prototype scope: Results are based on an off-the-shelf component prototype under laboratory conditions; broader operating conditions, long-duration tests, and alternative motors/converters are future work.
• Fiber attenuation: Although modest for short distances, fiber losses depend on wavelength and length; system design must account for transmission distance to maintain required power at PPCs.