Weak Electrical Signal Measurement
with Optical Lever Amplification Design & Implementation Based on Electromagnetic-Mechanical-Optical Conversion

Electrical signals at microampere level and below are easily obscured by environmental noise; increasing electronic gain alone amplifies front-end noise as well. This system performs electromagnetic, mechanical, and optical conversion before precision digital acquisition. It does not directly amplify the electrical signal: it converts current-driven deflection into an easier-to-measure spot displacement.

The paper then presents the overall structure, module implementations, calibration and linearity verification, error analysis, applications, limitations, and future work in order.

The system uses a rectangular frame as mechanical support with a modular design dividing the apparatus into four functional modules. The left side houses the suspended deflection structure; the right side accommodates laser emission, spot detection, and data processing circuits. The receiving and deflection modules employ an integrated coil structure wound from a single continuous copper wire, reducing connection losses. The entire apparatus incorporates electromagnetic shielding, with the data processing section isolated in a metal enclosure.

The generation module consists of a computer audio output, an Arduino control unit, and a Si4713 RF transmitter chip. The computer generates low-frequency modulation signals with controllable amplitude and frequency, fed into the Si4713's LIN and RIN inputs via a 3.5mm audio jack. The Arduino initializes and configures the Si4713 through I2C. The Si4713 integrates frequency synthesis, modulation control, and RF output — modulating the audio signal onto a specified RF carrier and transmitting via antenna.

The low-frequency modulation signal is generated by a Python script (100Hz sine wave via sounddevice). The Arduino handles Si4713 communication control. Compared to traditional digital waveform generation, this approach leverages computer audio output for high-stability modulation, improving frequency and amplitude control precision.

The receiving module adopts a lightweight, movable structural design, consisting of a multi-turn receiving coil with NdFeB magnets arranged on both sides with opposing poles. This creates a directionally stable magnetic field in the coil region. When an external weak alternating current signal propagates through space creating a changing magnetic field, the flux through the coil varies with time. According to Faraday's law, an induced EMF and current are generated, transmitted through a closed loop to the deflection coil.

The receiving and deflection coils use an integrated coil structure (one continuous copper wire), directly connected. This eliminates contact resistance and connection stability issues inherent in traditional separated connections, significantly improving system sensitivity.

The deflection module consists of a suspended deflection coil, permanent magnets, a reflector, and mechanical support. The coil is suspended by two fine copper wires; the reflector is fixed to the coil and rotates synchronously. A collimated laser beam from the right strikes the reflector and is reflected back onto a one-dimensional PSD.

When induced current I flows through the coil, the current-carrying coil experiences an Ampere force in the constant magnetic field B, generating a rotational torque. The suspension wire simultaneously produces a restoring torque. At equilibrium, the reflector deflects by θ, the reflected laser deflects by 2θ, traveling distance D to the PSD. Under small-angle conditions (θ ≤ 5°), spot position x and current I satisfy an approximately linear relationship:

Where k is the wire's equivalent torsional stiffness and N the coil turns. The deflection module converts weak current into observable optical position signals.

Normalized Position Operator P — The PSD outputs I_a, I_b from its two terminals. To eliminate laser power fluctuations and ambient light effects:

Within the operating range, P and induced current I satisfy a linear relationship (C from calibration):

Signal Frequency f — Continuous PSD acquisition with zero-crossing detection:

EMF Amplitude U — Receiving and deflection coils in series, impedance measured via LCR bridge:

The data processing module comprises a PSD sensor, AD8628 low-noise conditioning circuit, ADS1115 16-bit ADC, and Arduino. The PSD uses the lateral photoelectric effect to produce differential photocurrents. The AD8628 forms a transimpedance amplifier for I-V conversion (low input bias current, low offset voltage). The ADS1115 communicates via I2C with the Arduino for high-precision digital sampling.

The program uses modular design: ambient light compensation reduces background interference, a moving average filter (window N=16) suppresses random noise, and zero-crossing detection computes signal frequency. Filtered data is used for comprehensive computation of spot position, current, voltage, and frequency — output via serial.

A precision sense resistor R_s = 220Ω is connected in series with the deflection coil loop. The actual coil current is obtained via I = V_R / R_s. The Arduino outputs PWM drive signals at different duty cycles across 6 levels, collecting 17 (I_i, P_i) pairs per level. After accumulating 100 data pairs, least-squares fitting yields conversion coefficient C, zero-offset b, and R² correlation coefficient.

The calibration module consists of a current output section (PWM drive + sense resistor) and a coefficient computation section (reusing data processing hardware). Calibration requires no hardware changes — merely switching the Arduino program toggles between calibration and measurement modes, ensuring identical signal paths and avoiding errors from additional measurement equipment.

Accuracy Improvement Strategies

Hardware: wind shielding, optimized support, shortened analog signal paths, enhanced shielding. Software: moving average filtering, outlier removal, appropriate sampling parameters. Procedures: optical alignment and zero-point calibration before experiments, periodic baseline verification during long measurements.

The system combines receiving and deflection coils, an optical lever, and a PSD into a complete electrical-mechanical-optical-digital measurement chain. Calibration maps input current to PSD output, while modular hardware and digital filtering make the work a reproducible platform for weak-signal and physics experiments.