Answers are grounded in this paper and cite their sources.
Conventional battery alternating current internal resistance (ACIR) measurement faces challenges in generating adjustable high-current excitation with low energy dissipation and accurately acquiring weak alternating current (AC) voltage responses. This paper proposes an H-bridge-based battery ACIR measurement system with energy recovery. The battery under test directly supplies the excitation energy, enabling bidirectional energy transfer through the H-bridge and an energy-storage inductor. Fixed-frequency phase-shift modulation generates amplitude-adjustable bidirectional test currents at 1 kHz. An active direct current (DC) offset-cancelation sensing architecture improves analog-to-digital converter (ADC) input-range utilization for millivolt-level AC voltage responses, while a single-bin discrete Fourier transform (DFT) extracts the 1 kHz fundamental components for ACIR calculation. At measured peak test currents of 20, 100, and 400 mA, the charge return ratios were 8.53%, 9.81%, and 12.49%, respectively. Under the constant-terminal-voltage approximation, these values also estimate the fraction of energy drawn from the battery that was returned to it during the same excitation cycle. During 15 min repeated measurements, the proposed system exhibited lower apparent temperatures than the comparison instruments under the respective test conditions, with no detectable change in battery open-circuit voltage at the reported resolution. For the same conditioned signal, the 1 kHz fundamental voltage amplitude obtained from the ADC samples was 2.34% lower than the oscilloscope result. After channel calibration, the proposed system measured the 1 kHz ACIR of seven lithium-ion cells using the 20 and 100 mA excitation settings. Across the seven cells, the maximum and mean absolute relative deviations were 3.50% and 1.60%, respectively, relative to the VICTOR 3540, and 2.65% and 1.26%, respectively, relative to the FNIRSI HRM-10.
No takes yet. Share an insight, caveat, or question.
Lao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: