<p>Chaos is widely used in data encryption due to its distinctive advantages. However, most existing chaotic encryption models are based on dissipative systems, which are susceptible to attractor reconstruction and pose significant threats to encryption security. To address this issue, this paper proposes a conservative vortex-like chaotic system with tunable parameters that features conservativity and hidden characteristics, along with a concise model structure and rich dynamical behaviors. Through equilibrium point analysis and dissipativity verification, the hidden and conservative properties of the chaotic system are demonstrated, and complex dynamic behaviors such as vortex-like patterns, multi-stability, and transient transitions are observed. In addition, analog circuits and FPGA-based digital implementations are designed, confirming the engineering feasibility of the system. On this basis, an audio encryption algorithm is proposed, which utilizes chaotic sequences generated by the system to perform scrambling and diffusion operations on audio signals. Numerical simulations and performance evaluations show that the proposed algorithm significantly reduces temporal correlation in audio data, improves security and anti-attack capability, while achieving faster encryption speed.</p>