<p>Open Radio Access Network (O-RAN) has emerged as a disruptive paradigm for mobile networks by disaggregating traditional RAN functions into software-based components interconnected through open and interoperable interfaces. By enabling multi-vendor integration, cloud-native deployments, and programmability via rApps and xApps, O-RAN promises increased flexibility, innovation, and cost efficiency. However, due to the rapid evolution of specifications, software stacks, and experimental platforms, practical knowledge on how to experiment with O-RAN remains fragmented and difficult to access, particularly for laboratories with limited hardware resources. This paper presents a structured theoretical and practical study of the O-RAN ecosystem aimed at lowering the entry barrier to academic and small-scale research. First, we consolidate the architectural foundations of O-RAN, including its main components, interfaces, and control principles. Then, we comparatively evaluate five representative open O-RAN testbeds spanning simulated, hybrid, automated, and flexible emulated deployments, all analyzed from the perspective of reproducibility, deployment complexity, realism, and resource requirements. Rather than providing an exhaustive benchmark of O-RAN performance, the paper offers a reproducible, decision-oriented guide to help researchers select the most appropriate experimentation platform for their needs. Illustrative use cases are implemented on the most promising platforms to validate monitoring, control, and integration capabilities. All configurations and scripts are publicly released to support reproducibility and future experimentation on open and intelligent RAN solutions.</p>