This technical note presents a research-grade protocol to test the hypothesis of direct piezoelectric conversion of beta radiation into electricity using static ceramic PZT and engineered PZT-X formulations. The approach is fully solid-state (no macroscopic mechanics), with single-lamella and multilayer “sandwich” geometries, guard-ring electrodes, and low-loss fixtures. Irradiation is performed with a collimated Sr-90 source and with linac electron beams (0.5–2 MeV). Deposited power P_dep is operationally defined over the active piezo volume and estimated via Monte Carlo (e.g., Geant4, region tagging) and/or NIST PSTAR/ESTAR stopping-power data. Quantitative metrics include full I–V curves, open-circuit voltage (Voc), short-circuit current (Isc), maximum power point (P_max), and an electrical efficiency eta_elec = P_max / P_dep with a preregistered detectability target eta_elec >= 1e-4. To ensure rigor and reproducibility, the protocol specifies blind controls (glass, sapphire, amorphous silicon), poled vs. unpoled PZT, and above-Tc checks; lock-in detection at off-mains frequencies with FFT/PSD analysis; a geometric-mechanical transfer parameter (kappa) calibrated on the real geometry; ABAB randomized sequences; and explicit Go/No-Go criteria. Practical sections cover dosimetry and graded-Z shielding, minimum instrumentation specifications (low-noise electrometer, charge preamplifier, synchronized DAQ, triax cabling in a Faraday cage), open-data deliverables, and a minimal power analysis to size experiments. The study is designed to yield value regardless of outcome: a positive result would establish a new “radiopiezoelectric” conversion pathway; a null result would set meaningful bounds on the phenomenon. Format: bilingual PDF — full English text followed by the complete Italian translation. Note to Version 2:This version is presented in English only and includes several clarifications and refinements in the description of the experimental protocol. Note to Version 3This version further develops the protocol by explicitly reframing the experimental objective around a hybrid betavoltaic–ferroelectric (BV+FE) architecture. In this scheme, wide-bandgap betavoltaic junctions (SiC, GaN, or diamond) provide the primary conversion of β deposition into electron–hole pairs, while an ultrathin ferroelectric layer (HfZrO₂ or engineered PZT-X) enhances charge separation, reduces recombination, and introduces a rectifying field effect. The detection metrics are correspondingly updated: beyond the standard I–V characterization, the protocol specifies relative efficiency gain Δη_rel ≥ 20% of BV+FE versus BV-only devices as a preregistered success criterion, and the observation of a zero-bias current inversion with ferroelectric poling reversal as the unambiguous signature of the FE contribution. Blind controls, lock-in detection, randomized ABAB sequences, and graded-Z shielding remain unchanged, while the documentation of Go/No-Go thresholds has been sharpened. The experimental plan now emphasizes the value of a null result as well: failure to observe the FE-induced signatures will conclusively bound the role of piezo/ferroelectric assist in betavoltaic architectures. Format: English only.