LED lights have become very popular recently for smart farming applications, where they provide artificial light as the substitute for sunlight in a greenhouse or indoor farming environment. To ensure a low total operational cost and improve the efficiency of farming in those environments, it is imperative that the overall LED lighting system is energy efficient. LED is a dc system, whereas the grid is an ac system. As such, an LED driver is needed to perform the necessary voltage conversion. The boost power factor correction (PFC) converter is a popular LED driver that provides output voltage regulation and power factor correction at the same time. As the LED driver is grid-connected, its control system requires real-time estimation of the grid voltage parameter information for reference current generation. In this study, a comparison between frequency-and phase-locked loops as the grid detection method inside the converter control system is provided. For the phase-locked loop (PLL), a single-phase quasi-type-1 structure is considered. It is then compared with the conventional second-order generalised integrator (SOGI)frequency-locked loop (FLL). Comprehensive numerical studies are performed to evaluate the performance of FLL and PLL in challenging grid voltage cases. Results show that the source current has a lower total harmonic distortion when PLL is used as the synchronisation tool over the FLL counterpart. This can be attributed to the use of moving average filter in the PLL, which provides additional harmonic robustness compared to FLL. Lower distortion by the PLL method will the make LED driver, consequently smart farming more energy efficient.