A cross flow jet SOG has been developed in Miki Pulley Co. Ltd. to supply O2(1D) for different types of COILs. Performance testing of the SOG has been conducted through a wide range of gas pressures (5~40 Torr), specific surface areas (4~7 cm-1), gas velocities (5~30 m/s), and gas temperatures to characterize and optimize the device. The inflow and out flow of the reactants and products, including O2(1D), Cl2, H2O were measured using optical and conventional techniques. The gas temperatures in the measurement duct were estimated from stagnation pressures, mass flow rates, and critical cross section at the gas chocking point in order to determine the partial pressures of the gas products at the measurement point. Calibration method of the O2(1D) measurement suggested by Zagidullin is basically employed with a slight correction of upper limit definition of O2(1D) yield associated with the pooling loss, which remains even at the minimum P t condition of our device. Assuming that the gas temperature after passing through the jets is equilibrium with that of the BHP jets (-18 degree Celsius) in our calibration condition, the upper limit yield can be derived from the increase in the gas temperature. The estimated value of the yield limit was 94 %. A wide range of output values (40-95 % of Cl2 utilization, 50-90 % of O2(1D) yield) was obtained and analyzed to characterize the device. As a result of optimization, a 27 % of chemical efficiency was obtained when Cl2 utilization was 95 %, O2(1D) was 90 %, O2 partial pressure was 6.7 Torr, and N2 dilution ratio was 2. Discussion on the validity of the gas temperature estimation method is provided by comparing the results to the heat release based on the pooling model.