We fabricated a closed coils by using resistive-joint method and the joint resistance of the coils were estimated by field decay technique in liquid nitrogen. We used the Runge-kutta method for the numerical analysis to calculate the decay properties. The closed coil was wound by $(Bi,Pb)_2Sr_2Ca_2Cu_3O_x$/Ag tape. Both ends the tape were overlapped and soldered to each other. The current was induced in a closed coils by external magnetic flux density. Its decay characteristic was observed by means of measuring the magnetic flux density generated by induced current at the center of the closed coil with hall sensor. The joint resistance was calculated as the ratio of the inductance of the loop to the time constants. The joint resistances were evaluated as a function of critical current of loop, contact length, sweep time, and external magnetic flux density in a contact length of 7 cm. It was observed that joint resistance was dependent on contact length of a closed coil, but independent of critical current, sweep time, and external magnetic flux density. The joint resistance was measured to be higher for a standard four-probe method, compared with that for the field decay technique. This implies that noise of measurement in a standard four-probe method is larger than that of field decay technique. It was estimated that joint resistance was $8.0{ imes}10^{-9}{Omega}$ to $11.4{ imes}10^{-9}{Omega}$ for coils of contact length for 7 cm. It was found that 40Pb/60Sn solder are unsuitable for persistent mode.