Due to the unbalanced three-phase currents and the torque ripples and speed fluctuations to the drive caused by the current sampling errors of three-phase current sensors and corresponding signal processing circuits, the estimation on current sampling errors is of great importance to eliminate these adverse effects. The observer-based methods, however, present two defects: first, they could increase the computational burden to the drives, thus limiting their practical applications; second, their high dependence on the system speed or position feedback information might degrade the estimation accuracy. To solve these problems, this article proposed a fixed-point sampling method of lower branch flows and three phase-currents, which adopts the relations between the measured currents, rather than any complex algorithm, to estimate the current sampling errors. To coordinate with the proposed strategy, a fixed-point sampling scheme was accordingly rolled out. Compared with the observer-based methods, as a fixed-point one, much applicable to actual digital controllers, it could reduce the computational burden of the drive. Meanwhile, its effectiveness was verified by experimental results on a 1.5-kW IPMSM prototype, which showed accurate estimated current sampling errors and the effectively eliminated torque ripples and speed fluctuations.