We prove a total variation approximation for the distribution of component vector of a weakly logarithmic random assembly. The proof demonstrates an analytic approach based on a comparative analysis of the coefficients of two power series.