Hui (1998) discussed the steady state error occurring in the examples given in Tyan and Bernstein (1997). As was pointed out in Hui (1998), the steady state error was due to the under capacity of control amplitude due to the saturation nonlinearity. In this responding letter, we increase the saturation level and demonstrate that even though the requirement of saturation capacity is satisfied, the given LQG controller still gives totally unsatisfactory response compared to that of the controller derived in Tyan and Bernstein (1997).