I have this script which isn't making that much sense.

clear all
kp = 0.000065;
ki = 0;
kd = 0;
b = pid(kp,ki,kd);
d = (tf([1 0],[1]))
f = tf([11.53 1.068],[1 0.09834 6.263*10^(-5)])
velocity = feedback(b*f,d)
a = pole(velocity)
b = zero(velocity)
step(velocity)

Looking at it's step input, you will see that the final value is incredible large considering that response is due to a step input.. i know something must be wrong but what is wrong?