And if you choose b=x/10, x/100, x/1000 ... then each iteration will add the exact decimal point.