calculation for hairpin filter

Hi Guys,

I have read thru this paper as im currently working on Hairpin filters. I notice there is passband ripple occurs for this conventional hairpin design presented on this paper. Im suprise since the filter was contructred using butterworth configuration, isnt the filter should have flater passband. i even simulated base on the parameter given in this paper the passband ripple still occurs in my simulation. could any one advice why such a issue seen ? i attach the paper for your reference.

2. i also have some doubt on the width of the conventional hairpin as the width of the resonator increases with number of element for example the width increases form 0.86,0.99 adn 1.2, bas on by calcualtion using butterworth protype value doesnt yeild the same patern of width size as my calculation shows the width size decreases with increase in element number. Can any one advice me on any other formula tht can be used to calculated the width of resonator arm base on odd and even impedance?

thanks

regards

-danesh-