Losses Incurred by Actual Fabricated Microstrip Circuit

[magneto_cool]

Hi I have designed and fabricated a 5th order parallel coupled bandpass filter, however i've noticed that the gain of the actual measured circuit is much lower than that of the schematic simulated case (refer to attached diagram for simulated and measured results) thus i would appreciate it if anyone one can tell me what are the factors that contribute to the higher loss and how this loss could be reduced.

Any help would be greatly appreciated.

Thank You

 

[flyhigh]

Hi,

your filter is mismatched meaning that resonators are not on the right frequencies. You can tell this by examining the reflections - you don't have distinct dips in response. Regarding this, even your simulation doesn't look 100% OK - reflection dips for this class of filter should be symmetric in the passband. One reason for discrepacy between simulation and measurement might be that you did't include step discontinuity in your schematic. After this you should retune to achieve desired response. As a final check do EM simulations

flyhigh