Design of practical filters involves consideration of a large variety of disciplines and factors. These include the electrical, physical, and economic properties of resonant and coupled elements, the materials and processes used for fabrication (properties and cost factors), and the labor cost associated with assembly and adjustment. Modeling is a vital asset in the design process, but the real properties of filter elements must be incorporated into the modeling process, using an evolutional method in which the model is adjusted to compensate for the unavoidable nonideal nature of the elements, stray couplings, and the like. This is similar to the older, laboratory "cut and try" method, but far less costly and time-consuming. This implies that modeling should accept available measured data as input. Designers should always think about minimizing production labor, considering manufacturing tolerances possibly as a trade-off against tuning time and recognizing that availability of skilled labor is more than simply a cost but rather a constraint on delivery rate. Continued development of accurate and complete models of parts, enclosures, and interconnects, in conjunction with ever-better computational capabilities, could (if properly used) enable rapid and accurate designs of filters with very predictable and producible results
[1]
M.R.M. Rizk,et al.
An Interactive Optimal Postproduction Tuning Technique Utilizing Simulated Sensitivities and Response Measurements
,
1981,
1981 IEEE MTT-S International Microwave Symposium Digest.
[2]
Richard V. Snyder,et al.
Parallel coupled line notch filter with wide spurious-free passbands
,
2005,
IMS 2005.
[3]
John W. Bandler,et al.
A nonlinear programming approach to optimal design centering, tolerancing, and tuning
,
1976
.
[4]
John W. Bandler,et al.
Functional Approach to Microwave Postproduction Tuning
,
1985
.
[5]
R.V. Snyder,et al.
New Application of Evanescent Mode Waveguide to Filter Design
,
1977,
1977 IEEE MTT-S International Microwave Symposium Digest.