Bandstop filters are used in microwave systems to remove unwanted signals over a specific frequency range while passing signals with frequencies that fall outside of that range. Such filters can be used to reflect or absorb unwanted signals in a microwave system which may originate from co-site or externally generated interference as well as nonlinear components under high-power excitation in the system. For example, a traditional microwave bandstop filter can be composed of resonators coupled to a through-line with quarter-wavelength admittance inverters between each resonator. This bandstop filter topology can produce a symmetric notch frequency response and meet a wide variety of practical specifications. However, when the traditional microwave bandstop filter topology is used for high-order filters, the total through-line length becomes overly long.
Switched bandstop filters implement a reconfigurable frequency response at the front-end of an RF system that allows it to receive or reject a specific frequency band on demand. The conventional switched-bandstop-filter configuration is comprised of switches in conjunction with a bandstop filter and bypass transmission line in parallel. The switches route an RF signal to the internal direct path for all-pass mode or to the filter path to engage the bandstop filter. In this approach, the RF signal experiences significant insertion loss from the switches.
In the case of reconfigurable or frequency-agile systems, it is often a requirement that each bandstop filter have bypass capability. Using two signal-routing radio frequency (RF) switches for every bandstop filter typically results in considerable insertion loss when multiple switched filters are cascaded due to the losses of the switches.
Researchers at the Navy have devised systems and methods to minimize additional insertion loss from switches using all-pass networks in conjunction with SPDT (Single-Pole-Double-Throw) switches. In contrast to conventional switched bandstop filters, the switches in the Navy structure are located outside of the main signal path. Therefore the transmitted signal does not go through the switching network directly, and the proposed configuration drastically reduces the insertion loss compared with conventional switchable filters. It also minimizes the discrepancy in passband group delay between all-pass mode and bandstop filter mode without supplemental lengths of transmission line, which is only possible in conventional multi-path type switchable filter designs by increasing size.
- Self-switching bandstop filters that do not require signal-routing RF switches, allowing for very low passband insertion loss and improved power handling
- A key advantage of this approach is less discrepancy of the passband group delay between all-pass mode and bandstop filter mode
- Less through-line length required by the inductive coupling-cancellation structure than by previous approaches, resulting in smaller size
- Higher operating RF power levels can be achieved for a given switching element
- US application number 20170085241 available for license
- Potential for collaboration with Navy researchers