20-08-2012, 11:23 AM
RF & Microwave Filters
Lecture05Filters.pptx (Size: 3.12 MB / Downloads: 41)
Type of Filters
A bandpass filter channels signals with minimal attenuation through a range of frequencies known as the passband, and rejects signals at frequencies above and below the passband.
A band-reject filter (also known as a notch filter) is essentially the opposite of a bandpass filter. It rejects signals across one band (known as the stop band) and allows signals to pass with minimal attenuation at frequencies above and below the stop band.
A lowpass filter channels signals with minimal attenuation below a specifed cutoff frequency, while rejecting signals above that cutoff frequency. The cutoff frequency is commonly a point at which signal attenuation reaches 3 dB.
A highpass filter is essentially the opposite of a lowpass filter, rejecting signals below the cutoff frequency and passing signals with minimal attenuation above the cutoff frequency.
A diplexer is a dual bandpass filter that combines two bands (usually one transmit and one receive) into a common port. It can be constructed using LC filters.
A duplexer is a lowpass and highpass “dual-band” filter that combines two bands into a common port. One of the bands is usually the transmit port and the other the receive port. It too can be constructed using LC filters.
Key LC Filter Specifications
There are several key specifications that are essential when specifying any LC filter, since all will affect not just the filter’s performance, but how it is designed and fabricated. When taken together, they can even determine whether or not a specific problem can be solved using an LC filter or if another filter type is best suited for the task.
Insertion loss is the ratio of signal amplitude before the filter to the amplitude at its output. Insertion loss is almost invariably an important factor and should be as low as possible. It is equally important with low or high signal input levels. For example, heat dissipation increases at higher power levels, and lower insertion loss can help reduce it. When signal levels are low, high insertion loss could reduce the output after the filter to an unacceptable level.
Return loss is a measure of the filter performance. It is an indicator of how close the input and output impedance of the filter are matched. Return loss affects the ripple of the filter in the passband. A return loss or a VSWR of 1.5:1 is generally realizable.
Stopband rejection is the ratio of the unwanted frequency components at the input of the filter to those after it. It can be considered the key filter performance specification, since it equates to the filter’s rejection capability. Typical values can range from 20 to 100 dB, and will vary to some degree over the stopband frequency.
Center frequency is the region equidistant between the filter’s upper and lower cutoff frequencies.
Butterworth Filters
Butterworth filters are usually normalized for an attenuation of 3 dB at the cutoff frequency.
Because of its medium Q, initial attenuation steepness of a Butterworth filter is not as great as some types of filters.
However, its element values are more practical to achieve and less critical.
The rounding of its frequency response near the cutoff frequency can make this type of filter less desirable when a sharp cutoff is required, but its overall favorable characteristics make it widely useful.
Best in-band amplitude flatness, lower stopband attenuation than Chebyshev, better than Chebyshev for group delay flatness and overshoot (usually used as a compromise).
All of the above are realizable in parallel-coupled, direct-coupled, and interdigital filter topologies.
Inductor Selection
The vertically mounted type on nylon formers provide a better Q (about 150 with screening cans) and a better tolerance of about 5%, trimmable if an adjuster core is fitted. They are available with or without screening cans. There is no rated dissipation given by the manufacturer’s data sheet but practical harmonic filters have been found to get too hot to touch with an RF output power of 10W, suggesting this to be the practical limit.
Printed spirals have the advantage of controllable tolerance and low cost. The disadvantage is they take up a large area of PCB and only have a Q in the range 50 to 100.
An area with a height roughly equal to the radius of the spirals should be left clear above and below to avoid affecting the Q . The usefulness of printed spirals is limited to the VHF range.
The final type is not strictly a true inductor, but a transmission line used as an inductor. This method is useful at UHF and higher.