Monolithic crystal filter
- summary, tutorial or overview of the basics of the monolithic crystal filter describing its operation and giving its equivalent circuit.
Radio receiver selectivity and filters tutorial includes:
• Radio receiver selectivity • Receiver filter options • Filter specifications & parameters • RF crystal filters • Ceramic RF & IF bandpass filters • Monolithic crystal filter • Mechanical filters • SAW filter • Image rejection
Quartz crystal filters are widely used in many areas, and in particular in high performance radio receivers. They are able to offer unparalleled levels of performance at a cost that represents excellent value for the performance.
A development of the basic idea of a crystal filter is the monolithic crystal filter. These monolithic crystal filters are able to offer even higher levels of performance in some respects while costs are reduced slightly.
What is a monolithic crystal filter?
Traditionally a crystal filter is made from a number of discrete crystals with the circuit often based around the half lattice network. However these designs require the use of a number of individual crystals - often six or eight are used to give the required performance.
Rather than having a number of discrete crystals, a monolithic filter uses a single crystal element and two sets of electrodes plated onto the surface. There are two ways in which this can be done. The first is to have two sets of identical electrodes, top and bottom. Alternatively a single electrode can be plated onto one surface acting effectively as a common ground with two electrodes, one for the input and the other for the output at the top.
Circuit diagram of a half lattice crystal filter
Like most crystals used for radio frequency applications an AT cut is used - the cut of the crystal is defined by the angle at which the blank is cut from the original quartz crystal and the angle of cut defines many of the properties of the crystal blank and hence the overall filter.
The two sets of electrodes are placed onto the crystal. The electrodes are coupled to each other by the mechanical resonances in the crystal to give a highly selective filter.
Diagram of a monolithic crystal filter
The filter crystal is grown and cut in the same way as that used for normal crystals. Although quartz occurs naturally, and was used at one time for crystal manufacture, most of the material used today is manufactured synthetically. This has the advantage that the crystals are formed under much tighter conditions quality is more uniform. Additionally some of the flaws existing in natural quartz are not present making the overall quality much higher.
Once the raw crystal has been formed it is cut into blanks which are lapped and polished to a very high degree. The final stages of manufacture usually involve chemical etching as this gives a much finer finish. As a result the effects of ageing are greatly reduced. The size is very important because this determines the final resonant frequency.
Once the blank has been completed the electrodes are deposited onto the quartz. These are normally aluminium, silver or gold, and they have to be deposited under very controlled conditions so that they cover the required area and have the correct thickness. The thickness of the electrodes can be used to trim the filter to its exact resonant frequency. Making the electrodes slightly thicker reduces the frequency allowing the final performance to meet very stringent requirements.
Monolithic crystal filter operation
Like the standard quartz crystal, the monolithic crystal filter relies upon the piezo-electric effect that is exhibited in quartz for its operation. When signals appear across one pair of electrodes they set up mechanical vibrations on the crystal. These are affected by the mechanical resonances of the crystal element, and only those signals within the pass-band of the filter are allowed across the crystal to be picked up by the second pair of electrodes.
The primary resonant regions of the filter are between the electrodes on the upper and lower surfaces. The thickness of the wafer between the electrodes will be equal to nλ/2 where λ is the wavelength of the vibration in the crystal.
Monolithic crystal filter resonance
The operation of the filter can be explained in terms of an equivalent circuit. There are several elements to this, each adding to the overall performance of the filter.
Equivalent circuit of a monolithic crystal filter
There are several element to the operation of the monolithic crystal filter circuit.
- L1 / C1 & L2 / C2 : These are the two series resonant circuits. The actual values of these equivalent components are determined by the mechanical dimensions and properties of the quartz element.
- L3: This represents the internal coupling within the internal coupling between the two resonant circuits. It is typically equal to k x L1, assuming L1 = L2. It can be calculated from the bandwidth divided by the centre frequency (B / F0). Typical values for the coupling constant are around 0.0005.
- Co: This is the parasitic capacitance between the top and bottom electrodes of the input and output - they are assumed to be the same. This capacitance can be accommodated within the input and output matching networks that are outside the filter and within the external circuitry.
- Cp: Cp is the parasitic leakage capacitance across the resonant element of the monolith crystal filter. This capacitance needs to be kept as small as possible to prevent signal leakage across the filter which impairs its stop band attenuation performance
Monolithic crystal filters can be designed for use over a wide range of frequencies. Costs rise, though, as the frequencies increase as manufacturing becomes more exacting and reject rates rise as the crystals become much smaller and more fragile.
Where operation at high frequencies is required, these monolithic filters can be run in an overtone mode, and this results in the crystal elements being larger, and hence more robust, than if they were designed for fundamental frequency operation.
Monolithic filter impedance
When using any form of filter it is necessary to ensure that it is terminated with the correct impedance. The same its true for monolithic crystal filters. When designing the filter it is necessary to be able to calculate its impedance. Typically it is between 500Ω and 10kΩ. It is relatively easy to make a good estimate of the impedance of the input and output.
Z = impedance of impedance of filter
B = bandwidth in kHz
fo = centre frequency in MHz
n = overtone used for the filter
R is a constant - typically between 1 kΩ and 2 kΩ
Multi-stage monolithic filters
In order to improve the performance of a monolithic crystal filter, it is possible to add further poles. This will increase the rate of cutoff and stopband attenuation. However as the number of poles increases within a single crystal unit, unwanted modes become more difficult to control.
To overcome this, while still increasing he performance higher order filters are generally made by connecting several monolithic crystal filters in series. This will enable the performance of multi-section crystal filters to be replicated in a monolithic format.
Filter made using multiple sections are often called tandem monolithic crystal filter.
Although often cheaper than a crystal filter made from discrete components, a monolithic crystal filter can still be expensive. Nevertheless monolithic crystal filters are able to provide a high level of performance Also with quartz crystal technology improving, their cost is likely to fall somewhat over the years and the performance improve steadily.
By Ian Poole
Popular radio receiver tutorials . . . . .
|• Superhet radio||• Sensitivity||• Software radio||• Cognitive radio|
|• FM demod||• AM demod||• DSP tutorial||• Selectivity & filters|