33.8MHz ceramic resonator, what is it's bandwidth?
Have you got any idea what their bandwidth is?
I would like to find out if I can replace the crystal filted used in this circuit ../imgqa/eboard/Antenna/rf-5wsitjgxpcj.gif so that I have a much higher IF.
I'm not sure they will work well as filters, they are probably used as the clock resonators as there is no need for signal filtering in the CD/DVD ROM application. If they have 3 pins they almost certainly have internal loading capacitors which makes it difficult to match them together in a chain. They probably also have poorer frequency matching than Quartz crystals so even if they did connect in a chain, the bandwidth edges wouldn't necessarily line up together. "Your mileage may vary" as they say.
Brian.
I think that the Q of ceramic resonators is about 2000. Their frequency tolerance is about .5% (From Wikipedia). A crystals data would be Q = 20,000 and tolerance .01%, so it would be unlikely that the performance would be the same.
If you have the test equipment (mainly a frequency counter and some sort of RF voltmeter), you could test your CRs and see what they do when cascaded.
The crystal filter in the circuit, uses the same frequency crystal through out with different resonating capacitors. In professional brick wall filters, the crystals would each have a different resonant frequency, i.e. Fc +100 HZ, Fc +300 HZ, Fc+ 1.1 KHZ. .
Frank
These are 2 pins resonators. I suppose they were used as clock generators, since they were installed near high-pin-count ICs.
I was not clear, I am not talking about cascading these to create a filter, I am talking about using just one, instead of the whole crystal chain in the schematic.
Would that be of any good? I suspect it should be much wider, but sonce the IF is very high does it matter? I could just include an audio LPF select better the SSB signal of choice.
I doubt just one would work well enough. Obviously they have a frequency at which they exhibit low impedance but 'out of band' the attenuation would be quite poor, I'm guessing only around 20dB so you would need more than one in series to get good selectivity. For really low cost, you could consider either 3.57MHz or better still 4.433MHz quartz crystals as used in TV color decoders. They cost amost nothing, are quite accurately cut, have high Q and since digital TV came the normal, are abundant on the market.
Alternatively, consider one or more of these: http://www.surplectronics.com/products.asp?cat=96 . I have used them in the past to replace ceramic filters and the improvement in skirt response is very noticable.
Brian.
If you manage to preserve the Q when the resonator is in circuit of 2000, this would give a bandwidth of 19 KHZ. As mentioned these devices have a poor out of band rejection, so they need to have a couple of LC circuits to help with this. Your 3 KHZ AF filter will help with the band width, but not with the image within this 3 KHZ on the other side of the carrier. Their high insertion loss also means that either linearity or noise factor will suffer in trying to make up the gain. Its much better to have linear amplifiers for at 3.5 MHZ, the amount of man made noise will drown out the extra loss of a dB of noise factor.
Frank
Ah yes, I have some of these. Best case for a single device is 7.5KHz at 3db, not bad considering the simplicity (ready to go filters).
Ladder filters were always the problem in homebrew designs as determining the crystal parameters is not a so easy task.
I wonder if I place an inductor based eliptical LPF in series with an eliptical HPF, both designed for the same frequency, will they be able to be used as ssb filters effectively enough?
Elliptical filters have a deep notch, which makes their response quite sharp, biut I do not know if they would be suitable for ssb at all (even with some compromize)
So basically you want to make a bandpass filter by passing the signal through two eliptical filters. In theory it would work but I suspect you would have difficulty tuning the filters accurately enough. Eliptical has a steep roll-off toward the notch but to make two stable filters, working in opposite directions would be quite difficult. You would have to place a buffer stage between them and probably at each side as well.
Just a thought - I have never tried it in that kind of reciever but you could try a simple 'Q' multiplier LC filter, it would be far easier to tune and the bandwidth would be adjustable. The down side is it would have a symetrical peaked response. You really need a flat topped response with steep slopes a few KHz to each side.
Brian.
In low IFs even some opamps would work. I do not know how well will they respond to LC.
I have tried simulating an elliptical BPF, like you said, the passband is not flat and even at higher number of poles they cannot achieve the bandwidth of the crystal. I suspect any thermal or component tolerances will have a critical effect on the center frequency and maybe pull it too far.