Q: How does a quartz-crystal filter work?
Martin Biallas
Many books describe the topic "filters" very detailed, well at least the
LC-Filters. In most books the description begins with a simple LC-Filter.
Then the authors write about more special filters like "Bessel",
"Butterworth" and active filters. But at the end of the "filter" topic, you
can read something like "and then there are crystal-filters, which can be
used nearly everywhere when you need a sharp dropoff outside the limits of
a band ...".
Up to now, I couldn t figure out how they work. Maybe someone could try to
explain me how a crystal-filter works.
So, thanks in advance, and ...
Bye ...
Martin
bi...@clobber.unterland.de
To get my PGP Pubkey send a mail with Subject 'SendPGP' (NOT via FIDO!!).
Leon Heller
bi...@clobber.unterland.de "Martin Biallas" writes:
> Many books describe the topic "filters" very detailed, well at least the
> LC-Filters. In most books the description begins with a simple LC-Filter.
> Then the authors write about more special filters like "Bessel",
> "Butterworth" and active filters. But at the end of the "filter" topic, you
> can read something like "and then there are crystal-filters, which can be
> used nearly everywhere when you need a sharp dropoff outside the limits of
> a band ...".
> Up to now, I couldn t figure out how they work. Maybe someone could try to
> explain me how a crystal-filter works.
A crystal behaves just like an LC resonant circuit (it has both parallel
and serial resonance) with a *very* high Q (10,000 or more). Crystal
filters basically *are* LC filters, and can be analysed in the same way.
The two equivalent circuits look like this:
Parallel
Lm Cm Rs
------L---C---R------
| |
|------C-------|
Cp
Serial
Lm Cm Rs
-----L---C---R----
See W1FB's Design Notebook (ARRL) for simple techniques for designing
and building crystal ladder filters. Ladder filters are nice, because
all the crystals are identical. TV colour-burst crystals are very cheap,
and work quite well, as do microprocessor clock crystals.
A typical three-crystal ladder filter looks like this:
--C---X---X---X---C--
| |
C C
| |
---------------------
The three crystals, X, all have the same value, as do the capacitors
(C). This is in fact a Cohn filter configuration, which is sometimes
used with ordinary inductors and capacitors. By adding identical
sections, the shape factor can be improved. Using these techniques,
home-made filters can be constructed that have equivalent performance to
commercial filters, at a fraction of the cost. Many transmitters and
receivers constructed by radio amateurs use ladder filters.
Leon
--
Leon Heller, G1HSM | "Do not adjust your mind, there is
E-mail le...@lfheller.demon.co.uk | a fault in reality": on a wall
Phone: +44 (0)1734 266679 | many years ago in Oxford.
CLIF PENN
>Hello everyone!
>Many books describe the topic "filters" very detailed, well at least the
>LC-Filters. In most books the description begins with a simple LC-Filter.
>Then the authors write about more special filters like "Bessel",
>"Butterworth" and active filters. But at the end of the "filter" topic, you
>can read something like "and then there are crystal-filters, which can be
>used nearly everywhere when you need a sharp dropoff outside the limits of
>a band ...".
>Up to now, I couldn t figure out how they work. Maybe someone could try to
>explain me how a crystal-filter works.
>So, thanks in advance, and ...
>Bye ...
> Martin
>bi...@clobber.unterland.de
>To get my PGP Pubkey send a mail with Subject 'SendPGP' (NOT via FIDO!!).
If you can find an old enough reference book which shows the
equivalent circuit you will find that a crystal filter is equivalent
to a VERY high Q parallel resonant LC circuit (for the peak response)
and nearby another high Q series parallel circuit to greatly attenuate
nearby interfering signals.
A. Shiekh
wrote:
I found a quartz too narrow, and one might opt for a ceramic
resonator (smaller Q)
Robert Groover
>} >Up to now, I couldn t figure out how they work. Maybe someone could try to
>} >explain me how a crystal-filter works.
It's an electroMECHANICAL resonator: since the crystal is piezoelectric,
applied voltage will cause it to contract along one axis and store
energy. The amplitude of mechanical movement is very small, so there is
little energy dissipation and hence high Q.
There are sublteies about how the crystal is sized and oriented to achieve
a desired frequency, but that's the basic idea.
I think there's more about these in the old _Radiotron Designer's
Handbook_, but I'm not sure.
Robert Groover
Del Stanton
>Hello everyone!
>Many books describe the topic "filters" very detailed, well at least the
>LC-Filters. In most books the description begins with a simple LC-Filter.
>Then the authors write about more special filters like "Bessel",
>"Butterworth" and active filters. But at the end of the "filter" topic, you
>can read something like "and then there are crystal-filters, which can be
>used nearly everywhere when you need a sharp dropoff outside the limits of
>a band ...".
>Up to now, I couldn t figure out how they work. Maybe someone could try to
>explain me how a crystal-filter works.
>So, thanks in advance, and ...
>Bye ...
> Martin
>bi...@clobber.unterland.de
>To get my PGP Pubkey send a mail with Subject 'SendPGP' (NOT via FIDO!!).
A simple "mechanical analogy" explanation.
The crystal is piezoelectric, applied voltages make it bend,
bending it generates voltage. It is also a piece of
material with a mass and elasticity. If you had it mounted
appropriately and gave it a mechanical tap it would vibrate,
mechanically oscillate, at a frequency detemined by its
geometry and method of mounting. In this way it is just
like the tube of a wind chime, a hacksaw blade clamped in a
vise or a tuning fork.
When placed in an electrical circuit the driving force is
the AC voltage applied. If the AC voltage is varied in
frequency from below the crystals resonant frequency to
above its frequency the crystal will oscillate very strongly
(yes - mechanically oscillate - admittiedly with a very
small amplitude, but it moves, even at RF rates) only when
the frequency is quite close to its natural resonant
frequency. The sharpness of the peak depends on the
mechanical losses in the crystal. Since a crystal is almost
a perfectly elastic, when it is bent and released it returns
all the energy used to bend it, it has a very high Q. And a
high Q results in a sharp resonant peak.
Imagine that you have a variable oscillator driving an
audio amplifier connected to a speaker and a tuning fork
mounted in front of the speaker. If you watch the tuning
fork it will break into oscillation only when the
oscillator's frequency is at the resonant frequency of the
tuning fork. If you mounted coils on the tines of the
tuning fork and had then in a strong magnetic field then the
only time you would get any appreciable output from the
coils would be when the tuning fork was oscillating at its
resonate frequency, the rest of the time the output would be
vanishingly small. And that output would be at the resonant
frequency of the tuning fork. Thus your tuning
fork/coils/magnet filter would only pass one frequency,
rejecting the rest.
When you push a kid on a swing you have to time your pushes
to match the resonant frequency of the swing. Too fast or
too slow doesn't work, you just end up bruising the kid.
A hope this clarifies the operation of a crystal as a
filter.
Del Stanton - Burbank, California "sd...@pacificnet.net"