Post subject: FM Modulator Design Posted: Sun Oct 01,
2006 3:52 pm
Joined: Sun Oct 01, 2006
I am new here, I am trying to
design an FM modulator at 12MHz that could gives Deviation of +/-4KHz
and distortion not more that 2%. I tried direct modulation by using
Varactor in parallel or series Crystal. but I did not get more that
1 KHz. I tried 4046 with Divider N =1 but did not succeeded. Any
one have good idea might help. Also it is very strange that most
of RF perctical books avoid even discussing this topic. can any
one give me a good mathmatical design procedure.
noticed some circuits in which used a parallel LC resonent in parallel
to crystal. I tried that too but it was not stabe, I do not know
subject: FM Modulator DesignPosted: Tue Oct 03, 2006 6:12 am
Joined: Fri Feb 17, 2006 12:07 pm
Location: London UK
The problem arises because a crystal
has a very high Q, and the capacitance in the equivalent circuit
cannot be easily modified by a varactor without seriously degrading
the frequency stability. Typically only a few hundred Hz deviation
can be obtained.
There are some VXCO circuits that have overcome
this problem, although some trade-off occurs with average crystal
stability. Do a search under VXCO and you should find some circuits.
You could also try the web site "allaboutcircuits.com"
approach that has been very successful is to use ceramic resonators
at 12MHz in place of a conventional quartz crystal. Most of the
useful and detailed circuits for this requirement can be found in
Amateur Radio publications such as QST, Rad Com(UK) and Dubus (Germany)
Post subject: Re: FM Modulator DesignPosted:
Sat Oct 14, 2006 4:05 am
Joined: Sun Oct
01, 2006 2:41 pm
a conventional quartz crystal. Most of the useful and detailed circuits
for this requirement can be found in Amateur Radio publications
such as QST, Rad Com(UK) and Dubus (Germany)[/quote]
VCXOPosted: Thu Oct 19, 2006 4:00 pm
Joined: Wed Feb 22, 2006 3:51 pm
reference is the ARRL Handbook (almost any recent edition).
Most designers don't try to modulate a crystal oscillator's
frequency directly - even if you can get enough deviation, there
will be a (rather low) upper limit to the modulating (baseband)
frequency that you can use.
So what to do? The classic method
is to use a varactor-tuned LC oscillator in a phase-locked loop,
with the VCO output divided down to a relatively low frequency for
comparison to a (stable, crystal-derived) local reference.