In the old days, this is exactly how it was done, but even so, think of it
this way.............what had to be done was to use the same tuning
capacitor, but switch different resonating inductors for each tuning range.
The different tuning range inductors needed different ferrite mixes for the
different frequency ranges. Then, there was the R&D.......I believe that
much of this research data was taken from repeated experimantation. Even
today, in my own experiences with RF simulation programs, we would still
have to build and fine tune the circuit under development. Just to do a new
synthesizer design from start to finish would take between 6 to 18 months
(these are optimistic values). Figure an engineering team of at least 3
people that are being paid between 60k to 120k a year each, and it becomes a
little bit clearer just how high engineering costs can be. This doesn't
count the cost of multiple iterations of the design before the engineering
review team deems things to be complete.
This is one of the reasons that I decided to homebrew many different
receiver designs. This is a good way to learn "the hard way" many of the
things that a manufacturer must go through when working through a design,
from conception to final product.
Newer designs are a little bit easier now that upconversion topologies can
be used. This type of design became practical, once VHF crystal filters
became viable. The stumbling block with VHF crystal filters was that in
order to work, the crystal elements had to be ground so thin that they
became quite fragile. Enter the overtone mode crystal filter..........now,
you could use a much thicker filter blank, and resonate this crystal element
on one of its overtone modes.
What VHF crystal filters allowed one to do is use these filters at the 1st
I.F. The significance here is twofold. First of all, single
conversion receivers with a high I.F. and good selectivity could be
designed. Second, in a multiple conversion design, these same crystal
filters, because of their selectivity, are able to have good 2nd image
rejection (2 X 2nd I.F.)
If you look at the specifications of todays roofing filters (the filters
that follow the 1st mixer in a receiver), you will notice that a 910kHz
rejection spec is given. The reason for this is because the 1st I.F. is
always at some high frequency between 40 and 120MHz, while the 2nd I.F. is
almost 450, 455, or 460kHz (image response would be at 900, 910, or 920kHz
respectively).
After this is all done, the intermediate frequencies (I.F.) have to be
chosen very carefully, with spur chart analysis of the various mixers for
inband spurs (this had already been done by many manufacturers, so this step
can be pretty much considered done).
To make a long story short, upconversion eliminates the need for multiple
bandswitched coils in the LO section of a radio, because with a high I.F. a
synthesizer can be designed that will operate over less than a 1 octave
range.
The significance here is that you can now operate the Varactor diodes in the
VCO portion of the synthesizer over the most linear part of their range.
This allows you to have a VCO that has a K/V characteristic of less than 2:1
over its tuning range. When designing a loop filter for a synthesizer, the
K/V characteristic is one component of the design equations. This is very
important, so that the settling time of the synthesizer will be relatively
constant over its entire tuning range. With synthesizers designed for
frequency hop communications or digital modes, bandswitched Varactor diodes
are sometimes used for a small portion of the tuning range. Other times, a
dual mode loop filter is used, with a wide bandwidth loop filter used while
the PLL is acquiring lock, switching to a narrow bandwidth loop filter once
the system acquires lock. This improves the close-in phase noise of the
synthesizer, thus minimizing reciprocal mixing effects.
Today, settling time is not as much a factor as it used to be, since the
advent of Fraction N synthesizers. This design opens another can of worms,
since you now have to deal with different types of spurs. One approach that
has been used is what is called a Modulated Fractional Divider. This type of
design translated the Fractional N mixing spurs further out from the LO
carrier. This way, these spurs can be more easily filtered.
As an example, consider a radio with a 1st I.F. of 70MHz. To tune from 0 to
30MHz with high-side injection from the 1st LO, the tuning range of this LO
will tune from 70 to 100MHz. The lower sideband range from the 1st mixer
will be used in this case for receiving. The only tuned circuits required
for good image rejection will be a low pass filter that cuts off at about
35MHz.
Since your image band will be 140MHz away (140 to 170MHz), this response
will be far down on the low pass filter's skirts, as long as this filter
provides good out of band attenuation. Good shielding is important here.
Mind you, this only covers the LO portion of the receiver. In order to have
good strong signal handling performance, a good 1st mixer with good IMD
performance is required. Suboctave input bandpass filtering also improves
IMD performance of the receiver, because by limiting the aperture of how
much spectrum space the 1st mixer is actually seeing, you are limiting the
integrated power that is being applied to this mixer.
I could go on and on, but I don't want to hog too much bandwidth. I have
touched on only a very small part of the design challenges that a receiver
designer faces. I haven't covered roofing filter bandwidth options (the
other limiting factor besides phase noise performance that limits what is
sometimes referred to as dynamic selectivity). The 2nd mixer is where much
of this limitation occurs.
Oh, then there is the AGC system design. Typically, a receiver with AGC
applied only to the I.F. system will overload at around 3000 to 10000uV. To
extend the dynamic range of the AGC system, AGC needs to be applied the the
RF / 1st mixer stages so that it takes over at a level where the I.F. AGC
"runs out of steam". If you are not careful when you design this part of the
receiver, the two AGC systems can oscillate. Drake '7 Line
owners...........think about how critical that AGC adjustment is on your
units. If it is not set according to Drake's specifications, the AGC will
oscillate when in the fast mode.
I hope this small bit of "scratching the surface" helps. I still haven't
even covered the demodulator design and the audio stage design. And then,
there are the challenges of designing a low noise power supply!
Thanks to all of the other posters, for all of your good information!

Pete

"ShutterMan" wrote in message
oups.com...
Good answers from everyone, thank you. In looking at the simplest of
tank circuits for radio, I just couldnt understand why you just cant
add more coil windings and different capacitance to increase frequency
coverage.....but it looks like its alot more complicated than that.
Thanks again.