Ladder filters don't have as long an engineering history, when you try
for a wider filter they tend to be asymmetrical, and they are very
dependent on crystal series resonant Q. For narrow simple filters the
crystals can all be the same frequency but that's not true for better
response shapes, there the crystal frequencies need to be shifted. That
can often be accomplished with series capacitors.
Lattice filters use both the series and parallel resonant frequencies of
each crystal and each crystal has to be on its proper frequency. There
is a little bit of balancing needed if full lattice, e.g. crystals are
used in matching pairs. There's a balancing transformer needed. Lattice
filters have been studied probably half a century longer than ladder
filters and it the low frequencies where lattice filters were first used
for telephone multiplexing filters, ladder filters would not have
produced a useful bandwidth. At 4 MHz its hard to get a ladder filter
wide enough for SSB, though CW filters are falling off the log simple to
make work.
Its easier in a lattice filter to get the steep skirts than in a ladder
filter since both resonances of the crystals are in use. Super square
frequency response curves give better adjacent channel rejection, at the
cost of far more severe ringing in the presence of transients, like hard
keying, line noise and lightning static. In the minimum element bandpass
filter the poor time response and fine frequency response are intimately
linked. To improve the time response requires sacrificing shape factor.
Its not hard to toss in a few extra crystals in a ladder filter to make
the trade off much less severs. A couple years ago there was an article
in QEX about 14 pole ladder filters.
In the lattice filter, ultimate rejection is a product of balance in the
transformer and matched stray capacitances of crystals. In the ladder
filter, ultimate rejection is the ladder attenuator of the shunt
capacitance of each crystal being the series element and the capacitor
to ground being the bypass element. Wide filters and higher frequencies
make those two capacitors similar in size so the ultimate rejection is
compromised. Though it is possible to purchase crystals with smaller
elements or smaller metalization patches to reduce the C from the
crystal or to add a shunt inductor to tune out some of that "holder"
capacitance.
I have a ladder filter near my work bench, that has been there probably
20 years, that allowed a panadapter to sweep at least 8, maybe ten times
faster than the same crystals configured for the same bandwidth as a
flat topped SSB filter. The difference is all in the design applying
modern filter design criteria.
73, Jerry, K0CQ
--
FAQ on WWW: http://www.contesting.com/tentecfaq.htm
Submissions: tentec@contesting.com
Administrative requests: tentec-REQUEST@contesting.com
Problems: owner-tentec@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|