I disagree that the computation techniques were not up to the task of
designing good filters, the math that's used by today's computers was
already developed. Its all summarized in the 60s handbook on filter
design by Zverev. Including the lousy time responses of Tchebychev
amplitude response curves. But the decision makers relied on the theory
that the human ear is not sensitive to relative phase of harmonics,
hence the phase destruction that destroys digital data was not harmful
to communications. They neglected the effect of noise pulses on ringing.
Crystal filters with fancy response curves and SSB bandwidth are not
practical (but not absolutely impossible) at 455 or 500 KHz. The series
and parallel resonant frequencies are too close together for convenient
filter construction. Though Ma Ball used crystal filters at lower
frequencies for telephone multiplex filters. Collins also sold tons of
mechanical filters to compete.
Probably the biggest reason for using a Tchebychev amplitude response
curve is that you get the best arbitrary amplitude response (flat top,
steep skirts) with the fewest resonant elements and that makes for a
more economical filter, whether crystal or mechanical. But when the time
response is ignored the resulting radio is not useful in all weather and
static conditions.
The better ham receivers of the early mechanical filter era, including
the R-390 relied on many carefully tuned IF transformers and though they
didn't achieve the perfectly flat top and steep skirts response curve,
they often performed better in the presence of static and pulse noise
that poorly designed mechanical and crystal filters.
The ladder filter used by Tentec tends to have a little poorer shape
factor than the true lattice filter, but is easily designed by properly
selecting the coupling capacitors to have a very good impulse (time)
response and to not ring as the result. Eons ago, I was playing with an
APA-10 panadapter and wanted to add selectivity with a second IF. I
built a 5 or 6 crystal ladder filter with equal valued coupling
capacitors which would result from a near Tchebychev approximation. I
could only sweep it at 10 Hz sweep rate because it rang too badly. I
redesigned it for a Gaussian amplitude response which has very rounded
corners and could sweep it at 60 Hz or higher without ringing. Same
crystals, just different and unequal coupling capacitors.
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
|