Well, this E-mail program still doesn't do snippets, and I long ago gave up
trying to insert formulae, graphs, and the like in Email. Much less trying to
rember whether a given citation was in the second or third edition of
Terman's Radio Engineers Handbook. That said....
K0CQ disagrees, eloquently, with my strictly non-technical comments. I agree
with some of what he says, and have reason to disagree with other parts.
It's true enough that many filters, not just Chebeychev, severely distort
their pass band. Jerry believes this distortion does not affect signals
detected by ear because the ears are not sensitive to the relative phase of
different frequencies.
It was established (circa 1962) that a normal human ear could decode normal
speech phase shifting a monotone; the object being to design an effective
hearing aid for persons with profoundly frequency restricted hearing.
Obviously then, the ear can detect phase differences in a single frequency
arriving at one or both ears at the same time.
However, it was also found that the process was much more efficent when two
or three closely spaced tones were used. If my aged memory files are correct,
the frequencies were 880, 910, and 940 Hz, the phase shifts were less than 30
degrees, and the original speech input was separated into three frequency
bands; each band then phase shifted one of the tones. This was said to
provide near normal comprehension levels on the test subjects; albeit at
great expense and complexity.
While I have no further data on that subject it's been my experience that the
less overall distortion an IF chain has the better the intelligibility of the
recievers output - whether the desired output is digital or audio.
And yes, some well designed product detectors are very good indeed. Some of
the detectors in Amateur gear are quite good. But some that have shown up in
Amateur rigs seem to have been designed by either a first year engineering
student or a seriously hearing impaired person. Those things are far worse
for our purposes than a diode detector with a mininum level of BFO injection.
Let us remember that the OBJECT is the best possible intelligibility, and a
high end reciever should be at least as intelligible as a 1940's five tube
table radio. Many are not.
73 Pete Allen AC5E
|