> Real world transmitters always have limiters. Those DO change
> flatness. Just look at the EIA/TIA specication for testing
> transmitter pre-emphasis. The test is not run at system deviation.
> It is not even run at 60% of system deviation. It is run at 20% of
> system deviation. [that's +/- 1 KHz deviation for 5 KHz systems]
>
> Run the test at 20% into a modulation analyzer and you get a nice 6
> dB per octave line from 300 to near 3000 Hz.
>
> Run it again at higher deviations and see what the limiter does to
> your nice straight line - the pre-emphasis curve hits the limiter at
> progressiviely lower and lower frequencies as you increase the
> deviation. The result of this fact is going to alter the audio
> characteristic going through your repeater. It's one reason why
> level setting is so important. It is also a significant reason why
> it's difficult to get a repeater to sound like simplex.
That agument loses a lot of its bang when you consider that the audio was
already limited POST-PREMPHASIS in the user's radio. Assuming the
preemphasis curve in the user's transmitter is the same as the
deemphasis/preemphasis curve in the repeater itself (i.e. in terms of slope
or equivilent RC time constant), then the repeater will not alter the
end-to-end response.
> Another post suggested checking the frequency response of your
> repeater. Definitely - do that. Try it a various deviations. You
> may be surprised at how ugly it gets.
Seeing frequency response change as you get into limiting is completely
expected; "ugly" seems rather harsh. If you don't see an apparent "knee" in
the frequency response plot where there is a transition from positive slope
to zero slope as you approach system deviation, then the limiter isn't
working!
Bottom line: limiting after preemphasis results in a reduction in the
noise-limited dynamic range at higher frequencies; that's a natural
byproduct of a process which originated in the user's radio. The repeater,
following the same preemphasis/deemphasis curve as the user's radio, has no
further affect. Any variation or degradation in frequency response, S/N,
THD, etc. as the audio passes through the repeater is solely due to
imperfections in the design, implementation, or medium, not the
preemphasis/deemphasis process within the repeater since the two are
receiprocol.
--- Jeff
