Hi Brad,
> The nearest AM BC stations are in Austin, about 40 miles south of us.
> I don't think they were overloading the preamps I tested. It's also
> interesting that some, although not all, of the poorest performers had
> bandwidth-limiting circuits ahead of the active components. None of
> the broadband preamps with good IP3s sounded bad or stretched pulses
> excessively.
The overall accumulated power level from our antennas is what causes
problems.
Unless we are very close to a station (or stations) that reduces field
strength at night, the overall power level increases dramatically at night.
Even in a very rural location without any close BC stations active at night,
I measure in the 1-volt range of peak signal level on my Beverages. Even a
very good amplifer requires a band-reject filter for the AM BCB. Even SW BC
stations have enough level to cause overload problems in "weak" amplifiers.
When we get a night without static, I'll measure the total accumulated power
level from my antennas. I do know it is enough to dimly light a minature 12
volt bulb after going through a 15dB preamp.
> As Tom has suggested, the pulse-stretching I observed is due to the
> group delay characteristics of the filter networks.
We could *never* hear the unevenness in group delay on 160 meters. I swept
several multiple section L/C highpass filters and the phase difference is in
the order of a fraction a degree over the passband-width of receiver
filters. Some receiver filters, on the other hand, have 90-180 degrees
phase error over the passband!! I'll measure the time delay errors sometime
in the future, but they won't be over 10-15mS because the maximum delay time
is less than that. That <15ms delay time is through 16 poles of narrow
crystal filtering.
We have to keep in mind in order for unevenness in group delay to be
audible, it has to produce a large *difference* in group delay (propagation
time) within the passband of the receiver filters. The time error between
tones we hear caused by uneven group delay must be in the audible range of
our ears (well over 0.1 milliseconds), and that time error has to be within
the passband of the receiver filter and our ears!
73 Tom
|