On Dec 8, 10:58 pm, "AI4QJ" wrote:
"Keith Dysart" wrote in message
On Dec 8, 3:15 pm, "AI4QJ" wrote:
In this example, we have transmission lines, not an antenna or antenna
coil.
The total phase shift is 90 degrees or 62.5 nsec.

Only with great stretching.

The 10 degree 100 ohm line contributes 6.94nsec,

Correct.

the 43 degree 600 ohm line contributes 29.86 nsec.

Correct.

But now think in the time domain for a bit.
29.86 nsec after the signal is first applied it reaches the
discontinuity. 29.86 nsec later the first reflection arrives
back at the start. 13.8 nsec later the first reflection from
the end of the 100 ohm section arrives back at the start.
It takes many more reflections of reflections before the
impedance at the input starts to look like a short.

Nowhere in here will you be able to find anything that
happens in 62.5 nsec.

The key point is that since 62.5 nsec represents 1/4WL at the
frequency of interest, if the line was 1/4WL long, you would
find things that happen every 62.5 nsec.

The values returned back to the feed point are superimposed
in the time domain, adding and subtracting, but the whole system should
(*does*) still act like a 1/4W stub at 4MHz.

There is no doubt that all the variants discussed do share one
property with a 1/4WL open stub; the input impedance after a
long settling time is the same. But other properties, particularly
settling time, are quite different.

A 1/4W stub could consist of
1..N series transmission lines of different impedances provided they were
cut to the correct lengths.

Quite true. And if your definition of 14WL stub was anything that
produced a short, I would agree that it is self-consistent
but that it is probably not too useful.

Remember, the system was *measured* to be 1/4W so the reflections
must cause a short every 125 nsec, no matter what is happening in between.

If I understand what your are saying, then NO. Consider the short
itself.
There is no delay what-so-ever.

If you are saying that along the transmission line there is a point
every 125 nsec where the impedance is 0, I would agree for
a constant impedance line, but if you change the impedance then
the points on either side of the discontinuity are not 125 nsec
apart.

The task is to use the math to verify the measurement. The smith chart and
phaser diagrams should work and are equivalent to using math. Time domain is
possible too but no way I would ever go to that trouble on usenet at least!
Or do you think the measurement was wrong?-

I have no disagreement with the resulting numbers.

But 46.4 + 10 does not equal 90. Subtracting (46.4 + 10)
from 90 does not yield a useful number, though if you
decide that the sum MUST be 90 for some reason, then
you do have the number that will do that.

....Keith