I believe this is a good test and conclusion. I can only add two small
things.

1- the impedance gets further from 50 ohms, the 269 MFJ ( Mine has this
characteristic) shows more and more error. Try paralleling four 50 ohm
loads (with your Tees) or measuring a good 200 ohm load. The 269 shows
considerable imaginary at 4:1 real (:-( I think I have heard that the 259
is better.

2- The constant reference to the "pin" being much less than a 1/4 wave is
not the correct focus. It is the impedance (primarily inductance in this
case) which is the factor of interest relative to the frequency. What's
that old rule of thumb? One nano Henry per inch?? One nano Henry per 1/10
inch... Been off that bench too long.

73, Steve, K9DCI



"Dave Platt" wrote in message
...
If I can find a scrap BNC male connector, and make a shorting-plug out
of it, I'll run the coax-and-T experiment I suggested, and post some
actual numbers for the systems's behavior at those frequencies I can
coerce out of my MFJ-269.

OK, I ran the test, and the results are pretty much as I had
anticipated. Although the test arrangement and gear isn't
lab-standard, I think it's good enough to confirm the basic principles.

Test equipment: MFJ-259 antenna analyzer, with an N-to-BNC adapter.

Transmission line and load: 12' length of RG-58 coax, with a 50-ohm
Ethernet terminator attached to the end.

Shorting insertion: an additional 6' length of RG-58, with a couple of
BNC "T" adapters, which can be inserted between the analyzer and the
T-line-and-load. If the shorting plug is inserted it'll be 6' from
the analyzer and 12' from the load.

Short circuit: a male BNC plug, with a short inserted between center
pin and shell as far down inside the shell as possible.

First test: check quality of T-line and load, using only the 12'
section of RG-58 between analyzer and load. Note that the Imag(load)
numbers do not indicate the sign of the reactance... the MFJ won't do
that, alas.

Frequency Real(load) Imag(load) Indicated SWR
2.5 52 4 1.0
5 58 6 1.2
10 57 7 1.2
15 49 1 1.0
20 56 6 1.1
50 54 2 1.1
144 51 4 1.0
166 59 8 1.2
440 1.4

Impression: either the RG-58 or the Ethernet terminator isn't
exactly 50 ohms (not unexpected) or the MFJ's calibration isn't
perfect (likewise) but the figures are good enough to let us draw some
reasonable conclusions.

Second test: insert the 6' RG-58 and its T connectors between the
analyzer and the 12' section. See how much this additional length of
line, and the parasitics of the T connectors, affect the load and SWR.

Frequency Real(load) Imag(load) Indicated SWR
2.5 54 4 1.1
5 59 5 1.2
10 50 10 1.2
15 49 1 1.0
20 51 8 1.1
50 48 5 1.1
144 50 4 1.0
166 40 2 1.2
440 1.1

Impression: the measured values change a bit, but the SWRs are close or
identical (save for the 440 measurement). The additional length of
line, and the parasitics from the T connectors, are shifting things
around a bit (especially at 440)... not unexpected.

Third test: connect the shorting plug to one of the T connectors. See
what "pinning the line" does to the analyzer's view of the load.

Frequency Real(load) Imag(load) Indicated SWR
2.5 0 6 31
5 0 12 31
10 1 26 31
15 1 43 31
20 2 70 31
50 0 44 28.7
144 7 33 15.4
166 166 353 13.0
440 5

Impression: yeah, that looks like a short circuit, transformed via a
feedline. Nothing close to a 50-ohm-resistive load shows up at any of
the test frequencies.

Fourth test: disconnect the 50-ohm load from the end of the 12' line,
creating an abrupt change in the load impedance. See what this does
to the figures from the third test - how much of the antenna load
change gets back "around" the short?

Frequency Real(load) Imag(load) Indicated SWR
2.5 0 6 31
5 0 12 31
10 1 25 31
15 1 43 31
20 2 70 31
50 6 44 29.0
144 6 30 18.5
166 133 317 14.1
440 5

Impression: changing the antenna load from 50-ohms-resistive to
near-infinite made a slight change in the impedance seen by the
analyzer, but not much at all at any frequency. The presence of the
short circuit 6 feet from the analuzer is still dominating the load
that the analyzer "sees".

Special-distance test: with the short, and the 50-ohm load both in
place, sweep the analyzer frequency around the ranges at which the
analyzer-to-short distance is around 1/4 and 1/2 wavelength. See if
we can "see past" the short under these conditions.

Measurements: with the short circuit in place, and a 50-ohm load at
the end of the 12' line, an impedance peak (Z1500 ohms) is noted when
sweeping between 32.14 MHz and 32.94 MHz. An impedance minimum
(Real(load) of 2-3 ohms, Imag(load) of 0 ohms) is noted between 65.7
MHz and 65.85 MHz. The indicated SWR remains high (22 or above) at
all frequencies between the impedance maximum and impedance minimum...
there's no point at which the load resembles anything like "50 ohms
resistive, little reactive component".

Removing the 50-ohm load, and thus open-circuiting the antenna
feedline, has no significant effect on the impedance as seen at the
quarter-wavelength maximum, at the half-wavelength minimum, or at
points in between.

Impression: we cannot "see past" the short circuit, no matter whether
it's a quarter-wavelength from the transmitter, a half-wavelength, or
some distance in between these two. The analyzer "sees" only what
would be expected for a short circuit, transformed by 6' of coax.

Short-circuit quality test: measure impedance of the shorted BNC plug.

At 16 MHz and below, the MFJ-269 shows it as 0+0j. Above 16 MHz, some
reactance shows up... 2 ohms at 28 MHz, 3 ohms at 42 MHz, 4 ohms and
55 MHz, 8 ohms at 112 MHz, 10 ohms at 144 MHz, 12 ohms at 169 MHz.
The inductance of the shorted plug itself, and/or the length of the
N-to-BNC adapter on the analyzer, is having some effect. SWR remains
unreadably high: 31 up through VHF and 5 on 440. It's not the
highest-quality "short circuit" in the world, for certain, but at HF
and VHF it's close enough for our purposes here.

So... what do I conclude from this?

I conclude that at HF and VHF frequencies, if you accidentally short
your transmitter-to-antenna coaxial feedline (with a pin, nail, or a
loose strand of coax braid which "gets loose" inside a connector),
you're going to present your transmitter with an unrealistic load
(high, low, or nastily reactive) and that the power flow up the
feedline is going to stop at the short circuit.

Since no "short circuit" in the real world is going to have 0+0j
impedance, this isn't *absolutely* true, but you can probably consider
it to be *practically* true and correct at these frequencies, with
this sort of shorting.

I further conclude that with respect to the above, there's no 'magic'
about shorts which happen to occur at a quarter-wavelength distance
from the transmitter. In this situation, the transmitter will 'see'
something which behaves like a quarter-wavelength shorted stub... the
load further up the transmission line at the antenna remains
'invisible' to the transmitter.

At high-UHF and SHF/microwave frequencies, where a direct "short" is
likely to be a significant fraction of a wavelength, then the
additional reactance of the "short" will certainly affect how the
undesired connection affects what the transmitter "sees". It may
hurt, or help (helping to "tune out" reactance from the antenna
itself), or may be completely invisible. Allison is entirely correct
on this point, under these particular conditions.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

Even at 50 KW?
73, Steve, K,9.D;C'I

P.S. I suspect it is air line, no?
_____________

Yes. In the case of WJR, it was a length of 1-5/8" OD air-insulated rigid
line (20 foot standard lengths plus a custom length).