Several years ago I wrote an article for Satellite Operator about
9913 coax.
It seems that Belden supplied me with a four hundred foot length
of 9913 that displayed a high SWR at a particular frequency. 100 watts
of RF would go in to the cable, and with a calibrated 50 ohm load at the
other end,
11 watts would emerge.
Doing a Distance-To-Fault test indicated a glitch at a particular
footage,
in terms of hundredths of a foot, along the coax.
Cutting off the piece somewhat beyond the glitch distance and
reconnectorizing duplicated the DTF point again. Upon examination of the
coax no fault could be seen.
I then dissected the coax completely around the DTF pinpoint
and could still see no glitches! As you know, there's a spiral-wrap
of dielectric around the center conductor, and then another tube of
semiconductor around that, and then the shield and outer covering.
Belden replaced the entire hunk, but I never came up with
a reason for the "spike" at a particular frequency and at that
particular point.
Nowadays I deal with Andrews LDF4 through 7 and all
the vagaries of sweep test failures. Just handling a piece of coax
or hard line, with no sharp bending or manipulation in the midst
will completely change its reflectometry profile. From the very same
piece of coax with the very same calibrated loads and shorts, the
observer would not be able to connect the sweep test graphs as being the
same
piece of wire.
Just bending a piece of coax gently will change its
reflectometric
curve characteristic. A tiny chunk of metal from dressing the end of the
cable finding its way in to the connector will thoroughly discredit the
entire system.
Now that we are incorporating 500 to 10KW EIRP power levels
at thousands of locations in the area, the reliance on good quality coax
craft is heavily stressed. A glitch in a coax line that is perhaps part
of a multi-frequency (850/1900MHz) sharing scheme can cause what
we refer to as a phase-jitter error. The coax's physical problem will
cause a partial reflection and that will in turn cause a phase angle
error
of a signal received through the antenna. Its corresponding downlink
channel will be affected, so much so that a five degree phase angle error
will cause a 10dB attenuation of receive, or downlink sensitivity.
While we rack our brains to come up with the final 1dB of coverage
the last thing we need is to learn that bad coax craft has desensitized
our receivers 10dB.
Lastly, do not trust 1/2 inch coax to convey QRO levels of power.
If you are running high levels of RF you might need to rethink your
scheme of how the main run is getting the signal to the load. There is
no
1/2 inch class of coax that will handle 10KW. Despite whatever
publications you may quote, it will not work in the real world. Having to
replace coax runs up towers and buildings time after time because of
internal material inconsistencies making hot spots and eventual radio
failures have taught me that everything, yes everything turns to cheese.
Hal Mandel
KA1XO
Secaucus
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|