[TowerTalk] Coax beyond cutoff (waveguide region)

Healy, Rus RHealy@mdsroc.com
Tue, 30 Dec 1997 11:01:18 -0500

Del, KL7HF, wrote:
Very seldom is coax used above 6 Gigs or even at 6 Gigs
other than jumpers. That is into the waveguide region. 
The data supporting solid conductors over stranded due to diode
joints where corrision occurs, is certainly far more important
at the microwave frequencies, due to wavelength. However, it
lends credance <sic> to the threads we have had involving proper sealing
of coaxial cable ends to attenuate corrosion. 

Del certainly has a point when it comes to long runs of cable. However,
the cutoff frequency (transition from coaxial-cable behavior to
waveguide-like behavior) of coaxial cable varies with size and
frequency, among other things, including shield geometry and dielectric
characteristics. Then there are the practical issues of cable routing
(sizes and locations of bends) and what it does to a cable that the
manufacturer says is beyond cutoff.

Many of us who are active on the microwave bands use Heliax and
Superflex cables that are beyond cutoff according to Andrew literature.
For example, LDF4-50A Heliax (1/2-inch, non-Superflex cable with foam
dielectric) is rated for use up to 10.2 GHz, but works fine at 10.4 GHz
for everyone I know who uses it there. Interestingly, 1/2-inch
Superflex, FSJ4-50B, is rated to 10.8 GHz (as I recall). However, it has
measurably more loss than LDF4-50A at 10.4 GHz for even reasonably short
lengths (a few feet). For comparison, 1/4-inch and 3/8-inch Superflex,
both excellent cables for 10-GHz use, are rated to more than 11 GHz.

An example of where the economics of the coax-versus-waveguide argument
smacks hams in the face is for reasonably long runs (100 feet or so) at
5.76 GHz. At this frequency, it's quite possible to use 7/8-inch Andrew
LDF5-50A, which is beyond Andrew's frequency cutoff specification. It
works quite well, and costs much less than waveguide (when you can find
it) for this frequency. In fact, surplus 7/8-inch Heliax is widely
available, but waveguide, flexible or rigid, is rare and costly on the
surplus scene in most parts of the US.

Anyone who doubts the value of using such small cables at these
frequencies might be interested to know that 1/4-inch Superflex with
Andrew N connectors has substantially less loss at 10.4 GHz than even
high-grade, double-shielded RG-213-type cables with Amphenol N
connectors. I've measured 2.21 dB and 4.5 dB, respectively, for 10-foot
runs. (Both cables are operating below cutoff.)

As for the stranded-versus-solid debate, I believe that intermodulation
is the only significant RF concern, and here *only* at the multikilowatt
level with combined transmitters sharing a single feed line. That's
certainly what Andrew measurements pertain to. But even this is a less
significant effect than things like undesired mixing in transmitters,
which causes spurious reradiation or wideband noise at sites with
colocated, inband receiving equipment. Noisy antennas (those with bad
mechanical connections) and guy-wire attachment points even come before
coaxial-cable intermodulation.

In other words, for ham radio purposes, the difference is in the noise.

--73, Rus, NJ2L

FAQ on WWW:               http://www.contesting.com/towertalkfaq.html
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com
Search:                   http://www.contesting.com/km9p/search