Here is some great information about coax that relates directly to many of
the questions that come up here on our list. I must add, however, that the
Jefatec coax is not mentioned in this article, but the Jefatec Low-Loss 400
Flex is excellent coax. It is important to get the 400 Flex because it has
the stranded center conductor. -- Ron, K8HSY
***###***
Some Q&A About Coax for Your HF Station
By Jim Brown K9YC
Second Edition - October 2011
Q:
What do I look for when buying coax for my HF station?
A:
First and foremost, look for a manufacturer you can trust. Don't buy
un-branded or off-brand coax! Look for physical properties that match your
use - UV
resistant to stand up to sunlight, consider weight if it will feed a flat
wire dipole, look for "flooded" or armored coax if it must lay on the ground
and resist varmints. The most important electrical property is loss, and
loss below about 250 MHz is entirely the result of conductor resistance.
Q:
What do RG numbers mean?
A:
Not much. For all practical purposes, RG-numbers describe only the impedance
and approximate size. RG58 is a 0.2-inch diameter 50 ohm cable suitable for
very short runs. RG59 is a 75 ohm cable of about 0.25 inches diameter. RG8X
is about 0.25-inch, lower loss version of RG58, and RG6 is about 0.3-inch,
lower loss version of RG59. RG8, RG213, and RG214 are about 0.4-inch
diameter 50 ohm cables; 0.4-inch 75 ohm cables are called RG11. RG174 and
RG187 are
miniature (0.1-inch diameter) 50 and 75 ohm cables.
RG numbers were the original specification for coax during the military
buildup for World
War II, when not much happened above 30MHz (other than radar) and the only
commercial use of radio was AM broadcasting. In those days, RG numbers did
mean
something, and defined everything that mattered. That all changed in the
50s, with the
growth of FM and TV broadcasting, which brought about MATV (Master Antenna
TV)
receiving systems for buildings, and CATV (community antenna TV) receiving
and
distribution systems to serve entire communities. We learned how to reduce
UHF loss with
foam dielectrics, and to reduce cost and weight with lightweight foil
shields and
copperweld (copper coated steel) center conductors. A 1970 Belden catalog
lists 14 RG59
cables; their 2006 print catalog lists 52 of them. These cables differ in
many important
ways - braid or foil shields that may be copper, aluminum, or both, solid or
stranded
copper center, or copper coated steel center. Their outer jacket may resist
UV, or not.
Many are designed for use indoors, where their jackets and dielectrics must
not burn or
create noxious fumes (the real Towering Inferno was caused by burning
cables). Some have
beefy copper shields for use with broadcast video or in transmitting
applications, others
have thin foil/braid shields for use in those MATV and CATV systems.
Q:
Do I need cable with a "low loss" foam dielectric for my HF station?
A:
Not necessarily. At HF, all that really matters is big copper, both in the
center conductor and the shield. A foam dielectric reduces loss at UHF - it
provides no benefit on the HF bands, and it costs more! However - a foam
dielectric can require the use of a larger center conductor for a given
shield
diameter. That larger copper reduces loss, especially on the HF bands.
Q:
Published specs for many cables don't show attenuation below 50 MHz.
A:
That's because most coax is used at VHF and UHF, and because loss in coax is
relatively small at lower frequencies. For short runs on 160M and 80M, it
doesn't matter. But if you're running hundreds of feet, it can matter a lot.
If there's no spec for attenuation at HF, the spec to look for is the DC
resistance
(DCR) of the center conductor and the shield. Add those two resistances
together, and look for a value of 3 ohms/1000 ft or less for RG8, 7 ohms or
less
for RG8X.
Q:
Doesn't skin effect make DC resistance irrelevant?
A:
No! While skin effect concentrates current in the skin of the conductors, it
is the resistance of that skin that adds IR drop that burns power and
degrades
shielding. The R-value that plugs into that equation is the resistance at
the frequency of interest, and that resistance starts with the DC resistance
and increase with frequency! Bottom line - reducing the DC resistance of
that outer skin reduces loss and improves shielding!
Q:What's the relationship between size and loss?
A:
The characteristic impedance of cable depends on the ratio of the diameter
of the conductors and their spacing, and the loss is directly related to the
resistance. That means
we need big copper to minimize loss! Bigger coax can also handle more power,
but that's
not the only reason we use it - it also provides lower loss.
Q:
I'm only running 100 watts. Isn't RG8 or RG11 overkill for wire antennas?
A:
It depends on the length of your feedlines, the operating frequency, and how
much a dB (or two) is worth to you. Study the graph below, which shows the
loss for 100 ft of the best coax on the HF bands. If your feedlines are 100
ft or less, loss will be less than 1dB on 40M and below, so an RG8X like
LMR240,
Belden 9258, or the Wireman's CQ118 are a good choice. On the other hand, if
you're running 200 ft to a tri-bander on a tower, that smaller coax would
burn more than half of your transmitter's power on 15M and 10M! If you're
using a dipole to cover all of 80/75M, the increased SWR away from resonance
increases loss. Note also that these data are for very good coax cables by
major manufacturers. Off-brand cables are often made with much less copper,
and have greater loss.
999
Manufacturers Published Data, LMR400 and 3227 are measured data
Q:
What about copper coated steel and copper coated aluminum?
A:
For frequencies below 10 MHz, avoid copper-coated steel. At higher
frequencies, skin effect takes over, and all the current is in the thin
copper coating,
but at lower frequencies, the steel increases the resistance (and the loss).
Measurements of LMR400, which has a copper coated aluminum center conductor,
show no increased loss due to the higher resistance of the aluminum, even as
low as 1 MHz.
Q:
What about shielding?
A:
The shielding provided by coax depends primarily on two factors. First, the
resistance of the shield - the lower the resistance, the better the
shielding.
Second, the density and homogeneity of the shield - that is, a shield that
is more dense and more uniform provides better shielding. This is where
quality
of manufacture comes into play - cut into cheap coax and you'll find
relatively thin copper braid. For best shielding, look for coax that
combines a heavy
copper braid with a dense layer of foil.
Q:
I have a high power, multi-transmitter station. Can I use RG58 or RG8X
between my transceiver and amplifier?
A:
Smaller coax cables like RG58 and RG8X provide less shielding than larger
coax, thanks to the higher resistance of their shields. To minimize
inter-station
interference, use coax with a beefy copper shield for all cables.
Q:
How can I use 75 ohm coax for a transmitting antenna? Aren't all
transmitters 50 ohms?
A:
While most transmitters and power amps are designed to match 50 ohm loads,
most of
us use antennas that require a tuner to extend their bandwidth, or to load
them on
additional bands. That tuner will handle 75 ohm coax just as easily as 50
ohm coax. Some
antennas are a better match to 75 ohms, and if you're trying to cover a ham
band that is
broad on a percentage basis, you'll achieve lower SWR over greater bandwidth
within the
antenna system with 75 ohm coax. Also, for the same copper (that is,
resistance), 75 ohm
cable has about one half the loss of 50 ohm cable because the current is
about one third
less. Looking at it the other way, 75 ohm coax will weigh a bit less than 50
ohm coax for
the same loss, because it requires less copper. I have two 80/40 fan dipoles
up at 120 ft,
both fed with Belden 8213 (RG11 foam). In addition to reducing losses on
those bands,
these antennas work quite well on 30M, 17M, 15M, 12M, and even 6M. I've made
a
dozen transcontinental and KH6 QSOs on 6M with 100 watts using these
antennas.
Q:
Aren't RG8 and RG11 too heavy for use with flat-top wire dipoles not
supported at the center?
A:
The additional weight of big coax is certainly significant, and requires
that the construction of the antenna be more robust. My high dipoles are
supported
between tall trees that sway in the wind, so I use pulleys in the trees and
a 100 pound weight on one end. After several years trying various
constructions,
I've settled on #10 solid or stranded copper, egg insulators at each end,
and center insulators like the Alpha Delta or those sold by The Wireman. The
Budwig coaxial feedpoint attachment works electrically, but is not easily
rigged to handle the pulling stress on the wires. If you're going to use the
Budwig for this kind of antenna, you'll also need an egg in parallel with it
to relieve the stress.
Q:
How does moisture affect coax? What should I do about it?
A:
Moisture affects coax in two important ways. First, moisture accelerates
corrosion of the braid, which degrades shielding and increases loss. Second,
moisture
in the dielectric (the insulation between center conductor and shield)
increases dielectric loss. To keep moisture out of coax, all connections
should
be thoroughly weatherproofed.
Q:
How do I keep moisture out of coax?
A:
First, install connectors in a manner that minimizes the openings. If you're
using solder-type connectors, try to completely fill the holes in the center
pin and the shield. When installing cables outdoors, wrap every connection
thoroughly with Scotch 88 or 33, and cover that with a rubber mastic tape
like
Scotch 2228. Alternatively, tape the connection with Scotch 88 and coat that
with a product like Scotchcote. Prevent moisture from running down the cable
to the connector by putting a downward loop in the cable that forms a "rain
drip" below the elevation of the connector.
Q:
I've heard that coax degrades with age, causing loss to increase
drastically. Is that true?.
A:
As far as performance on the HF bands is concerned, it's another one of
those old wives' tales that, while based on a grain of truth, should not be
taken
as gospel. What's true are that 1) as copper braid corrodes, its resistance
can increase, which increases loss and degrades shielding; and 2) moisture
in the dielectric can increase dielectric loss. However
-
I recently acquired a lot of rather old coax from a neighbor who had become
a silent key, and when I tried to sell it, no one was interested. All of it
was good quality RG8, RG9, RG11, and RG213 from major manufacturers, and Mom
and Dad, who had both been through the Great Depression, taught me to be
frugal.
So rather than throw it away, I used it to make stubs, and study their
performance. So far, I've built more than a dozen stubs from this old coax,
and
with the exception of one piece that was obviously of poor quality to begin
with, every one of those stubs performs very well, including those made from
a length of RG9 (Belden 8242) that had been stored outdoors in a shed,
probably for a decade or two (see photo at right). This is very expensive
cable - the center conductor is #13 silver-plated copper, an inner
silver-plated braid
2068
shield, a outer copper braid shield, and a non-contaminating grey jacket.
That green stuff
on the outer braid is copper oxide, and even the braid without the green is
a dull brown.
Note the gap between the connector and the braid - the double shields make
the outer
diameter of this coax 0.424 inches, too big to fit in standard PL259
connectors. HF stubs I
built from this coax produced a very deep notch, indicating very low loss.
Why are my results so different from conventional wisdom (old wives tales)?
First, because
dielectric loss is primarily a factor at UHF, not at HF, so degradation of
the dielectric
doesn't matter. Resistance of the shield certainly is important, but my stub
data suggests
that the real world degradation is a lot less than predicted by those old
wives tales! Does
this mean that we shouldn't be careful to keep moisture out of coax? Of
course not -
moisture and corrosion are a bad thing. But if I had runs of good low loss
coax on my HF
antennas that were installed and trouble-free, I'd find other ways to spend
my money.
Q:
What's the deal with Commscope 3227 and 2427K?
A:
Several years ago, a large warehouse full of this cable on 1,000 ft spools
came on the market at a fraction of its normal cost as the result of a
telecom
bankruptcy. Some of that stock is still around, selling well below its
normal price. It's excellent coax, equivalent in quality and construction to
the
best of Belden and Times. The two types are identical, except that 2427K ,
the plenum version, has a higher loss dielectric. The differences only
matter
above about 300 MHz. The center conductor is #10 solid copper, so it's not
very flexible. 2427K is only 0.35-inch o.d., which makes it nice for coaxial
ferrite chokes.
Q:
You don't talk much about RG58 and RG59. Why not?
A:
Primarily because smaller cables have more loss, thanks to the higher
resistance of their smaller diameter center conductor and higher resistance
shields.
The higher resistance shields also degrades shielding. RG8X and RG6 cables
are the smallest cables that make sense in a ham station. I recently bought
a mag mount antenna so I could work a network of 1.2 GHz repeaters here in
the SF Bay Area. The mount came with 12 ft of permanently attached coax that
burns 3dB (half of the transmit power) at that frequency! On 20 meters, 160
ft of a good RG58 will burn 3dB, 110 ft will do it on 10M.
Q:
I've heard about low loss 75 ohm "hard line" that is often given away by
cable TV (CATV) companies in scrap lengths. Is this useful in my ham
station?
How do I use it?
A:
The half-inch size of this cable has about the same low loss as 50Ù
half-inch hard line. It's a bit delicate - the center conductor is 0.12-inch
copper
coated Aluminum, the shield is aluminum, and the dielectric is foam, so it's
easy to kink it if you're not very careful. It's easy to fit PL259
connectors
onto this cable. VF depends on the dielectric used, and is typically on the
order of 0.84.
Q:
Can I use 75 ohm coax on a 50Ù antenna?
A:
Sure. For monoband antennas, we can take advantage of the fact that even
with a mismatched line, the impedance seen by a transmitter feeding a line
that
is a half wave long, or some multiple of half waves, will be the actual
antenna impedance. In other words, if the line is some multiple of a half
wave,
it essentially disappears. To use this property, you'll need to carefully
prune the feedline to the desired length. A good way to do this is to
connect
a short to one end and look for a short at the other end with an impedance
bridge like the MFJ259. If you need more length, simply add 50 ohm cable at
either or both ends.
mc-ref
Mfr data for some 75Ùcables, LMR400 shown
for comparison
Q:
How can I attach PL259s to half-inch CATV hard line?
A:
Begin by removing about 3.5 inches of the outer jacket, then, strip the
cable so that at least 1 5/8-inches of the center conductor is exposed. The
hard
part of this is that this cable often has a dielectric that is bonded to the
shield and to the center conductor, so it is very difficult to strip. The
only good way to do this is with a coring tool. My neighbor, K6XX, loaned me
his, a Cablematic CST500, which works quite well. It's available from US
sources
for about $75. You'll also want a basic tubing cutter to make the coring and
stripping go faster. Be gentle with the tubing cutter - it's easy to crush
the shield if you over tighten it. Once the center conductor is exposed,
you'll need to scrape the foam that is still clinging to it. Be careful not
to
scrape off the copper coating.
For each connector, you'll need two hose clamps just large enough to fit
over the cable.
Slide them onto the stripped cable, then the retaining ring of the
connector, then the
connector body, positioning it snug up against the end of the cable shield.
Now, solder the
center conductor, slide the retaining ring over the connector body and screw
it all the way
onto the body. Find a UHF barrel, mate it with the PL259, and carefully
tighten it with
wrenches. Next, find some copper braid about 3/8-inch wide and cut it into
two lengths of
about 1.5 inches. We'll use this braid to bond across the joint between the
connector shell
and the cable shield. Using good electrical tape, carefully tape the two
lengths of braid to
opposite sides of the joint, and wind multiple layers of the tape so that
the outer diameter
is about the same as the diameter of the cable, going all the way to cover
the connector.
Now, slide the two hose clamps over the joint, making sure that one is
around connector
body and the other is around the cable shield, and tighten them carefully.
Apply additional
weatherproofing to the complete joints after mating it to cables at each
end.
3174
For more great advice on using CATV hard line, including methods for pruning
to length
and alternate methods for installing connectors, see QST, Jan 2000, p91, and
websites by
N1BUG, N9ZIA, W9XT.
Q:
So if RG numbers don't mean anything, how do I know what coax to buy?
A:
First, stick with the major manufacturers - Belden, Commscope, Times, and
with the few cable vendors (Davis RF, The Wireman) who sell quality
private-label
cables that are well documented on their website. Study their data sheets,
and pay only for the specifications that you need. For transmitting on the
HF
bands, look for beefy copper center conductor and shields, and don't
consider any cable smaller than RG8X or RG6. Don't pay extra for a foam
dielectric
(unless it also gives you bigger copper), and don't buy a cable with
unpublished specs. "LMR400 equivalent" is not a specification, it's an
advertising
claim, and should treated with the same mistrust as a sideshow barker! Some
high quality coax cables suitable for use on the HF bands are listed below.
RG8X: Belden 9258, Times LMR240, Wireman CQ116, CQ118 are approximately
equal.
RG8: Lowest loss - Commscope 3227, 2427K, Times LMR400UF, LMR400; Belden
8268.
RG8: Slightly greater loss - Belden 8267, Belden 9913, Wireman CQ1000,
RG8 for stubs: Belden 8237 or 9251, Commscope 3227 or 2427K; Davis RF 213.
RG11: Lowest loss - Belden 8213;
RG11: More loss - Belden 9212
RG6: Lowest loss - Belden 8238, 8261
Q:
What are some good types of coax for my HF station?
A:
The table below lists a known good cables with full published specifications
from known good vendors. Costs listed are those I found with internet
searches
in Feb 2010 from reputable vendors. Shorter lengths are more expensive per
foot. Note that RG58 and RG59 are not on this list - they have too much loss
for long runs, and their shielding is not as good as the larger cables. As I
have time, I plan to add some RG6 cables to the list. For
longer runs, you should consider bigger cables like LMR500, LMR600, and
various hard
lines. Within each cable type, cables are listed in order of their total DC
resistance.
table with 8 columns and 24 rows
Manufacturer /number
Total DCR
Shield DCR
Vf
% Braid
Shield Mat'l
Center Conductor
Cost 1,000 ft
RG8 Cables Solid PE Dielctric
Belden 8268
2.4
0.7
.66
97%
2-Cu/Ag
#13 Cu str
$1,800
Belden 8267
2.9
1.2
.66
96%
Cu
#13 Cu str
$1,200
Belden 8237
3.1
1.2
.66
97%
Cu
#13 Cu str
$1,030
Davis 213
3.1
1.2
.66
Cu
#13 Cu str
$570
RG8 Cables Foam Dielectric
Commscope 3227, 2427K
2.3
1.4
.84
90% + Al
Cu/Al
#10 Cu
See text
Times LMR400UF
2.7
1.65
.85
Cu/Al
#10 Cu str
$1,090
Wireman CQ102
2.7
1.8
.84
Cu/Al
Cu/Al
$675
Wireman CQ106
2.9
1.8
.82
Cu/Al
#10 Cu str
$693
Wireman CQ1000
2.9
2
.85
Cu/Al
Cu/Al
$765
Belden 9913F7
3
1.8
.85
95%
Cu/Al
#10 Cu
$1,242
Davis Buryflex
3
1.9
.83
Cu/Al
#10 Cu str
$720
Times LMR400
3
1.65
.85
Cu/Al
#12 Cu/Al
$730
RG8X Cables Foam dielectric
Times LMR240
7.1
3.9
.84
Cu/Al
#15 Cu
$540
Belden 9258
7.6
3.3
.82
95%
Cu
#16 Cu
$560
Wireman CQ116
7.6
3.3
.78
Cu
#16 Cu str
$270
Wireman CQ118
8.2
3.9
.8
Cu/Al
#16 Cu str
$270
RG11 Cables
Belden 8213 (75 Ù)
3.7
1.1
.84
97%
Cu
#14 Cu
$1,030
table end
Q:
What coax should I use on a Beverage or K9AY loop?
A:
Neither loss nor impedance are important, but shielding and durability are.
This is a great application for good quality RG58 or RG59. Even better, use
coax that is "flooded" with a compound to resist moisture penetration,
especially for runs that may be attacked by varmints.
Q:
What PL259 connectors should I use? Are hamfest connectors OK?
A:
Stick with branded connectors from Amphenol. Buy PL259s with a silver plated
body - they are much easier to solder and cost only $2.74 each in lots of
100, only slightly more than nickel plating. Cost is about 10% higher for 50
pieces, another 10% for 25 pieces. 83ISP is the Amphenol part number.
Off-brand
and unbranded connectors often have poor mechanical fit, are often made with
cheap plating that oxidizes and/or makes poor electrical contact, soft
metals,
dielectrics that melt when heat is applied.
And don't trust vendors just because they advertise in ham magazines. About
six years ago
I restocked my parts bins with an assortment of DIN connectors from The RF
Connection.
As I used them to make patch cables for my radios, I found contact plating I
couldn't solder
to and dielectric that melted, even with my temperature controlled iron, and
even when I
did think I'd made a solder joint, it soon failed. When I called to
complain, I was told "Hams
are cheap." Last I looked, they seem to be selling the same junk.
Q:
I've heard that every connector adds a dB of loss. Is this true?
A:
FALSE! This is an old wives' tale that has it's grain of truth in the use of
junk connectors, which CAN introduce excessive loss, can overheat because
the center conductor is too small or goes intermittent, and so on. REAL coax
connectors do not introduce measurable loss at frequencies below high UHF.
As a test, I recently spliced together 12 100 ft lengths of low loss coax
that I had made up for a DXpedition, and measured the loss using HP
equipment up to about 500 MHz. The total loss in that cable, including 24
PL259s and 11
barrels was a dB or so less than the loss specified for the cable by the
manufacturer at 500
MHz. All connectors were Amphenols.
Dr. Ronald E. Milliman
Retired Professor of Marketing
|
