In article
,
Telamon wrote:

In article .com,
wrote:

Telemon I was going to Email this info but your addy
is bogus. I wanted to inform you of my test results
and to give you the few additional refferences I located.

A few comments are at the end.

To everyone else:
I won't be responding as I have grave concerns over the
new cyber stalking laws and really have less then no desire
to get into arguements with how relevent this is to SWL.
Take from it whatever you can use and ignore the rest.
The r2000swl and r2000swler at hotmail and at yahoo
accounts are dead.

I will be greatly shocked if I ever post again.
Too much risk and way too little reward.

---------------------------------------------------------------

In an ideal universe signals outside of a coax would not
effect those inside the coax. Sadly coax is not perfect,
and outside signals can and do intrude to the inside.
The name for this effect is "transfer impedance".
I will abreviate it to "TI" to save space. TI is real and
in some situations can cause all sorts of problems.
I posted on this topic back in Dec 05 and have been
trying to find more info and gain a better understanding.
TI is a measure of "shield effectiveness"
I will start with a quote from my Dec 05 post:
"From Fluke:
http://www.flukenetworks.com/us/

"Transfer Impedance - For a specified cable length, transfer
impedance relates to a current on one surface of a shield to
the voltage drop generated by this current on the opposite surface
of the shield. Transfer impedance is used to determine shield
effectiveness against both ingress and egress of interfering
signals. Shields with lower transfer impedance are more effective
than shields with higher transfer impedance."

Blue Jean Cables has a good simple article with chart.
http://www.bluejeanscable.com/articles/shielding.htm

Beldon has more info at:
http://www.belden-wire.com/Catalog/TechInfo/TechTransfer.htm"

----------------------------------------------------------------------------
--
----

Another Belden source:
www.magma.ca/~emccons/docs/beldenTiAndSe.pdf

-----------------------------------

Some patent discriptions for improved coax:
http://www.freepatentsonline.com/4477693.html
http://www.freepatentsonline.com/4187396.html
http://www.freepatentsonline.com/4414689.html

------------------------------------

One way it it is tested:
www.nema.org/stds/wc61.cfm

------------------------------------
A patent that shows one way to measure TI:
http://www.delphion.com/cgi-bin/view...=fstv&OUT_FORM
AT
=pdf

(My test setup is pathetic in comparision)
------------------------------------

From http://www.ct-magazine.com/archives/ct/0599/ct0599l.htm
"Transfer impedance

Drop cables are shielded cables with combination foil/braid
shields designed to operate in two electromagnetic environments
-one is the desired RF inside the cable, and the other is the
ambient over-the-air environment. Transfer impedance, Zt, is a
means of characterizing how well the shielding works, or how
electromagnetic energy transfers through the shield.

Transfer impedance relates the current flowing on the outside
shield surface (such as the common mode interference signal)
to the internal voltage it develops on the other side of the shield.
Figure 1 shows how the center conductor is susceptible to the
voltage on the inside of the shield produced by common-mode
current on the shield's outer surface. This voltage is the result
of a diffusion current through the shield."

------------------------------------

A good overview of the subject:
www.scte.org/documents/pdf/ANSISCTE782003IPSTP011.pdf

------------------------------------

A PDF with real math that discribes TI.
www.its.bldrdoc.gov/pub/ntia-rpt/01-382/appendix_b.pdf

-----------------------------------------------------------------

My own tests consisted of a RF "white" noise source, buffered
into a 24:1 step down transfomrer that I coupled to the cable under
test in a test jig that was 12' long and with 10' of coax( or , triax,
balanced pair, heliax etc) held in place with a 10' injection loop
that was spaced 1/2" from the test cable over the 10' test length.

My set up is very crude but even with my meager level of
instrumentation
I have found that common coax will allow significant engrees.
The most important thing isn't to try to find a perfect coax, but to
play
attention to the way we route coax. Running receive coax feedlines
in parallel with "noisy" cables or wires is a real bad idea. Keep
receive coax at least 5 feet from CATV, power, or telephone
lines. Cross noise cables at right angles. For significant noise
injection to occur with "normal" noise signal levels requires
the cables to be routed "close" and for more then a "few" inches.
"Normal", "Close" and "few" are relative.

"Close" as in a foot an means a length of inches can cause problems.

Think inverse square law.

TI is mainly going to be an issue with those who have "long",
say over 50M(150') coax runs, or who have noisy cable drops
fairly close to your coax. In most instances TI is only an issue
after you have solved all the other noise problems. Noise is
MUCH more likely to be directly radiated to your antenna or to
"creep" up the outside of your coax shield, get into the the antenna
and then be a problem. The same steps used to keep signals from
creeping up the shield and reaching the antenna will also reduce TI.

The most important step in stopping TI is to prevent or reduce
common mode currents on the coax outer shield. It is usefull
to understand the shield is really 2 conductors. The ouside and
the inside are seperate conductors. Common mode currents on
the outside can be coupled to the inside and if and/or when that
happens noise is added to the desired signal.

The use of feritte "beads" on coax(or the techniques shown
by Bryant at http://www.dxing.info/equipment/coax_leadin_bryant.pdf )
can greatly reduce the risk of common mode noise coupling
through the shield and adding noise, or creeping up the coax..

I found that in addition to beads, different feedlines can offer
much greater isolation. Mini heliax, hard line(as used by CATV),
Triax and twinax all have much lower levels of TI then coax.

While double or quad braid coaxes tend to be better, single braid
with a good foil can be much better for the 1~30MHz arena.

I could not inject any noise into heliax or hardline.

I had some TI allowing RFI from my 20" VGA monitor to get
into my system. The feeline ran "real close" to the monitor,
a very strong RFI source. By rearanging my shack so the monitor
and PC are well away from the feedline and patch bay I was able
to reduce the RFI to a level I can only detect with lengthy audio
FFT runs to see what I can no longer hear.

I hope this helps others to avoid some problems I have fought
for the over 25 years. While I really knew better, I had operated
like coax was a perfect faraday shield. That illusion stopped
me from resolving some minor RFI issues that proved impossible
to fix until I understood how the RFI was getting into my system.
RF can get through the shield to cause problems. Under "most"
conditions and at most locations I suspect it is a minor, at worst,
problem. The noisefloor on HF is so high that the minimal noise
added by TI, at most locations, will be undectable. It is something
to keep in mind, and a good reson to pick the route of receive coax
with some care and attention to TI, but to not switch to hardline
or heliax to solve "problems" that aren't there. In my opinion feritte
"beads" have great utility for many, maybe most SWL antenna
feedlines. The fact that they will help to reduce TI is nice, but not
thier
most usefull aspect. I think that noise conducted up the outside of the

shield getting into to the antenna is the bigger and more common
source of RFI. And ferrite can greatly reduce that noise level.

I was moved to post this after reading the post about "Coax to
coax noise transfer" and accepted that I had something that might
help someone. Hint: Try to find another route for you coax. Broadband
has lots of noise in HF and this could be a big problem. Do a test
before comitting to this route. Place your coax parallel to the
Comcast broadband line and terminate the coax with 50 (or 75)
ohms and see if you have any noise. I suspect you won't. I would
be very concerned about the comcast cable putting noise on your
shield that gets directly into the antenna. The use of feritte and a
~9:1UNUN to couple/"match" the coax from the antenna can help
isolate the coax, and any noise it might carry, from the antenna.

From my experiences it is much better to avoid noise then to fight to
reduce it.

I warned you all that this was long!

Thanks for your research and thoughts on this subject.

Generically transfer impedance refers a EM wave in this case traveling
from one medium, air outside the cable to the inside of the coax where
the dielectric constant of the inner insulator is the other medium but
specifically since the topic is coax cable it also refers to the outer
shield effectiveness.

No shielding is perfect and has an attenuation value associated with
it. Normally the attenuation value is large enough compared to the
signal in the coax that it is practically considered infinite but
apparently not with strong external interference sources around and
very small internal signals from the antenna in micro-volts. Here is
does not take much leakage to mess with a few micro-volts.

Look like the first problem NEMA had to solve was coming up with a
repeatable standard. The problem with the appears to have been
controlling the mode of propagation of the external EM wave energy as
this kept changing on them over frequency. The net result was
unrepeatable results. The standard became another pipe, which with the
coax cable kept inside and out side EM energy in TEM mode. Once this
was adopted the results became repeatable.

Here is an example of a fixtu
http://www.dcmindustries.com/products/TI-3000.htm

Now before anyone get to excited about this development this is a
specific standardized method of comparing cables and uses a specific
mode of energy to accomplish this comparative measurement. This is not
absolute as the cable can be placed in a different environment and will
not have the same effective shielding.

Yup, this is a good link on the subject, I looked at this months ago. I
don't remember if I found it myself or maybe you pointed it iut in the
past.

Beldon has more info at:
http://www.belden-wire.com/Catalog/TechInfo/TechTransfer.htm

Take a look at the sample report at the bottom of the DCM page
http://www.dcmindustries.com/products/TI3000brochure.pdf

No huge surprise here at the plot (red trace) showing the shielding
becoming less effective the lower you go in frequency. Any coax you use
will not shield as well at lower frequencies. They don't say what kind
of cable is used in this example but around 1 MHz its about 8 ohms so
if you have the strong AM station around as you complained about I see
how the coax would not be as effective against it as another source at
a higher frequency. The shield is only so thick and the skin depth
increases with lower frequency.

In any event on the Belden link the plots of those coax cables ends at
5 MHz but you can see the value of the transfer impedance going up into
the single digit ohm range so it looks comparable to the DCM example.

The problem with your crude test setup is keeping the EM energy outside
the test coax (your noise source) is that the mode of coupling will
change over frequency and change your results. Not only that but
changes in the room around it will change your results. This is the
first problem the standards committee had to solve.

Yes, your advice to limit the coupling to a nearby CATV cable to the
radio receiving coax is all related to mutual inductive coupling.

I learned two things from you bringing up this subject on the news
group. At AMBCB frequencies coax shielding is an order of magnitude
worse than I ever expected. Number two is a reason ferrites along a
long run of coax can help against intrusion of a AMBCB signal.
Previously I thought that ferrites on the cable ends would be effective
but in the middle of a run not helpful.

But note that as soon as you are above a few MHz the shield on most
decent coax looks like a short and not much will get through it with an
attenuation factor in the 90 to 110 dB range. The surprise for me was
the drop in effectiveness in the AMBCB range. I can now see that extra
shielding could be warranted if you have a AMBCB station interfering
with your reception.

I just re-read the Belden page and want to point out these concluding
statements " To determine how these two regions interact with one
another, you need to know how an electromagnetic signal in one region -
with its associated voltages and current - couples to the other. The
transfer impedance gives this relationship."

The whole exercise on this page refers to a specific coupling mode where
the transfer impedance does put a value on that particular relationship
and it is valid for that mode but that coupling mode may or may not
exist in a real world situation so they follow up with "Currently, no
ambient models are sophisticated enough to be used for broadband system
applications. However, simplified models can be used to help analyze, if
not quantify, interference problems." Yeah it does that.

--
Telamon
Ventura, California