Topband: Problem with compression F connectors on Quad RG-6

[Tom W8JI]

> I'm interested in your comment about LMR400 and soldered braids, Tom. I
> understand the point about shield current flowing on the insde of the foil
> or braid closest to the center conductor, but if the shield connection is
> faulty, oxidized, or has high resistance, then it seems to me the outer
> braid (presumed to be soldered to the PL259) would carry a potion of the
> return current. It would be an interesting physics problem to work out the
> relative return currents carried by a coaxial cable with two concentric
> shields, each having non-zero resistivity, or a thickness comparable or
> thinner than the skin depth..
>
> I don't recall the skin depth at 1.8MHz, but my guess is it's probably
> longer than the thickness of the coating on aluminized mylar. I don't know
> what the foil thickness is on LMR400.


2 MHz skin depth of copper is .0018 inches.
2 MHz skin depth of aluminum is .0023 inches.

The LMR400 type cable I have is about .006-.008" foil. I chemically stripped 
the foil off, and the foil was two wraps thick around the cable.

LMR400 or any cable, in sensitive applications, requires a solid bond to the 
shield that carries the vast majority of return current. In the case of 
almost all cables on HF and higher, that is the innermost foil. Of course it 
is different at audio or lower frequencies.

One common connector problem comes from not forcing the woven shield tight 
against the foil at the connector, or having the foil or woven shield 
tarnish or corrode. The path to the inside of the foil is out on the braid 
to an eventual contact point, then back on the outside of the foil to the 
foil edge. At the edge current can go inside.  This is like adding 2X the 
length of the path to the connection point in overall shield connection path 
length.

 (Current can also "get in" across the edge of a longitudinal seam, if the 
seam's overlap is insulated. The problem with that is the seam can kill UHF 
performance.)


If you solder to the shield of LMR400, and put it on a network analyzer and 
measure the "stub" characteristics, many times (not always) it will move 
around as the cable is flexed. This is because the soldering heat contracts 
the dielectric, releasing pressure between the braid overlay and the foil. 
Now you have a crummy connection that changes electrical length of the 
connection to the "real" shield.

Even if you do things right, once the foil and braid develop an oxide layer 
the connection goes away. This can work its way out for several feet of 
cable length, really messing up a cable. This will not show with a single 
shield.

Cables with foil have to be installed and treated correctly. The more layers 
you add, the more careful we must be. Since the extra layers are pretty much 
meaningless, the best practice is to avoid them. Use a good shield against 
the center and connect to it at the connector.

> I also wonder about the ability of a thin foil shield to carry Amps of RF
> without appreciable loss!

RF flows on the surfaces nearest opposing (differential) current flow. You 
have the small outer surface area of the center conductor carrying the same 
current as all of that wide area of the foil. The shield has a great deal of 
surface area compared to the center.


73 Tom