Andy Wade wrote:
Ian White wrote:

Yes, Andy, it is to make a ground mat for a radio aerial...

How did I guess...?

well, four aerials actually!

Ah, phased verticals (?)

That's right. Four verticals at the corners of a 10m square, phased
together to create a directional beam that is instantly switchable in
any of eight directions.

Part of the fun of playing with aerials is that it's a mixture of
electronic engineering and DIY (though I must admit there's not a lot of
scope for paperhanging).

1. Solid, not stranded
That is an absolute requirement - please take my word for it. (At
radio frequencies, the resistance of stranded wire becomes much
higher when the strands corrode, as they inevitably will over the
years. Solid wire is much less affected.)

Hmm, I've not come across that one before. It used to be quite common
practice to use 7/36 stranded bare copper for (MF & HF) radio earth
systems. Presumably you're suggesting that the inter-strand corrosion
increases the RF resistance somehow - yet the current flow is
longitudinal and surface corrosion of a single wire will also increase
the RF resistance. Further explanation and scientific evidence would
be of interest.

Sorry, this is getting a bit OT, but it might be interesting...

Something called the 'skin effect' makes radio-frequency (RF) currents
flow only on the outside skin of a conductor. Skin depth decreases with
frequency (square root of). No current flows within the wire, so its
current-carrying cross-sectional area is much smaller than its physical
cross-section. Therefore its effective resistance is higher at RF than
it is at DC or 50Hz.

The difference between solid wire and stranded is that the current has
to hop from strand to strand in order to flow straight down the wire,
because the lay of the strands is slightly spiral. There is only a small
mechanical force holding adjacent strands together, so the contact is
never very good and so the RF resistance is just a little bit higher
than for the same-size solid wire.

Exactly the same happens in braid, as used for the outer shielding of
coaxial cables. Because the lay of the strands is more diagonal, more
hopping is required so the RF resistance is a little bit higher still.

Now add a thin layer of surface corrosion, which has a much higher
resistivity than metal. In the solid wire, the current simply pulls back
into the metal skin beneath, so the RF resistance is almost as low as
before. But in the stranded wire, the individual strands are forced
apart and the RF resistance can become a lot higher.

The most dramatic demonstration is what happens when the outer plastic
jacket of a coaxial cable is cut and rainwater gets in. When the braid
corrodes, the strands are forced apart and the RF resistance increases
dramatically. The braid also wicks the water down along the rest of the
cable run. If this has happened, the only solution is to replace the
whole length - it's shot.

This effect would be nothing like so severe for my earth mat
application; but given the choice of solid or stranded wire, it's
something I'd rather avoid. However, the D10 telephone wire might change
my mind, because the insulation is extra-strong and would protect the
wire much better.

BTDT - enamelled is a pain to scrape and connect, especially when
doing it about 200 times, outdoors at ground level.

But surely almost all 'enamelled' wire these days is solderable
polyurethane insulated and won't need scraping.

That only applies to 'instrument' wire on small reels. The main source
of enamelled wire on big reels still tends to be from motor winding
shops, and they often use a more modern version of the brown enamel
which is double-coated, temperature-resistant and tough as old boots.
(But the Dremel trick is duly noted - thank you.)



You can get 6491X PVC insulated conduit wire in solid (single-strand)
up to 2.5 mm^2. (But PVC is a poor dielectric and might also add
significant RF loss.)

Thanks for the correction - I'd thought it was all stranded, so that is
yet another possibility. No problem about the PVC, because the main RF
losses are in the ground that it's touching.


Or get some 7-strand conduit wire (say 10 or 16 mm^2) and strip out
the strands.
Because of the total length of wire involved, I'm afraid that would
be far too labour-intensive.

YIK, but it did meet the "cheap" part of the specification.

I may well do that for the eight "busbar" wires to which all the others
are soldered. For the other 120 wires, I'd prefer less preparation work
if at all possible.


--
Ian White