On 11/2/18 9:35 AM, Richard (Rick) Karlquist wrote:
On 11/2/2018 8:29 AM, jimlux wrote:
Tungsten is 3 times the resistivity of copper, and results in about a 5%
Whenever skin effect is involved, the RF resisitivity
is proportional to the SQUARE ROOT of the DC resistivity.
Therefore, the RF resistance of tungsten is only SQRT(3) or 1.8
times as much as copper.
yes.. that's true.. the skin depth gets deeper with higher resistivity,
(as sqrt(rho)) so you wind up with a rho/sqrt(rho)
This only applies if skin depth is << wire diameter (say, wire diameter
is >8-10x skin depth.
For 10 MHz, skin depth is about a mil, so unless you're using AWG30 or
smaller, probably not an issue.
Another example, aluminum has 58% more DC resistance than copper
but only 26% more RF resistance. These numbers happen to neatly
translate into 4 wire gauges and 2 wire gauges respectively.
For example, 4 gauge aluminum wire has the same resistance at RF
as 6 gauge copper wire.
And, it's handy to remember that 10 gauges is 1/10th the area, 20 gauges
is 1/10th the diameter. and AWG10 is 0.1" in diameter.
For DC resistance
3 gauges is twice/half the resistance (just like dB!)
This calculation doesn't apply to magnetic alloys like nickel, iron, or
stainless, because the permeability affects only the RF resistance,
so there is no simple relationship.
Well, the skin depth goes as sqrt(rho)/sqrt(mu) If you have a good
number for mu (see below) you can calculate it. Non-magnetic stainless
steel is, unfortunately, not mu=1, it's just "mu is small compared to iron"
The real challenge is finding a good permeability number for some random
wire alloy - not only does the resitivity of steel vary a lot with
carbon content, so does the mu, and both vary with frequency too.
If it's a critical thing, the approach is "measure it"
We spent some time at work attempting to characterize the RF properties
of steel tape measures - they make handy deployable antennas -
ultimately, we wound up just bounding the problem and showing that in
the worst case, it wasn't "too bad".
just like antenna traps, *measuring* the RF properties of some arbitrary
"thing" that's not a discrete component can be challenging.
73
Rick N6RK
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|