[TowerTalk] Silver vs Aluminum vs Copper Re: OT: Inductor Calculator

[jimlux]

On 3/1/19 9:49 AM, Jim Thomson wrote:
> From: jimlux <jimlux at earthlink.net>
> To: towertalk at contesting.com
> Subject: Re: [TowerTalk] OT: Inductor Calculator
> 
> 
> <  And how to get something silver plated once out of high school
> <chemistry class I don't know?
> 
> < Silver plating doesn't affect the RF resistance very much unless the
> <silver is VERY thick - particularly at 2 MHz.
> <Copper skin depth at 2 MHz is 1.82 mils (46.1 microns)
> <Silver skin depth at 2 Mhz is 1.77 mils (44.9 micron)
> 
> <Resistivity is 1.59 for silver, 1.68 for copper
> 
> <QQ-S-365, ASTM B700 silver plating thickness is .05 mil, 13 microns
> 
> <Complicating this is there's usually a nickel flash/strike plate under
> <the silver, to prevent silver/copper migration, and nickel is magnetic
> <and lossy.
> 
> 
> <But for low frequencies, you're probably better off just making your
> <conductor 6% bigger in diameter and that will provide lower overall AC
> <resistance.
> 
> ##  personally, I hate the look of bare  copper, drives me nuts...so I  silver
> plate  all my  cu strap, coils, etc, etc.... with the ..cool amp  goop  from the
> cool amp company in Ore.   I also use it on brass  machine screws  like
> 10-32, and  .25-20, and also phosphor bronze  etc.  Also used it on contactors
> and relays,  fuse holders  etc, and  dc resistance drops like a rock.

Silver plating is much easier to solder than copper


> 
> ##  44.9 microns  =  0.001767717  inches.   Im sure I can pile on at least
> .001  inch..or more of  silver plating..esp with multiple applications.

The paper that was linked by another list member comments that plating 
conductivity is often (much) lower than the plain metal, because of 
stuff added to make the plating actually work.

And you really need 5 skin depths - 1 skin depth only gets the first 37% 
of the current - 5 skin depths is >99%.
There's a whole raft of icky things too about silver migration, weird 
alloys, and so forth.


> 
> ##  what about freqs higher than 2 mhz ??  

skin depth goes as the square root - 20 MHz is about 1/3 the thickness. 
When you get to microwave frequencies, or even VHF, UHF plating starts 
to make sense for things like resonant cavities (which are high Q).

There's a whole bunch of folks trying to make resonant cavities & 
filters with high temperature superconductors plated on the inside. 
Liquid nitrogen temps are fairly straightforward.



   A   .25 inch  OD  tubing coil  will handle
> 41 %  more  current, due to skin effect, on 7 mhz... vs the same  .25 inch OD tubing coil
> on  14 mhz.   The same  .25 inch OD  tubing coil on 1.8 mhz  will handle  a whopping 2.82
> X more current... vs the same .25 inch OD tubing coil on  14.350  mhz.....again due to
> skin effect.   It goes to the  sq  rt of the ratio of the freqs.

I don't know about "handle more current" - the RF resistance goes up as 
the square root of frequency, so the dissipation goes up as sqrt(f) too.


> 
> ##  re  Aluminum vs  copper.  You require  58-60 %  more  cross  sectional surface area
> if  Aluminum is  used  vs   copper.   Again, taking skin effect into account, the tubing  used
> could be paper thin...like   aprx  .00176 inches thick..on  2 mhz.   The only way to increase the
> puny cross sectional surface  area of the  tubing is to increase its circumference  by  58-60 %.
> And since  circumference is directly proportional to diameter, the diameter has to be
> increased by  58-60 %.   IE  instead of  say using .25  cu tubing,  you would have to use .4
> Al  tubing.  Closet thing to that would be  .375   al  tubing.   If  instead of using say  .375  cu tubing,
> you would have to use  .6 inch  Al  tubing.   Closest thing to that would be    .625  Al tubing.
> You can also do similar calcs for  solid  copper wire  vs  solid AL  wire.  Same deal, the AL wire
> has to have its  diameter increased  by 58-60%.

Sure - it's a classic engineering tradeoff - if you look at a dollars 
per dissipated watt sort of metric,  it takes 60% more aluminum than 
copper to get the same resistance, but copper costs a lot more than 60% 
more than aluminum.   So if you've got room, it's cheaper to make it 
bigger out of aluminum.  And, in self supporting antennas, aluminum is a 
better structural material than copper<grin>

I might note that you shouldn't build antennas out of stainless steel or 
titanium. Titanium has a great strength to weight ratio, and is fairly 
inexpensive, but is a terrible RF material.



> 
> ##  I have run loads of tests to verify the above over the last  10 years, and also again recently.
> But  you also dont want to lose it all by using lousy connections  to the given coil...or using puny gauge
> wire  to make taps on a tubing coil.    I use cu strap for  connections to tubing coils..or other  RF
> connections and terminations.   Strap is unique in that it conducts  RF current on both sides of the strap.
> Tubing only conducts  RF  on the outside.   .375  wide  cu strap is aprx the eq  of  .25 tubing.

And, because the skin depth is so shallow, you can use *really thin* 
strap - it becomes more of a mechanical issue than a electrical one. 
Lots of high performance inductors are wound with tape, not round 
conductors for all sorts of reasons.  Tubing *is* nice for HV - edges 
are bad.



I think my take home is that silver plating is nice for appearance and 
solderability, but probably not worth it for antenna components at HF, 
and maybe not worth it for high Q resonant circuits (you'd have to 
analyze it).

For a lot of things, whether to use copper or aluminum is going to be 
determined by mechanical considerations (or convenience... If you've 
already got 10,000 ft of copper or aluminum wire, you'll tend to use it)