NEC is in FORTRAN IV and runs just fine under Linux. You may need to get sources and compile it with gfortran.  There might be precompiled binaries out there.  You won't get a fancy GUI and stuff, but there are utilities out there to plot patterns with GNUPLOT, etc.

Electrical code requires AWG 6 (or bigger) for bonding. Exothermic welding or listed compression clamps to do the connection.

That said, hard solder (brazing) probably meets the intent of the code (you gotta use a torch, after all) and is a practical approach. Clamps are a pain, because they could loosen, so you can't bury the connection.

daisy chain or parallel depends more on the physical arrangement of them.  If they're in a line, then daisy chain (i.e. there's not much point in running parallel wires). If they're around a central point, then parallel (star) is better. 

I suggest you look up the Onderdonk and Preece equations. 
Which commercial power company?  A name.  I've got most of the reports generated by, for instance, the Rural Electrification Administration, which would be a likely candidate for that time span.  You don't think perhaps that economics have changed in the last 80 years?  Copper and labor costs might have changed?

Fused #6 on a *power pole* is a very different situation from an antenna.  For the fault current reason.  On a power pole, a lightning strike can provide an initial conductive path from an energized line to the ground wire, so the high current can persist much longer than from an antenna.  Grounding practices in electrical utilities are different from those for antennas, for good reasons.
But hey, it's in the NPFA 70/National Electrical Code.  And the code making panels do know their stuff. There are hundreds of pages of discussion around changes in the code and why they make them.  Someone claiming, as you do, that the code requirement is insufficient has a higher burden of proof.   So show us.
Hmm, I would point you to the work of E.B. Rosa, in 1907, at the National Bureau of Standards.  He did the analysis of inductance, based on fundamental physics.

You, like many, are confusing AC resistance (flat strips better than round conductors) with inductance (flat strips not that much better than round conductors, for the same cross sectional area).  Hey, up until a few years ago, I thought, too, that flat strips had low inductance.  Then I went out and started calculating it, and gosh, it's not that big a deal.   

Simple example.. Consider a wire AWG10... it has some inductance per unit length.  Now split that into two AWG13 wires (same total cross sectional area). Lay the wires next to each other, and hook them in parallel. What's the inductance?   say that AWG 13 has *exactly the same* inductance as AWG10 (it's slightly more, but let's assume for now).  Two inductors in parallel.. you might think, at first, that the total inductance is 1/2 (two parallel inductors).  but it's not.. The magnetic field from one wire is tightly coupled to the other wire: it's right next to it.  Turns out, the *mutual* inductance is almost the same as the inductance, so what you have is two inductors with twice the inductance, in parallel, so the net is the same as before.

In reality, the inductance is sllightly (emphasis on slightly) less.  AWG 13 has slightly more inductance than AWG 10, and the mutual coupling isn't perfect, but it's still remarkably close.

Go do the math.  Read the book from people who spent their life doing this stuff 100 years ago.
I'm going to say that the inverse is the case.  I can cite the basis for my statements.  Go calculate the inductance or AC resistance of a 90 degree bend. You'll find it's not much different from that from the same straight length of cable.  That is, say I've got a 1" radius 90 degree bend.  That is, the wire in the bend has a length of 1.57 inches (roughly).  The inductance of the 1/4 turn coil is pretty close (as in less than a few % different) from a straight length of wire 1.57 inches long.

Even 180 degree turns don't have that much inductance, beyond the length of the wire involved.  You need to start getting towards 360 degrees before the inductance starts to rise.  It's because inductance is all about magnetic field coupling, and the field from a wire doesn't couple well to a wire that's not parallel (why we put antennas at right angles to avoid interaction).

On the other hand, a sharp bend dramatically reduces the breakdown potential. Breakdown voltage for a curved surface is roughly 70kV/inch for the radius of curvature.  A fairly short wire with a fast rising edge from lightning (rise times of 1-2 microseconds) can have a fairly high voltage induced on it.  (1uSec @ 20 kA @ 1 meter of wire is about 20kV, from L*di/dt).. it's easy to get a potential difference of several hundred kV on a long wire carrying a lightning impulse, and if the surface next to it is at ground potential, then breakdown from the sharp edge is likely.

This, by the way, is why the NFPA 780 (lightning protection) and NEC require lightning down conductors to be spaced from other wiring and metallic things (downspouts, plumbing, window frames). 

But I'm always willing to look at new (to me) analyses and calculations, perhaps that cover some aspect I've overlooked and that hasn't been covered in the literature I've seen.  So if you have that analysis from the 40s.. tell me where I can find it, and I can figure out why what they recommended then is different from what is recommended now.  It very well could be relevant to ham installations (which are different than power stations, FAA control towers, and commercial radio stations: all of which have widely disseminated grounding handbooks and standards)

AWG 10 has a fusing current for a lightning waveform (short pulse some tens of microseconds long) very much higher than the stroke current, except perhaps for a 300kA megastroke, and even there, it's probably good enough.  Larger conductors are used for mechanical reasons, not electrical.

Upon what basis do you recommend #6?  where did the 96% number come from?   If there is a published study I'd like to read it, because I've not run across an economic analysis like that.  Most recommendations basically copy someone else's recommendation, which in turn isn't based on physics or analysis, but more a "it worked, it wasn't too expensive, so we keep doing it" strategy.

Electrical substations use heavy grounding because they have very high fault currents from the power line.  The flashover may be triggered by lightning, but when you have a transmission line carrying 10kA, if it breaks down, it's going to burn for a long time before the breaker trips and interrupts the current.  A very, very different scenario from a lightning impulse which lasts a fraction of a millisecond.  (google for "Lugo station arc"... that's why they have heavy grounding at substations)

More surface area does NOT markedly decrease the inductance.  It *does* reduce the RF resistance for high frequencies, but for lightning, the inductance dominates.  The difference in inductance between a square bar that is 1x1 cm and a flat strip that is 20cm wide by 0.05 cm thick (i.e. same 1 sq cm area) is about 40%. 

Sharp bends don't increase the inductance.  The concern is with arcs from "pointy things"

it depends on which section of the NEC they were wiring in accordance with.  CATV (Art 820), for instance can use a different gauge than radio and TV antennas  (art 810) which is different than amateur antennas (also 810) which is different than telephone wires (art 800 and 830)

AWG 10 is what's required for an antenna support.

telephone - insulated AWG 14 (800-40)
antenna structures - bare or insulated AWG 10 (810-15) - TV, Radio, Amateur
discharge unit/leadin cable - bare or insulated AWG10 (810-21)
CATV - *insulated* AWG 14 (820-40)

stranded or solid doesn't make much difference.

One thing to remember is that all earth grounds have to be interconnected with AWG6.. so if you drive a new rod (for whatever reason), bond it to the system ground with #6.

the primary purpose of that ground wire is to reduce the number of deaths and injuries when a powerline hits the antenna or structure (one of the most common causes of death by electrocution, as it happens), it's not really a "lightning grounding conductor".

gets more of the antenna away from the shielding/capacitive influence of the vehicle
makes it easier to avoid whacking things accidentally
more convenient to tie tip towards back, than towards front
doesn't get in the way of loading things through the rear hatch/tailgate/etc

I had my screwdriver installed on the front bumper of my Passat (visual feedback of coil position was useful).

the theory is straightforward, the practice is not. In particular, choosing appropriate magnetic components is hard.

The other design challenge you'll face is good behavior at different loads.  It's fairly easy to get a switcher to work into constant load, but into a varying load (e.g. a high power RF amp for SSB or CW) it is much more difficult.  The regulation loops that control the switching are more complex than a simple linear pass regulator.

proper component selection (particularly for the capacitors in terms of ESR, ESL, and di/dt ratings) is also part of the game..

EMI/EMC is also a challenge, but at least brute force can work.



You probably won't find a proven design for your need.  While you could probably wind up building a multikilowatt DC power supply, your first one probably won't be efficient, small, or light.   As others have suggested, building a smaller one first, to get a feel for the various design parameters and "care-abouts" is where it's at.

Clean may not be the issue.  It will almost certainly be -48V.  The real issue will be how well it works with a rapidly varying load.  Telco and networking stuff tends to be almost constant current draw.  Set up a little test with a suitable switch and load and watch the output on a scope. If you have a high power SSR for DC, and a bunch of  lightbulbs for a load, that would work.  A mechanical relay and a digital scope would work.

ANd the study done modeling a ground burst at the Los Angeles/Long Beach port found that more deaths would occur from people getting in car accidents rushing away. Or, for that matter, towards grossly overloaded hospitals.

But to the original question.  It's been studied, it's been experimented with (Teak shots).   The Glassman book (Effects of Nuclear Weapons) has a fairly decent summary.

As SAV pointed out, the long wire is a terrible lightning ground.  But that's not why it's there.  It's there to keep everything "moving together" voltage wise.  Since the bonding conductor is long and so are the other wires, the "difference" in voltage is small, and survivable.  There's no illusion that in a transient situation that anything that's bonded remains near "earth" potential.  Just that bonded stuff stays similar to other bonded stuff in the same general vicinity.

For the lightning strike dissipation, you need a grounding system specifically designed for that.  Then, that gets bonded to the electrical safety ground (the primary purpose of which is to make sure you don't get a shock when touching something metal while standing in your bare feet) on the general principle of "all grounds bonded together".

This is really more of a equipotential bonding conductor.  Presumably the lightning current is going elsewhere.  And lightning doesn't care much about corners. The added inductance from a 90 degree bend is negligible. The problem with corners is that they are a "sharp point" which can promote sideflash to other conductors.  A 50 foot wire has huge inductance, no matter what shape or configuration it is.


A qualified electrician wouldn't ask about bonding. Not only is it the code, but it also makes sense (yes, there are things in the code that don't always make sense..)

Get yourself some 2.4" toroids with #31 mix and use those to make a choke for your feedline.  5 or 6 turns of the coax works just fine.
(Fairrite 2631803802)
Works better than almost anything else: a heck of a lot better than any of the air core coil o'coax things, and generally better than the W2DU string o'beads (just because you get almost N^2 the effect from N turns through a core).

31 mix is relatively new (10-15 years?) and is almost ideal for HF chokes.

Check out K9YC's writeup on RFI and chokes: http://www.audiosystemsgroup.com/RFI-Ham.pdf

Test data for all kinds of different baluns and chokes. Part numbers to order from FairRite or stocking distributors like Arrow, Mouser, Newark.  I wouldn't get them from Amidon.. they tend to be higher priced than the big mailorder places.

Mouser part #623-2631803802 $7/each in qty 1..  They're NOT in the catalog page, but they are online.

There are other sizes that are smaller, and snap on varieties, etc. 

Yes.. this is my reason why you shouldn't sleep with your phone on next to your head. 

All those little pulses when the phone interacts with the cell site that we all hear as EMI on computer speakers, PA systems, etc.  And of course, in a weak signal area, it jacks up the power and tries harder.

Absorbed dose wise, I think the energy is fairly low.  For *most tissue* the blood flow can carry the heat away: so the odds of you getting injured from your phone in your pants pocket are more from the lithium battery catching fire than from RF absorption.

But putting the transmitter next to your head..   I can live without my hand or leg, but not without my brain.  So even though I'm pretty much a believer that there's no effects,   if there *is* an effect on nervous system tissue, you've got an awful lot of it next to the transmitter.  Hence the ALARA strategy..  Putting it a meter away instead of 2 cm is a pretty big field difference.

And, I gotta say, having made measurements on a lot of antennas and EMI/EMC nearfield measurements over the years, I'm not so confident of those absorption models.  There's an awful lot of assumptions about how one holds the phone relative to the head and hand, and small changes might have big effects on the fields.

I'm also curious about the fields when you've got headphones hooked up.  Sure, the antenna is farther away, but do they measure how much energy couples into that wire?

And I am sorry.. I did respond somewhat uncharitably.   As you can tell, this is something that I care about, since there is so gosh darned much misinformation out there. 

The people in my neighborhood ranting about remote reading electric meters causing cancer, as they hold their iPhone to their ear with their WiFi enabled laptop sitting on their lap.  It *is* unrealistic to expect those folks to spend the time understanding RF dosimetry, models, etc. and I have great sympathy for the poor engineers at the utility who have to try to respond to it. (same thing for power line EM fields.. "Currents of Death" did us no favors)

 I also do this stuff for a living (be in EM fields, have to do safety analyses, etc.) so by now, the whole validation of the limits is something I'm personally very comfortable with.  I am not totally convinced on the epidemiology and physics for cell phones and heads and long term effects, so I'm a member of the ALARA category (As Low As Reasonably Achievable).  My teenage daughters and I have a difference on "Reasonably Achievable" as far as distance from phone to head while asleep: I'd like to get my daughters to at least put their phone on the bedside table, instead of on the pillow.  But I'm also not going to freak out about it.


It is also good if hams, as a group, become well educated on this, and how the standards get set, and why they are what they are (i.e. not speculative based on some theoretical harm), because that WILL be the next way that people will attempt to "zone" us out of existence.  "That radio weirdo down the street has an antenna that's not only ugly, it's killing my babies with radiation."

Good, then, you'll find it interesting (albeit tedious in some cases)

It's good that the C95.1 doc is available free now.  Forking out $100 for it in the past was a significant disincentive.  It really DOES lay out the theory and the test and the whys and wherefores.  This rev (compared to the previous 1999(?) version) has some significant differences (which haven't been reflected in OET Bulletin 65, at least as far as terminology)..   There is a LOT more stuff essentially coming from the concerns about cellphones and the enormous amount of research done in the last decade on specifically that issue (especially with respect to what field causes what heating where, in the head)

For most hams, though, not much changes.  The science and measurement stuff for HF hasn't changed in decades.  And the change from "controlled/uncontrolled" to "MPE/action level" is more of a "making it like other safety standards" interpretation thing. (harmonizing, as the "term of art" has it)

ANd that ground rod hooked to the transient suppressors needs to be bonded to your electrical service ground with #6, too.

Clamping is good. solder is bad.

Watch where the wire runs.  Don't run the grounding wire next to any other wires or conductors you care about (if the lightning comes, you don't want it to sideflash over).

No neat curled coils, just run it reasonably direct and avoid loops.  Bends don't have a huge effect, although if they're sharp, the corner of the bend might flashover easier than just a straight run.

Get yourself a copy of the Mike Holt "Low Voltage Handbook":  it's on his website, it's free, it has all the relevant electrical code explained with examples, etc. http://www.mikeholt.com  Much cheaper than buying a copy of the code, and better explained, too.