[TowerTalk] My Yearly Mistake

[Jim Lux]

On 8/14/15 12:40 PM, Roger (K8RI) on TT wrote:
> We were having an off line discussion, but I thought I'd posy my answer
> to one.
>
> I have mentioned several times that I never scored 100% on a math test.
> Lots of 95s, but I could always count on one dumb mistake and that's
> exactly what I did here.  Once I found the mistakes, it was so obvious
> that I fail to see how those schooled in mathematics missed it/them.
> It's been over 25 years since I graduated and I often hear the median
> referred to as the middle value, when in reality it is the middle,
> measured value. IOW you first arrange all the measure/observed values in
> order of the parameter you are keying on.  Then take the middle
> measure/observed variable.  IOW if there are 3 values, you take the
> value of the second. If there are 11 measured values. you take the
> 6th.   What I did was to take the middle value between the minimum and
> maximum of the set.  Why?  I don't know.  Minor in math, but brain fart?
>
> Well at least I have my allocated mistake for the year out of the way,
> so, I should be spot on for the rest of the year <cough><cough>
>
> At any rate, as you said, we are playing the odds with lightening and we
> need to know the likelihood of those outliers.  At first the return Vs
> cost is great, but the law of diminishing returns quickly rears its ugly
> head where it costs a lot for a little increase in return.
>

Which is why knowing the distribution, or something like the 95th 
percentile stroke might be more useful than the median (50th 
percentile).  That makes it more of a "expected value" calculation which 
is what insurance premiums, etc. are based on.




> One thing I do think missing in the ground rod Vs UFER comparisons is
> the area covered by the system.  How does that system area compare with
> the contact area of the ground rods and UFER contact area once the
> desired resistance has been achieved?

It also kind of depends on the objective of the system. Is it to 
minimize damage to the whole system (or some part)? Is it to minimize 
personnel hazard?  For instance, electrical substations have big 
grounding grids because they are interested in minimizing the "step 
potential" (how much voltage difference there is between your feet) in 
the event of a big fault.

There's also the issue of "availability/mean time to repair" ... a ham 
might tolerate taking a direct hit, having some small amount of damage 
in the system, being off the air and being ok with repairing it when the 
storm is over. An installation for safety-of-life applications (public 
safety repeater, FAA control tower) in a hard to get to location might 
want a system that is "can take multiple licks and keep on ticking" .

My general observation is that hams are primarily concerned with "don't 
blow up the rig" and "don't light the house on fire", "don't blow up the 
rotator", and, to a lesser degree, "don't break the antenna".  (because 
it's hard to get up the tower to replace the rotator).


so looking at "how well grounded is the tower" probably isn't as 
important as "what wires/cables/conductors come into the shack" and "how 
do we minimize voltage differentials within the shack".

After all, airplanes take lightning hits without being grounded.





>
> As hams with out a government budget where monstrous cost over runs are
> often tolerated, our auditors/wives are not nearly so forgiving of
> estimated costs, let alone, over runs.  <:-)
>
> In corporate America, avoiding cost over runs is a good way to assure
> continued employment.
>
> 73
>
> Roger (K8RI)
>