[TowerTalk] Lightning GROUND switch wanted RE:DE K0FF

[Tom Rauch]

> I have always built my own relay boxes using the Collins 5KW relays that
> are (were) available from SSof Nebraska and double up the contacts. The
> best I can hope for is reasonably low loss when in use, and of course it
> grounds the antennas when deactivated. 

One nice think about high power is any excessive loss instantly 
shows up. With 1500 watts at HF, as little as .1 dB loss would 
quickly destroy almost any relay contact suitable for switching RF 
with a size that would provide a reasonable RF performance.

> everything that CAN be grounded IS, it wont have a tendency to build up a
> static charge and attract lightning. The exception is the fiberglass
> covering over the typical 2/440 collinear antenna. It's a non conductor
> but I feel that it picks up a static charge from the wind etc. and causes
> lightning to hit it.

"Static" charges migrate through real-world insulators as well as 
conductors, so it makes no difference at all if the covering is 
fiberglass. The rate of charge buildup is so slow that wet fiberglass 
doesn't impede the charge movement. As a matter of fact, a tall 
structure made from a non-conductor is LESS likely to get hit than 
an identical structure that is conductive. If you don't want those 
fiberglass covered antennas to explode, cover them with metal!

We basically can't do anything at all to prevent charge buildup, and 
decease the likelihood of a strike. The only thing that would work is 
discharging the cloud, or changing the voltage gradient around the 
target.

A low voltage gradient, cause by a large blunt smooth surface, 
decreases the likelihood of a hit by increasing the voltage 
breakdown.

> something going on there if my interpretation is correct or not. PS on
> more than one occasion I have seen these antennas completely blown to
> smithereens with pieces found a block away....and the equipment was
> totally undamaged...protected by a Polyphaser.

My aluminum dipole arrays don't get blown to smithereens. The 
aluminum dipole array on my tower now has a one inch hole 
melted in it, but it never shattered and still works just fine.  Never 
ever mount a fiberglass antenna above the tower top, or within a 
few feet of the top, unless you plan on replacing it on occasion. 
 
> But back to the relays for a moment..I was asked by a nearby HAM to help
> him lash up his station, using the MFJ remote (wired) relays,and was
> shocked to see that both his brand new units do not ground ANYTHING when
> deactivated,. 

You'd better look again. The armature of the relay comes up and 
contacts ground. The center conductors are isolated, so the only 
possible arc path is through the grounded armature. The armature 
is the only path for the fault current on the center conductors, and 
it is double insulated with the armature in the middle of that double 
insulation a totally grounded barrier that allows about one inch 
separation.  

> ground plane is not attached to it in any way.  Is this normal? They can
> easily be modified to to so by providing a heavy jumper from the contact
> pad to the ground pad, but careful examination reveals no intentional
> means to accomplish this. Are they all that way? Another thing I don't
> like on the X strap is that it's thin brass and riveted.

The shorting bar is beryllium copper, not brass. It has to be to 
allow "spring action" so the contacts seat properly. Around 1983 or 
so the jumper was brass, until custom relays were obtained. 

There was no measurable difference in performance when that 
change was made. It was mostly done to simplify manufacturing, 
and make people "feel better".

> brazing rod crimped and soldered to 1/4 inch appliance right angle FASTON
> connectors which fit the spades on the Collins* relays, and use silver
> plated copper straps insulated with Teflon tubing for the wiring to the
> connectors. The lead length is longer than I like, but seems to have no
> bad effect on HF that I can determine. Looking at Tom's comment on spacing
> has me thinking that I'll add yet another relay to the box, in series on
> the rig side, and arrange it to give the maximum separation when
> deactivated.

What works best is isolating the center conductors from the 
common ground point. Use a grounded barrier between the two 
contacts.

Copy the RCS-8, and it will greatly reduce any possibility of 
currents flowing inside the cable to the rig. Of course it does 
nothing for currents on the outside, for that you need to use a very 
good ground or unscrew the cables and move them away from 
everything else. 

By the way, the relay in the RCS-8 is good for well over 6 kW into 
50 ohms at HF. Loss below 50 MHz is immeasurable on my 
Wiltron Network Analyzer, which resolves .05dB and is marked at 
every .01 dB.   

The exception would be switching them when full RF power is on. 
They will be toast quickly if hot switched at more than a few 
hundred watts.

> at the tower to select the various antennas into the hardline. The second
> is in a similar box on the grounding panel at the shack end, and this
> selects the various tower boxes. Maybe that's enough separation but I
> doubt it. Don't forget that the vast majority of the destructive energy
> from a direct hit will be on the shield. This is best dealt with using a
> bulkhead and single point grounding system.

Very true! And very well said.

The rig-killing pulse on the
> centerconductor comes from the lightning bolt "ringing" the resonant
> antenna, so the power pulse is more or less a high RF field ON Frequency.

Actually it can be on more frequencies than that. Virtually all 
systems have multiple resonances, and the antenna itself does not 
have to be the coupling mechanism into the center conductor.

Remember the shield is only a shield when the frequency is high 
enough that the shield is several skin depths thick. Not only that, 
any gaps or poor connections in the shield can become an entry 
point for energy into the center conductor.

You grounding advice was very good. The ground at the entrance 
point, which is where the relays should be, is of the utmost 
importance. Everything, even the power line and telephone lines, 
should "enter" the shack at that point.   
 
What I do, since my house was here before the radios, is bring the 
telco and power feeds from outlets in the room to that common 
point. I keep those wires away from everything else. At that 
common point, absolutely everything is bypassed and grounded, 
and then the leads leave there and distribute around the shack.

The most damage I've ever had...even with the unwise practice of 
leaving things connected (including modems) ....has been limited 
to an occasional 2N3904 or 3055 transistor that drives the antenna 
switching lines leaving the shack.

It would be smarter to unplug and disconnect, but I almost always 
don't do that. I have far too many cables and too much wiring.


73, Tom W8JI
w8ji@contesting.com

--
FAQ on WWW:               http://www.contesting.com/FAQ/towertalk
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com