Antennas: Surge arrestors and lightning]

[Andre' Kesteloot]

Alan Melia wrote:

> I am no expert on lightening protection but I did work about 30 years ago on
> high voltage protection of (the pre-optical) submerged repeater amplifiers.
> These are power-fed in series and when your friendly Trawlerman gets snagged
> and puts his axe through the cable, it is in effect shorting 7 miles of coax
> (100pf per foot or so) charged up to anything up to 15kV. We used gas
> discharge tubes to protect the power supply together with BIG Zeners. The
> gas tubes have to be specially made (I have never quite understood what the
> little neons on the old wartime BC348 and command sets were supposed to do)
> The problen with discharge tubes in the dark is that eventually all the odd
> ions recombine, then when you want the tube to strike it has to reach very
> high voltages before it will take current. Purpose built  (gas) surge
> arrestors have a small amount of radioactive material in them to ensure that
> there are always a few ions floating around to start the breakdown avalanche
> off. I think that gas tubes took somewhere in the order of a few hundred
> microseconds to fire, but they took the brunt of the power.
>
> The amplifier input was a 3:1 step up transformer feeding the base of a
> bipolar transistor. Having found some diodes that would absorb 10kV for
> 5usecs we put a pair across the the transformer input. We found that whilst
> they protected the transformer from fusing, in this position they did not
> protect the transistor. We found the transformer 'rang' and produced a surge
> in the secondary that destroyed the transistor. So a second pair of back to
> back diodes was connected directly across the base and ground. As these
> where wideband amps, intermod was inportant so we actually used a pair of
> diodes in series in parallel with another pair with reversed polarity. The
> reason for this is that the capacity of a junction diode varies with the
> square of the voltage applied, so if you half the voltage swing you reduce
> the capacity variation by a factor of 4, and a resultant improvement in the
> intercept points.The diodes used were a standard in the 1N3595 range from
> Fairchild. The secret being a large 'S' shaped spring with a reasonable
> thermal inertia that made contact to the top of the diode chip, and absorbed
> the rapid rise in temperature at the junction. The output stage was
> protected by a zener and small diode in series, the zener providing a
> 'hold-off' voltage.
>
> I noted that my old FT101ZD had a small 'lamp-fuse' in series with the
> aerial feed to the receiver section. Induced voltages took that out but it
> obviously protected the rx front-end.
>
> I would suggest a spark gap might be the best safety device ....2 sharp
> points about half a cm apart (it probably depends on the rf voltages around
> but air breaks down at about 10kV per cm gap , I think) What you must watch
> is that if the static causes a flashover whilst you are transmitting the
> spark will be maintained by the RF even though the voltages are not enough
> to initiate a discharge. The safest way is Mal's....ground it all and go and
> make a cup of tea until the storm passes.
>
> Cheers de Alan G3NYK
> Alan.Melia@btinternet.com