a summary of antenna tips for LF work
Tue, 09 Jun 1998 22:10:14 -0400
Walter Blanchard wrote:
> You shouldn't try reading across from your HF experience to LF.
> At LF practically any size of antenna is so short you are in a
> different business. Trying to feed a 15m antenna at 136 kHz is
> like trying to feed 30 cms at 1900 kHz.
> Think of it as a power generator (the RF power amplifier)
> trying to push current around a (simplified) circuit comprising
> (in series) :
> Loading coil resistance (RL),
> RF resistance of connecting wires (includes antenna
> wire and earth wire) (RW),
> Radiation resistance of antenna (RR),
> Ground losses (RG).
> You are trying to dissipate as much power as possible
> in RR and as little as possible in the others.
> You can't do much about RR which is always a small
> fraction of an ohm.
> RL depends on Q; for Q=300 might be 5 or 6 ohms
> for Q=150 could be 12-14 ohms.
> RW could be only tenths of an ohm if you use nice thick
> copper wire.
> And RG is what all the argument is about. It can be got down
> to a few ohms if you have nice conductive soil and are prepared
> to spend a bit on hundreds of metres of thick copper wire.
> If neither are true and all you can do is put a stake in the
> ground then you may well only get 100 ohms or worse.
> Or you could go and live on a salt marsh!
> So nearly all the power is lost in the earth and coil.
> Coil loss only gets important when the earth loss gets to
> be comparable so you spend a lot on earth wire and then
> find you have to spend some more making a good coil.
> If you find that making a good coil doesn't make much difference
> it means you have a lousy earth!
> The poorer your soil the more you need to bury lots of copper
> wire in the ground. This gets so expensive that professionals
> often choose LF sites on the basis of good ground conductivity
> more than anything else. If you can't run out long lengths of
> earth wire then put in more radials. There are tables available that
> tell you what the trade-off is but for amateur purposes they are
> rather irrelevant since they assume you've got almost unlimited
> ground space.
> It is VERY important to get good contact between your earth wires
> and the soil. So, no insulation, and bury them at least 150 cms.
> Use as thick wire as you can get, and certainly copper. If they have
> to be rather short then connect them at their ends to a 2m copper
> stake and if you connect intermediate stakes so much the better .
> As they get longer this becomes less important. My figures were
> based on a wire diameter of 3.5 mm buried not less than 300 cm.
> Yes, change your iron for copper if you can. As your
> earth becomes better it becomes more important to do this
> because the resistance of the iron becomes more
> dominating (RW above).
> However, there is at least one little trade-off in favour of a
> rather inefficient system - the bandwidth goes up! If you succeeded
> in getting a 1-ohm earth AND a coil of loaded Q=300 you could
> probably radiate a true 1watt from a 15m antenna with only 100w
> input but the bandwidth would only be a few hundred Hz.
> But if you have a bad earth and a poor coil requiring a kilowatt
> or so the bandwidth might go up to2-3 kHz .
> For amateurs this is actually A GOOD THING in practice because the
> whole thing is then much less critical in almost every respect. Like being
> to use RF-lossy but cheap insulators (as long as they can stand up
> to the voltage), not having to worry about whether your coil-winding
> technique is perfect or the wind blowing the antenna about and
> de-tuning everything! Or rain getting on the coil. Not to mention QSY'ing.
> There is no easy way of measuring earth loss without actually setting
> up some sort of system. Then, measure current going into the antenna,
> measure coil Q and from that estimate series resistance, take the
> power dissipated in the coil away from total power, and nearly all the
> rest will be going into heating up the earth. The little bit that's being
> radiated you can ignore for this purpose!
> Good luck!
> Walter G3JKV.