[TowerTalk] Cushcraft/MFJ Traps

[jimlux]

On 9/12/19 12:42 PM, Jim Brown wrote:

> I have two very different takes on this. First, traps are an inefficient 
> way to build a multi-band antenna. That means they suck up gain. Also, 
> because the elements are shorter, their radiation efficiency is reduced. 
> The best multiband antennas don't use traps.

I'm not so sure about the efficiency aspect for shorter antennas - sure, 
for "very short", the matching network losses will increase, but the 
actual antenna efficiency isn't different (I^2*R losses are usually 
pretty low)

Take a 6 meter long dipole as an example. REsonant at roughly 24 MHz - 
here's the feedpoint impedance
f	r	x
23.6 	77.6359 -12.4921
23.8 	79.7763 -2.8992
24 	81.976 	6.6949
24.2	84.237 16.2932
24.4 	86.5613 25.8988

Now let's drop to 18 MHz, so the dipole would be 75% of resonant length
f	r	x
17.6 	33.6036 -316.66
17.8 	34.6034 -305.433
18 	35.6276 -294.325
18.2 	36.6769 -283.331
18.4 	37.7518 -272.445

So you'd need some sort of matching network to cancel out the 300 ohm 
reactance. It's pretty easy to come up with a coil that has a Q of 200, 
so the 300 ohm coil would have a resistance of 1.5 ohms. Compared to the 
36 ohm radiation resistance, that's about 4% or 0.2 dB.

At 50% length:

11.8 12.9553 -732.472
12 13.4523 -713.564
12.2 13.9618 -695.14

Now we're starting to be significant, a inductor Q of 200 is going to be 
around 3.5 ohms loss resistance, and against 13.5 ohms antenna R, that's 
a 20% loss (1 dB).

Of course, for many HF links, on receive, the SNR is determined by the 
atmospheric noise, and antenna gain (for lowish gain antennas < 10dB) 
doesn't change the received SNR - the reduced gain drops both the 
desired signal and the noise level.

For transmit, of course, it does affect the SNR that the other end sees.