End fed antenna and KX1

[WA7CC]

Elecraft says that the range of the KX1 tuner is limited and will not tune a high Z antenna such as an end fed dipole.

It looks like the tuner is an L network, so I'm wondering if anyone has tried connecting an external capacitor across the antenna connector to enable the tuner to match an end fed dipole.

[AA4PB]

Just shorten the end fed antenna a little (i.e. don't make it exactly 1/2 wavelength). That will lower the feed impedance and allow the tuner to match it. Be shure to use a counterpoise connected to the case of the KX1.

[KC8LTL]

All antenna tuners have limitations.  While my K2's ATU can tune some antennas that some other tuners can't, it did not handle it well when it had an end fed 1/2 wave antenna.  The KX1's ATU has less range, so, no, it would almost certainly not do well.

But why do you want it to be a half wave antenna?  Add or subtract 5 to 10 feet and you will still be on the air with close to the same pattern and the tuner will thank you for it.

[K4AHO]

Best bet is to try something like the PAR EndFed for Portable 40/20/10 meters...   I use one with my K1 and with PAR's matching unit works great...   Does not seem to require a counterpoise...   The coax feed from the matching unit to the radio is plenty enough...

72

Jim
K4AHO

[WX7G]

The best all band, 40,20, and 20 meter, antenna I have modeled for the KX-1 consists of two 25' wires. One slopes upward from the radio while the other lays on the ground. Both wires are routed in the same direction.

[VE3WMB]

The Elecraft Manual suggests a mostly vertical wire in the range of 24 to 28 feet long for use with the KXAT1. I have used both 24 and 28 foot wires and they will match and work quite efficiently on 40m/30m/20m. My experience is that if you can get the wire almost vertical the 24 ft wire will match fine. If you intend to erect it as an inverted-L or run some portion horizontal then go with a 28 foot wire as it will offer a more stable impedance on 40m.
I normally work the wire against 5 X 16ft radials laid on the ground
(made out of computer ribbon cable).  A semi-vertical wire as short as 12 feet will load fine on 30m and 20m if you don't care to operate 40m. The 24 to 28 foot wire is in a length range that is not near a multiple of a half wave on either of the three bands so it provides a reasonable impedance for the KXAT1 to match.

Another trick that I have used for a portable vertical is to use a 20 foot crappie pole like the Black Widow or Shakespeare Wonderpole as a support for the wire. I use a 20 foot length of TV twin-lead with both
conductors shorted at the top as the vertical wire. For 30m and 20m I short both of the conductors at the bottom of the twin lead together to make a "FAT" 20 foot radiator. On 40m I disconnect one of the two conductors of the twin lead from the rig at the bottom, giving me a 40 foot long wire folded back on itself. The additional capacitance from the linear loading effect will allow this to match on 40m using the same 5 X 16 foot radial system as mentioned above.

I use a Pomona binding post  to BNC adaptor and banana plugs on both conductors on the twin-lead. The Pomona allows two plugs to be connected to the RED terminal (i.e. center conductor on the BNC) so this facilitates switching the antenna between 40m and 30m/20m, by
connecting or disconnecting one side of the twin lead at the pomona adaptor.

I have taken this antenna idea one step further and created a 14 foot version of this linear loaded antenna for KX1 Pedestrian Mobile operation. I use a 14 foot length of indoor Twin Lead attached to a lightweight 13 foot fishing pole and I also drag a 29 foot wire and this setup allows me to operate 40m/30m/20 while walking around.

Michael VE3WMB

[WA7CC]

I've decided to build AA5TB's tuner and give the end fed dipole a shot. I'll let you all know how it works out.

http://www.aa5tb.com/coupler2.html
http://www.aa5tb.com/efha.html

[AA4PB]

I don't think you'll find any noticable performance difference between an end fed antenna that is exactly 1/2 wavelength long and one that is a little shorter or longer to lower the impedance enough to let the KX-1 tuner match it.

In any case, any end fed antenna requires a counterpoise of some sort. If you don't provide one then the coax shield, the radio case, etc acts as the counterpoise. How efficient these are depends on a lot of variables - which is why I prefer to use a well defined counterpoise with known characteristics.

[WA7NCL]

I scanned the posts and nobody answered the question directly.

The answer is yes a shut cap will allow it to match high impedances.  I have demonstrated this with my K2 ATU.  Theoretically the L network should be able to match the end fed antenna.  I have built manual tuners that will do it.  However tuning becomes very touchy due to high Q.  

Evidently the ATU tuning algorithm is not able to deal with the very sharp tuning characteristic as it tries to converge to a match.  When you add a shut capacitor to the output, you lower the resistive part of the impedance down which makes the tuning less sharp and the ATU can find a match.  You need to be a little carefull not to add to much capacitance or you will end up with too low of resistive part and efficiency will suffer.  As an aside, I have seen the shunt cap at the output used in ATUs for marine random wire tuning.  Some marine ATUs have a jumper to add a fixed shunt cap at the output to help "problem" matching

I brought this subject of not matching even multiples of quarter wave wires up on the Elecraft list some years ago and got the Elecraft response of "don't use half wave wires".  They didn't seem to feel the limitations were a problem and didn't see any merit in the shunt capacitor solution.

Lastly, if you look at the matching system usually used for the end fed half waves, the tapped parallel tuned circuit, you see it can be decomposed into an analogous pi network which is the capacitive analog of the tapped inductor.  The only advantage to the usual tapped circuit is there is often a coupling link that attempts to offer isolation from common mode problems.

Pat yourself on the back, you were correct in your thinking.

[AA4PB]

How does adding additional capacitance lower the resistive part of the antenna impedance? How is this better than just not cutting the antenna wire length to exactly 1/2 wavelength?

An end-fed 1/2 wave antenna should have an impedance somewhere around 2000 -j0.

[WA7NCL]

If you shunt the Hi Z resistive antenna impedance with a capacitor you get a resulting impedance that has a lower resistive part and a capacitive reactance.  In an L network you then "tune out" the capacitive part with a series inductor leaving the (lower) resistive part without reactance.  My reason for adding the shunt reactance is to help the ATU L network match to the very Hi Z antenna.

You can change the antenna length and it will "work".  I was just answering the question of the poster.

The theory is that if you are anti-resonant (half wave) you have very low ground current in the return path.  There is a certain amount of controversy over whether this is a benefit or not.  I don't really think we want to go there in this post.

[WB6BYU]

The typical autotuner network for a high impedance wire would be an L network,
often the low-pass variant with a series coil and shunt capacitor.  It isn't
immediately obvious why adding additional shunt capacitance in parallel with
the tuner output capacitor would make any difference, except on those bands
where the available C in the tuner is insufficient (which is unlikely to be the case
on 20m with a tuner designed to go down to 80m.)

The only cases that I can think of would be
(1) adding the external capacitor enables the internal tuner to switch to a
step-down L network, making the combination a pi network.  I'd think that
the required capacitance would be fairly critical for this to work well.

(2) The tuning algorithm is designed not to match impedances over a certain
level, either because the tuning is too sharp (perhaps the available step size
is too large?) or to protect the tuner parts against the high voltages that
might be encountered.

I'm not saying that the method doesn't work, but it isn't clear WHY it works.

[K0MF]

I just got my KX1 kit (20/30/40M).  Much of my operating will be from hotel rooms.

I'm wondering if anyone has experience with using the end fed wire right from the rig and dangling it out the window (it would hang vertically downward), and a counterpoise wire thrown on the floor in this type of environment.  I'm guessing that current maximums could occur at or right near the rig, which is inside of an (often) metal / concrete building, so it might be best to choose a wire length that push the current maximum "out the window" so to speak, for each band to maximize efficiency.

Any real-world experience / recommendations out there..?

Mike K0MF