Coaxial stub matching of the .64 lambda antenna

Hello Folks,
At a tangent to my last post that was very kindly dealt with via the many knowledgeable hams on these pages.. Has anybody tried Joe Carrs coaxial stub matching of the .64 wave antenna? I know the coil in series with the radiator system is often cited, but have not read of this other approach from users. In addition there is a US hams web site who advocates the .64 lambda over the 5/8 wave vertical as being an 11 metre secret that hams using the lower bands other than 11 metres don't seem to be party to .. The difference between the .62 lambda and .64 seems neglidgable to me.. But does this guy have a point? He points out that the extra length is a definate plus. I would also like to know if the standard radial length should echo the radiator length or be the 1/4 wave lengths?

Thank you, God bless, Mark M6AWG

 

For those of us who don't have Joe Carr's book, can you describe that stub matching configuration in detail?

 

Dear Cecil,
I would indeed, but the said publication had been literally worn away over the years with constant reference use..
Maybe someone out there has a reference to it?

Very best wishes, Mark M6AWG

 

Sure, here it is for free download; http://www.eagle3.net/n4ywn/docs/PracticalAntennaHandbook-vol4.pdf.
It'll take a bit to load. It's 625 pages of good information.
The information on what Mark is doing is on page 199.
Have fun
73
Gary

 

Thanks Gary - for everyone's benefit, let's explain how that stub works. Below is the graphic from Joe Carr's book:

http://www.eagle3.net/n4ywn/docs/Pra...dbook-vol4.pdf

That "stub" is actually two stubs, one series section transformer (L1) and one shorted parallel stub (L2). The series section transformer (L1), which carries a very high SWR, transforms the feedpoint impedance from e.g. 150-j500 ohms to e.g. 50-j300 ohms. The shorted parallel stub is inductive with a reactance equal to +j300 and neutralizes the capacitive reactance in the above 50-j300 ohm impedance thus leaving 50 ohms resistive and a 1:1 SWR on the coax to the transmitter. That is ideally how it works perfectly.

If one has an antenna analyzer, like the MFJ-259B, one can adjust L1 for a 50-jX value and then adjust L2 for a +jX value. That's why I love my MFJ-259B.

Don't expect the L1 and L2 length formulas to be perfect because Joe Carr could not possibly know all the variables at your QTH. Both L1 and L2 may have to be adjusted to obtain a perfect SWR=1:1 but there indeed exists particular lengths for L1 and L2 that will accomplish that feat.

 

There isn't much information in the book to help you determine exactly what you need for the length of these lines. The formulas he gave probably only applies to his particular antenna. The feedpoint impedance of the vertical you are trying to match varies depending on how it is built. A ground mounted vertical will be different from and elevated one. Sloping radials will make a difference, and the diameter of the vertical element will make a difference. Experimentally determining the lengths is a little difficult, and if you happen to be very far off at the beginning you may never find a solution. Fractions of a foot make a big difference.

Examples for 5/8 wave 20 meter vertical:
Example 1: Ground mounted using 43.6 ft of #12 wire, RG213 coax, L1 = 9.4 ft, L2 = 1.35 ft
Example 2: Ground mounted using 43.6 ft of 2 inch diameter pipe, RG213 coax, L1 = 8.1 ft. L2 = 2.1 ft

A solution is easily obtainable with a Smith chart if you know the feedpoint impedance of the antenna. A solution can be determined using EZNEC by manually inserting various lengths of coax, assuming you can get in the ballpark to start. In both cases, you will probably have to use NEC to get the feedpoint impedance of the antenna. Don't forget to include the coax loss in the model. That could make a difference between getting SWR = 1.0 and SWR = 1.4 (with RG213). Coax loss will be unexpectedly high for such a short length of coax.

Jerry, K4SAV

 

Jerry, re my experimental 20 metre.64 lambda monobander.
Thank you to you and Cecil. I am going to have to go the stealth route and use 28 guage wire for the element..and RG 58 for the coax feed line and stubs I have on hand. I also will NOT be able to slope the radials down at 45 degrees(only Horizontally)as the local geography will not allow me that situation. What lengths do you think the stub lengths would have to be with a 28 guage wire element.
Thank you so much guys. Ps radial length 1/4 or 5/8 wave?

Mark M6AWG

 

Based on an EZNEC analysis, for a 20 meter, 0.64 wavelength vertical (44.3 ft) of #28 wire with 4 radials 10 ft above average ground, RG58/U coax, L1 = 9.35 ft, and L2 = 1.45 ft. Loss in that RG58 is about 2.7 dB. Loss in the copper wire is only about 0.2 dB.

Maybe this info will save a few head hairs. Fine tuning the coax lengths is tricky because they interact. Assuming you are close to the right lengths, L1 mostly affects the resonant frequency. Longer lowers the frequency. L2 mostly affects the impedance. Shorter lowers the impedance, except when it gets too short the resonant frequency goes up, and when it gets too long the impedance goes up and the resonant frequency goes down.

Use 1/4 wavelength radials, not 5/8. That provides the lowest impedance at the feedpoint to best decouple the feedline. Still with only 4 radials you may need a choke on the feedline.

Jerry, K4SAV

edit: You might want to reconsider building this. What you are proposing will have less gain than a 1/4 wavelength vertical with 4 radials.

 

Gerry,
What if I abandon the coax stub route and use a 2.7Uh coil..is the 5/8 wave then still inferior to the 1/4 wave vertical ground plane as your footnote suggested?
So much conflicting oppinions everywhere?

M6AWG

 

If the "5/8" (actually .64 wave) wave vertical has less gain than the 1/4 wave (I assume this is at lowest launch angles)
then why did the FCC require "5/8 " wave minimum tower height for the highest coverage class (1A) AMBC stations , and not 1/4 wave towers ? (This also assumes 120 1/4 wave radials)