[CQ-Contest] Coax Stubs for SO2R

[Jim Brown]

On Mon,10/5/2015 7:20 AM, Jukka Klemola wrote:
> Jim,
> I thought placing double stubs on the TX line is not so critical.
> The two band stopping stubs are supposed to be about quarter wave 
> apart; on the harmonic band; and that should effectively remove the 
> need for extended optimization.
>
> Also, in my experience, if you get approx 30 or 35dB attenuation using 
> one stub, you will get more than 45 dB attenuation over the whole band 
> when measured in a 50 ohm system.

Sure -- using my DG8SAQ VNWA, I measured peak attenuation of these 
double stubs of 55 dB for the 80M stub pair and 59 dB for the 40M pair. 
BUT  a monoband antenna is NOT a 50 ohm system at the harmonic 
frequency, and the transmitter is NOT a 50 ohm source at the harmonic 
frequency.

Most monoband antennas present a very high Z to the line at their 2nd 
harmonic, which establishes a very high SWR for the harmonic, so the Z 
varies over a wide range along the line. A stub works by placing a short 
on the line at the harmonic frequency, and is most effective if placed 
on the line where the Z is high. It is MUCH less effective when at a 
place in the line where the impedance is much less than 50 ohms.

SO -- with double stubs separated by 90 degrees at the harmonic, the 
second stub will always be at a high Z spot on the line, but the first 
will depend on where it is along the line.

There is a second issue.  By their nature, the output stages of modern 
power amps, both tubes and solid state, produce 2nd harmonic that is 
only about 6 dB below the fundamental, and must be filtered by the 
amplifier's output network. Most of these networks are designed for 50 
ohms, and that is how they are tested.  If the last element of that 
network is a capacitor, it will be most effective if it sees 50 ohms or 
more as a load AT THE HARMONIC, and least effective if it sees a short. 
Likewise, if the last element of the filter is an inductor (Pi-L), it 
will be most effective if it sees Z of 50 ohms or less.

I rigged two power amps, a Ten Tec Titan and an Elecraft KPA500 with a 
voltage tap at their output terminal into a dummy load and into a double 
stub network in line with my 40M dipole. With the stub feeding either 
the antenna or the dummy load,  the second harmonic at the output of 
both power amps varied by +/- 10 dB as I added short sections of coax to 
vary the length of the line between the amp and the stub over more than 
180 degrees at the harmonic frequency.

SO -- if we put the stub in the "right" place for the power amp, we get 
full performance of it's harmonic suppression network plus the 
suppression of our stub(s), but if we put it in the "wrong" place we can 
lose as much as 20 dB less of the 2nd harmonic suppression of the output 
network.

> At least my measurements show these results:
> -placement not critical for double stub

As you can see from above, that's only true if the antenna is near 50 
ohms at the harmonic.

> -attenuation almost doubles in dB compared to one stub

Yes.

>
> --> I have been building only double stubs nowadays.

Me too.

Another point. When you have done this amount of suppression, you will 
likely begin to find other sources of 2nd harmonic, as W3LPL recently 
observed (perhaps on another reflector), AND you may also find leakage 
paths in your antenna switching system.  I recently replaced a vintage 
Six Pack with a 6x2 sold by 4O3A. I measured crosstalk in the 4O3A unit 
before installing it and the Six Pack after removing it and found the 
4O3A unit had more than 20dB better isolation on 20M.

In hopes of picking up another 6-10 dB of isolation, I am also planning 
to replace some of the random vintage patch cables in my station with 
newly made jumpers using high quality RG213 and Amphenol 83-1SP connectors.

73, Jim K9YC

>
>
> 73,
> Jukka OH6LI
>
> 2015-10-05 9:57 GMT+03:00 Jim Brown <k9yc at audiosystemsgroup.com 
> <mailto:k9yc at audiosystemsgroup.com>>:
>
>     On Sun,10/4/2015 2:59 PM, Jeff AC0C wrote:
>
>         That's a great way to start.  Especially as the solar cycle
>         fades and 10/15 are not open.  Then in the day, you run 40/20,
>         and in the night you run 80/40.
>
>
>     Yes, but there's a LOT more to it if you want to maximize the
>     effectiveness of the stubs. It can matter a LOT (20-30 dB) where
>     along the line stubs are placed, both with respect to the antenna
>     and to the power amp.
>     See my piece in NCJ one issue back, or download it from my website.
>
>     For our CQP expedition, I made up double stub packs for the 40M
>     and 80M CW stations only. Each pack was two stubs cut to kill the
>     second harmonic, with a quarter wave (at the harmonic) connecting
>     them. I didn't have time to optimize their placement, but using
>     two stubs insures that you'll get at least 25-30 dB, and with
>     optimized placement, another 25-30 dB.
>
>     Stubs are less important on SSB because the likely operating
>     frequencies are not directly harmonically related, whereas the
>     harmonic of 3525 hits 7050, and 7025 hits 14050.
>
>     73, Jim K9YC
>
>
>     _______________________________________________
>     CQ-Contest mailing list
>     CQ-Contest at contesting.com <mailto:CQ-Contest at contesting.com>
>     http://lists.contesting.com/mailman/listinfo/cq-contest
>
>