http://elecraft.85.s1.nabble.com/KPA-1500-SWR-fault-above-700-800-watts-on-160m-tp7664618p7664664.html
Thanks for the new link. I must say we made several big chokes using the "old" method and our analyser wasn't good enough to measure the common mode impedance, so our conclusion was they were all above the analyser limit of several thousand ohms. I'm about to make a single core choke, so I'll take a look at the new info.
> On 03 September 2020 at 17:05 Jim Brown <
[hidden email]> wrote:
>
>
> On 9/3/2020 4:21 AM, CUTTER DAVID via Elecraft wrote:
> > I use K9YC-style chokes, eg several turns on 240-31 core, or, if it's really bad, several turns on several cores using 7" diameter turns (from memory, check his cookbook). I'm told this is cheaper and better than anything you can buy. Of course I know nothing about your layout but it has always helped me to ground the coax before it comes into the house.
>
> Hi David,
>
> Note that, based on extensive research in 2017 and 2018, I no longer
> recommend the older style of choke that you're using, and I revised the
> RFI tutorial to reflect that. I published both that revision and the new
> Cookbook late in 2018.
>
> k9yc.com/publish.htm
>
> The problem with the older style chokes (multiple turns of coax on
> multiple #31 cores) is that 1) any "scrambling" of the winding reduces
> choking impedance, and 2) variations in winding style (diameter,
> spacing) make the chokes non-repeatable.
>
> The new choke designs ARE repeatable, AND take into account the wide
> manufacturing tolerances of all ferrite parts, which can significantly
> affect the choking impedance and where the chokes are tuned. The work I
> did in 2017 and 2018 included characterizing nearly 200 #31 cores
> obtained from multiple vendors over about ten years, selecting cores at
> the limits of those tolerances, winding and measuring hundreds of chokes
> on those cores, tabulating the results, and making recommendations for
> each band and each transmission line type based on worst case results
> for each design. Chokes wound following the new 2018 Cookbook should
> provide at least the specified response on any #31 core you buy. Note
> that #31 mix is made ONLY by Fair-Rite.
>
> Another point about measuring common mode current. As you have observed,
> it varies along the line, simply because in the common mode circuit, the
> feedline is part of the antenna, and both voltage and current vary along
> the line conforming to wavelength relationships at the operating
> frequency. A good choke forces a current minimum at the point where it
> is inserted. How common mode current varies along the line depends on
> the quality of the choke, the electrical length of the outer conductor
> of the coax (VF~0.98), and how the shield is terminated on the other end
> (grounded or floating using a link coupled tuner).
>
> A choke is most useful at the antenna feedpoint (up in the air, or at
> the feed end of a vertical or other long wire), which tends to minimize
> common mode current all along the line, because it "disconnects" the
> feedline from the antenna. A choke at any other point is FAR less
> effective -- all it does force a current minimum at the point it is
> inserted, leaving the feedline connected to the antenna, allowing common
> mode current between the choke and the antenna, and coupling any noise
> current to the antenna.
>
> For this reason, the first choke in any system should ALWAYS be at the
> antenna IF the antenna is matched to the feedline. The power handling of
> my chokes are ONLY applicable if the feedline is matched at the point
> where they are inserted. A choke inserted into a poorly matched or
> un-matched line is much more likely to overheat and fail. Dissipation
> (heating) is addressed in the 2018 Cookbook.
>
> 73, Jim K9YC
> ______________________________________________________________
> Elecraft mailing list
> Home:
http://mailman.qth.net/mailman/listinfo/elecraft> Help:
http://mailman.qth.net/mmfaq.htm> Post: mailto:
[hidden email]
>
> This list hosted by:
http://www.qsl.net> Please help support this email list:
http://www.qsl.net/donate.html> Message delivered to
[hidden email]