Tom,
When you refer to displacement current due to the weak
electric field, are you refering to the electric field
between the side of the doublet driven by the center conductor
of the coax and the beads which are in series with the shield?
To me this implies that there is a shunt component to the
bead current albeit perhaps small. As Bill mentioned, do you
think this would explain the non-uniform heating?
An interesting experiment would be to move the whole string
of beads down the feedline away from the feedpoint a couple
of feet and then remeasure the heat distribution. Presumably,
this would reduce the displacement current a little.
Mike, W4EF
----------
From: Bill Coleman AA4LR[SMTP:aa4lr@radio.org]
Sent: Thursday, October 01, 1998 8:53 PM
To: w8ji.tom@MCIONE.com; towertalk@contesting.com
Subject: Re: [TowerTalk] Weird choke "balun" failure
On 10/1/98 9:56 AM, w8ji.tom at w8ji.tom@MCIONE.com wrote:
>> I know I've suggested this before, but why not use a mix of bead
>> materials? Have the beads closest to the feedpoint be 43 material, and
>> the ones away from the feedpoint be 73 material.
>
>I don't understand what any particular sequence of beads would do, since
>they are all in series without shunting components to "sap off" current. RF
>is just like low frequency circuits so far as series and parallel
>components are concerned. Whatever current flows in one end of the beads,
>provided the string is not heavily coupled to something via a strong
>electric field, flows right out the other end.
Hmm. I base this suggestion on the observation that others had offered
that the beads nearest the feedpoint of the balun were heating up and
destroying themselves, while beads away from the feedpoint were fine.
I see now that the suggestion infers that each bead is doing something in
isolation -- removing current from the sheild and generating heat. But
that's not true, as you point out.
Since that is the case, why DOES the bead closest to the feedpoint get
the hottest and fail first?
>(As a matter of fact, that IS why some traps have poor performance even if
>measured in circuit Q seems good. That's why I never use coaxial stubs or
>traps or put loading coils in tight metal shields or make them with large
>metal end caps or put large capacitance hats very near loading coils.)
(Hmm. The R7 doesn't have metal shields on its traps, but the R7000 does.
I take it you wouldn't endorse the change in design?)
>There is a small taper in current, do to the small reduction of current
>through a "displacement current" through the weak electric field.
Enough so that the beads close to the feedpoint are blistering hot, and
those far away are cool?
>> If you had an infinite variety of bead materials, seems the optimal way
>> to approach the bead balun would be to start with a low-permability
>> material at the feedpoint and increase the permability as you move away
>> from the feed. Done right, this would evenly distribute the loss (and
>> thus heating) across all the beads. None of the beads would get "hotter"
>> than the rest, and fewer beads would be needed overall.
>
>I don't understand why that would be true. It is a simple series circuit,
>and power dissipation is I^2 R.
>Since current is equal throughout the string, heat would be proportional to
>the ESR (R) of each bead. It always works that way, unless Kirchoff and Ohm
>were wrong and their laws are just suggestions.
You're absolutely right. For some reason, I was thinking each bead works
in isolation.
Bill Coleman, AA4LR, PP-ASEL Mail: aa4lr@radio.org
Quote: "Not within a thousand years will man ever fly!"
-- Wilbur Wright, 1901
--
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com
Search: http://www.contesting.com/km9p/search.htm
--
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|