On 4/17/2012 2:20 AM, Steve Hunt wrote:
> I just double-checked my results for the 6 off #31 bead choke. I looked
> carefully at the ratio of the Resistive (loss) component of the choke
> impedance to the Reactive component, across the range 1MHz to 50MHz.
> Between 1MHz and 30MHz it closely tracks the complex permeability charts
> published by FairRite: in other words, below about 4MHz the Inductive
> reactance predominates, but anywhere between from 4MHz to 30MHz the
> Resistive component dominates and by the predicted ratio. Only above
> 40MHz does the choke become predominantly Reactive again, due to the
> stray C involved.
Following up on Jim Thomson's love for strings of beads. It's quite
instructive to study the Fair-Rite data sheets for representative
parts. Go to the link below and enter the part numbers, then go to the
bottom of the sheet and study the impedance curves. Steve and I agree
that the curve for a string of single turn beads will be simply the
curve for one multiplied by the number of beads in the string.
Now, here's where we diverge a bit on our analysis of the usefulness of
strings of beads. Yes, a single turn through #31 (that is, simply
clamped onto the cable) shows resistance the same general order of
magnitude as reactance between 4 MHz and 20 MHz, you've got to have a
LOT of beads in that string to hit 5K ohms resistive, which is known to
be a good design value to make a dent in RX noise. By contrast, it's
easy to hit 5K with the multi-turn chokes on #31 or #43. The only times
I use a "string of beads" in single turn fashion are either at VHF, or
where it simply isn't possible to wind turns.
BTW -- I did all this work FIRST, and published it first. So a more
accurate description of your measurements is that they confirm mine, and
those of my colleague who produced the lovely families of data for 1-14
turns of wire. :) We published that in 2005, and that publication noted
it was a parallel resonant circuit, and the importance of resistance in
providing suppression . That fact was well known by EMC engineers in EU
as long ago as the 60s, but lack of communications between them and the
ham radio world, no one seemed to know about it.
Likewise, the first place I saw any reference to the benefits of a
serious common mode choke to suppress RX noise was a publication by
W1HIS around 2006, and I immediately built some chokes, put them in
antennas here, and confirmed his assertion. He also said that 5K ohms
was a good design target.
A few years ago, some EEs working at CIA made contact with me, later
visited here for a brain pick. A year or so later, the leader of that
group sent me a DOD engineering report from the 70s that went down the
same trail I did, coming to nearly all of the same conclusions as I had
(and none in conflict), but missing dimensional resonance. They were
also designing to suppress RX noise, and came up with design target on
the order of 5,000 ohms and used multi-turn chokes to get there. That
report was probably never classified, it was simply a matter of lack of
communication between EEs working in different places.
73, Jim K9YC
TowerTalk mailing list