RadioBanter - View Single Post - Sangean ATS-909 external antenna impedance??

Lostgallifreyan · December 24th 09, 07:56 PM posted to rec.radio.amateur.antenna

(Dave Platt) wrote in
:

In article ,
Lostgallifreyan wrote:

I'll pass. I think the reason no-one knows is that as you (and
others) say, it's not important enough. What does seem important is to
try to reduce localised noise, and to break the current link to protect
the radio input from static discharges. Whether I use coax or a balanced
loop made from speaker wire, it loooks like my next step is to get
Amidon FT-114-75 ferrite cores to play with, and in Britain I can't
easily do that, but if anyone knows a local direct equivalent to them I
can try that. I'll Google for things that fit the description (AL about
3000, permeability u=5000, about 1.14 inch outside diameter) but I think
it's wise to ask here to try to save time.

You can probably at least start your experimentation using the common
sort of interference-suppression ferrites that are found in many
computer accessories - e.g. molded onto DC cords, VGA cables, USB
cables, and so forth. Here in the U.S. these are easily available at
electronics surplus stores, ham-fest flea markets, and other such
sources.

In my experience, these tend to be a ferrite mix such as type 43,
which is optimized for use at somewhat higher frequencies than
HF/SWL... so they will probably not be optimal for your purposes.
However, they can be made to work.

A complete ring is better, air gaps might make it harder to assess how good
it is unless you know enough (or have equipment) to make comparisons. I
don't.

But I did find an appropriately sized type 43 ferrite ring on eBay
and put it on my watch list a few hours ago, so I guess this is what I'll buy
if I don't find better quickly. The seller has a few, I think.

A few years ago I constructed a common-mode feedline choke for our
local ARES/RACES ham station, to try to keep 40- and 80-meter signals
from being carried back down the outside of the coax and into the
building (our signal was audible on phones in the city's "911"
emergency dispatch center... *not* good). I took several tubular
computer-interference-suppression ferrite cores (large inner
diameter), glued them end-to-end with cyanoacrylate, let them dry, and
then ran some RG-8X coax through the center and back around the
outside and through the center again. The coax looped through the
tube of ferrites three times.

This resulted in an extremely effective common-mode choke. According
to my MFJ antenna analyzer, the impedance looking up through the coax
in the usual way (standard hookup, into a 50-ohm dummy load) was 50
ohms... the ferrites had no effect at all on the differential-mode
signal in the coax.

But, when I measured the impedance along the braid (i.e. from the
ground shell at one end of the coax, to the ground shell at the other...
a DC short circuit), I couldn't get a reading at any frequency... the
meter just said " 1500 ohms". Even at the lowest frequency of
interest, these non-optimal ferrites added so much inductance to the
common-mode signal path that they were blocking the feedline current
flow very effectively.

[Unfortunately, we determined that the phone interference was caused by
direct RF pickup by the phone wiring, which was in the "near field" of
the antenna above the roof. It occurred even if we completely
disconnected the building feedline, and fed the antenna directly from
a radio located up on the roof. The feedline choke couldn't help us.]

In your situation, I'd guess that you could probably make an efficient
feedline choke by using almost any surplus ferrite toroid which is
sufficiently large to wind your feedline (coax or speaker wire)
through it a few times. Or, use several surplus ferrite cores,
end-to-end, and if they're large enough in diameter, loop the feedline
down through the center more than once.

I hadn't thought of this for common mode noise suppression but I will try it,
sounds like a good idea. The rings I'm after are for the coupling at each end
of the line though, 9:1 or 10:1 at far end, 1:1 at receiver end. But I guess
that the clamp-on ferrite slug could be useful on the balanced line loop
that runs between the rings. I've also wondered if that balanced line might
better be a twisted pair with each twist perhaps 10 cm or so long to be sure
that common mode really is common, but I have no idea if that's wise or
required. Haven't seen anything to suggest I need to do it.

It won't be perfect (nor as good as if you used a ferrite optimized
for use at lower frequencies) but it will probably help matters, and
will give you some sense as to whether it makes sense to go to the
trouble and expense of buying ferrites that are better for your purpose.

80 turns.. plus another 8 for the secondary.. I read they should be close
turns, and the ring size was probably specified mainly to accomodate them
all. I will try other ferrites and carbonyl iron power type cores and
whatever else comes my way perhaps, I found four small toroids in a broken
computer SMPU that might be fun to try. Ultimately though, I'm aiming for
best shots because it's like plotting an unknown graph curve with only a few
samples, so the better the accuracy of each move, the better I can understand
any discrepancy in any one of those moves.

I have a basic philosophy that is similar to Jeff Liebermann's 'Learn By
Destroying' thing, as I also used to take stuff apart as a kid, and destroyed
plenty, but actually apart from seeing what sort of build quality was
acceptable and how discrete components were built, I didn't learn much
beyond tool handling, the thing that taught me most was when I took apart
something I had to put together again for fear of punishment if I failed and
got found out. So the closer I get to starting with something like an ideal
working part or whole, the faster I can figure out the meaning of significant
excursions from that ideal. It's kind of like the argument of whether a kid
should get his first bike new and working, or self-built from bits. I think
the first way is right because it's the fastest way to know what working
(and safety) really is, then start changing things, knowing what path was
taken and how to get back without help. Works great with computers too...