[RFI] An RFI solution report
Sat, 10 Jan 98 19:33:37 +0000
here some of my practical solutions to a number of interferences I
had to front with. Probably not everitying I've done it's exactly
against RFI in the strict sense of the word since a part my work was
devoted to avoid HF RFI to PC and other devices, while usually RFI is
intended as the interferences to HF from other devices, but I believe
finally the two things are overlapping.
As a preface, my shack is at 5th floor of a concrete building just
12 meters below the KT34 and less than 2 meters away from the
supporting tower used also as a shunt feed vertical, for 40 80 and
160m. The VHF UHF antennas, used for packet, are equally close to the
shack, having the UHF 20el on a 1.5 meter pole placed some couple
of meters from the main tower and a dualbander vertical in a
very similar situation.
This description has been made to explain that the distance to
ground is more than 20 meters and consequently a connection to the
ground, if good for safety reasons is definitely bad to keep "cold"
my shack by RF.
The symple connection of devices to the ground heavy wire doesn't
help leaving unsolved 95% of the interferences.
My goal was to have VHF/UHF TNCs and computer working in the same
while without damages first (...had to repair a TNC2 blown because of
RF), and free by any mutual interference, even if using the home made
Amp with the 3CX1200.
What I got, after a number of tries, is a situation where this is
nearly 100% obtained unless few restricted HF spots (a total amount
of 15-20 KHz in the whole HF spectrum) where some birdies and noise,
due to computer, monitor and one of the TNC, is present.
The level of this interferences is not very big since none of them
is strong enough to move the s-meter needle of my rig, when most of
the times QRM does it.
The first "good idea" was to discard a TNC in a plastic box, the
second to place a ground bus between all the metal cabinets of all
the devices including Computer, PS and rotators control boxes.
Most of these devices has no ground screw and GND passes only through
power cables. Altough this is good for safety reasons, often this
cables are too long, in term of fractions of wavelengts so resulting
inefficent for effective RF screening purpose.
While the above operation took away most of RFI from TNCs toward
HF and some of the PC "noises", it was definitely insufficient to
prevent RFI from HF (40 80 & 160) transmitting to VHF/UHF rigs and
Since my tower (including rotator and KT34) work also as a shunt fed
vertical, the only effective way I found to solve this trouble was
to create an artificial earth on its base.
Due to the omega match I used, SWR was not a problem even when I
used, as counterpoise, only a group of 16 radials each of them only
8 meters long.
Altough maximizing the counterpoise current didn't give appreciable
benefits in the performances of the antenna itself, the upgrade of
implementing 3 separate ground radial systems, each one raised in
current flow (I used an RF ammeter) on its band with variable
discrete components, gave a sparkling improvement in RFI creating an
RF nearly-no-voltage point at the antenna (tower) base.
This action reduced the rf flows, trough antenna cables, to the shack,
and was the only effective one I could find.
To be precise I have to tell that coaxial cable shield of KT34 and
all the rotator cables (by use of 10,000 pF cap.) are connected to
the base of the tower, likewise the gnd of the 3 omega matches.
Concerning RFI from PC to HF the great improvement was obtained
with the use of RF chokes along the cables to TNCs, monitor and
Having the HF antennas at a nearly zero distance from the shack,
any radiation by any of the cable between PC, monitor, TNC and
PS is big enough to be received.
Here it is the need of stopping the rf flow and cosequent radiation,
along the interconnecting cables. The only possible (practical) way
to do it is to keep 'em short, and placing chokes as close as it is
possible to their ends.
To realize effective chokes in few turns it is a need to use
materials whose permeability is very high, 2000 or more, I used
5000 toroids where I could.
Some of the cables, i.e. that one of the monitor, are pretty big
or some other cable had sealed connectors beeing unpractical to
wound, also for the biggest ferrite toroids I could find.
Trying to solve this nasty problem, it occurred to me I saw ferrite
elevation transformers (TV ones) realized by cores in two pieces
in a form of C then kept toghether, to ensure the concatenated flux,
by long screws and bolts.
They are pretty easy to find out, at least here. Shops or TV repaiar
are very happy to sell the old models at symbolic prices.
I had two doubts, the first about the losses of a ferrite material,
good at 19KHz, when used at 30 MHz, the second about the detriment in
the magnetic flux because of placing the two halves faces one against
the other. Infact, if the two couples of faces are not perfectly
matching one against the other, there is a space of air in which the
concatenated flux drops dramaticly.
I wasn't also very enthusiast to use the original metal fixture to
combine the two halves the meanwile keeping the windings.
Thinking about an alternate metod, I saw a little plastic bottle
containing Loctite (TM) liquid.
Since the cyanoacrylate bonds effectively only if the surfaces are
perfectly matching, if the two halves of the ferrite core would have
solid joined by Loctite (TM), then the magnetic flux won't have been
too much reduced and the problem of fixture inherently solved.
I wounded the monitor cable on the two halves of the ferrite core, and
after 60 seconds from bonding I tried, with all my forces, to separate
the halves but it was not possible any more by symply use of hands.
By tests, this rudimental but easy to manage cores, are enough to
effectively stop RF in the HF spectrum in big cables.
I hope this report can help someone in solving problems similar
Administrative requests: rfi-REQUEST@contesting.com