[RFI] Anyone correlated feedline common mode current levels with RFI / noise received?
davearea51a at gmail.com
Sun Dec 9 15:12:33 EST 2018
I don't have quantitative measurements, but some experiences that may help:
1) CM Probes: I've built a number of these using clamp-on ferrites meant
for EMI remediation. I usually wind four to six single-layer turns of AWG
#18 enameled copper wire around half the clamp-on. The pigtails are
passed through a relatively large ferrite 'bead' to act as a common mode
choke. I then wrap the windings and clamp-on half with copper tape to form
an electrostatic shield. Do not fill the gap between halves when clamped
in place. This shield is connected to the braid side of the BNC
connector. Most of these come out to be 10 dB below the level of CM
current on the line around which they are clamped when connected to a
50-ohm system (spectrum analyzer). Most prove to be relatively flat from
100 kHz (or so) through 100 MHz. You can build a calibration fixture and
actually calibrate these if required.
2) My HF antenna system consists of a 450-foot long doublet (doublet as I
don't want to use the term "dipole" whichimplies resonance) fed with a
combination of open wire (made of AWG# 6 'ground' wire across the back of
the garage with spacing of about 1-inch) and 450-ohm window line (from
outside garage to feed point and inside shack from wall to CM choke). Once
inside the shack, the window line is fed to a CM choke (quadfilar wound,
but coinnected in bifilar manmner). that feeds a home brew twin-T matching
network (NOT, PLEASE, and "antenna tuner" !!!!) the configuration of which
consists of series (both legs) ganged but isolated series C's, shunt roller
inductor in the center, then another set of seriec ganged but isolated
C's. The reason for the two CM chokes, one on each side of the matching
network, is that the new appliances produce an unreasonable amount of RFI
and the matching network is open to the house volume. Certainly the choke
at the50-ohm (once adjusted) port of the matching network functions as a
balun as it transforms differential mode from the matching network to CM
for the coaxial cabling which connects the rest of the equipment. there is
no more than 5-feet, max, of coaxial cable in the whole system. The CM
choke on the antenna feedline side ofthe matching network isused on the
lowest bands, 160, 80, and 40-meters to further keep the CM energy on the
open wire feeders to a minimum. Generally, it is not needed on the higher
The CM chokes are home brew with 14 windings of Davis SuperFlex stranded
and insulated 'antenna' wire on two stacked 4-inch cores of #43 material.
As mentioned, they are wound in quadfilar manner but connected in bifilar
fashion: windings 1 and 3 in parallel and windings 2 and 4 in parallel. I
did this for two reasons: 1) to decreases I^2R losses and 2) to reduce
the effective Zo of the transmission line on the cores (without using
ribbon windings, it is difficult to realize Zo's much lower than 80 to 100
ohms). Both will withstand key-down at legal limit for a very long time.
BTW: I have an HP vector network analyzer to properly characterize what I
build. I have a write-up on these CM chokes, but its too large for this
group.io to post for others (about 4Mb with images of both the chokes and
the Smith Chart results).
3) 1:1 flux-coupled transformers: I tried this......once. I had a very
nice 6" ferrite core of #43 material which was a full inch thick on which I
would the transformer. Worked well until it cracked due to too much
(magnetic) current in the core - local thermal stressing. Believe me, the
core was large enough for CM choke application to withstand
multi-killowatts. But, it would not take the localized magnetic current at
power and cracked. RATS....Don't ask for my language at that moment. This
was years ago. Based on this experience and having used this design for
receive-only ELF and VLF active systems, I don't believe these are good for
transmit of any significant power unless the localized thermal effects of
large magnetic currents can be properly addressed.
4) Where to place the CM chokes: Again, I don't have quantitative data,
only experience. I've been licensed for some 59 years - Jan. 03, 2019 will
be 50-years. Many old-timers from pre-coaxial cable days recommend placing
CM chokes at the feedline connection to the antenna in the shack, at the
feed point of the antenna, and midway between. I have often entertained
the idea of placing yet a third CM choke in my installation at the feed
point of the wires. I have the chokes,but so far, haven't dropped the feed
to install it. As mentioned, I do have a CM problem from the new
appliances (no thanks to FCC who explicitly exempts home appliances - or is
it the appliance lobby? ). The ONLY band I have not tamed in that respect
is 30-meters where the appliance-generated RFI is just viscous.
Fortunately, much of the outside 'finishing' on the house is steel with a
steel roof and the antennas are far enough away that, if I keep CM noise
off the open wire feeders, is not a problem.
Dave - WØLEV
On Fri, Dec 7, 2018 at 8:45 PM JW via RFI <rfi at contesting.com> wrote:
> I didn't see my favorite.
> The 1:1 Flux coupled transformer.
> I use these as "ground-breakers" in the shack on the end of a long coax
> cable to a low-gain rx-only antenna.
> This style was particularly effective a few years back on the 630m band.
> de Jim WB5WPA
> From: Alan Higbie <alan.higbie at gmail.com>
> To: Rfi List <rfi at contesting.com>
> Sent: Friday, December 7, 2018 2:11 PM
> Subject: [RFI] Anyone correlated feedline common mode current levels with
> RFI / noise received?
> I know that we can minimize RFI / noise pickup by minimizing the common
> mode currents on our feed lines. And, that we want the choke to reduce
> common mode current to near zero.
> I have seen ***different types*** of common mode current chokes:
> -- ferrite bead balun (e.g. W2DU, W0IYH)
> -- wound-coax ferrite chokes (e.g. K9YC)
> -- broadband 1:1 current baluns (Balun Designs)
> -- binocular core coax common-mode chokes (e.g. W1HIS)
> There are also suggestions of ***where*** along the feed line baluns should
> -- at the antenna feedpoint
> -- at the amplifier output
> -- an "at least one other in-between." (W1HIS)
> There are RF current probes used for measuring common mode currents.
> +++ How do we, as a practical matter, measure common mode currents to
> confirm that what we are doing is effective?
> +++ Is there more than one method for measuring common mode currents, i.e.
> on the actual antennas we use?
> +++ Has anyone measured feed line common mode current before and after
> inserting balun? what results?
> +++ for placing common-mode baluns "in-between" antenna feed point and
> amplifier output - - is there an empirical method for determining where
> this should be? (assume a mono band antenna)
> +++ And, most importantly, has anyone correlated common mode current levels
> with the noise floor levels, etc.? (e.g. A-B testing)
> 73, Alan
> RFI mailing list
> RFI at contesting.com
> RFI mailing list
> RFI at contesting.com
*Dave - WØLEV*
*Just Let Darwin Work*
More information about the RFI