On Feb 28, 1:26 pm, wrote:

I have the (incredible fast and reliable) Verizon fiber optic into the
house for my internet connection. So far so good. However at the
access point there is a battery backup unit for the fiber-Copper
converter. This battery unit is kept charged by a powersupply that
makes an ENOURMOUS amount of RFI that radiates onto the (15 feet) lead
into the unit and from there into the shack. (S6)

Does anybody have any suggestions on how to silence this? Or... a
suggestion for a cheap and 'quiet' powersupply? (It's output is a
regular 12v)

Over the years, I have seen a number of posts that cite strong RFI
from various sources. In most cases, where I have obtained a unit and
made measurements, the units have complied with the FCC limits.

Most of these devices are switchmode power supplies. In the US,
because they intentionally generate signals above 9 kHz internally,
they are classified as unintentional emitters and must meet the FCC
Part 15 conducteed emissions limits below 30 MHz and the radiated
emissions limits above 30 MHz. (The premise is that on HF, small
chasses don't radiate much, while the wires they are connected to
do.)

Unless they are used as part of a computer system, they are "verified"
as described in the Part 15 rules. This means that the manufacturer is
required to test them and keep the results on file. Switchmode power
supplies used in computers are authorized under a Declaration of
Conformity (tested by the manufacturer as verified devices, but only
by labs the FCC has approved) or, rarely nowadays, under
certification, where the manufacturer sends test data to the FCC,
which then authorizes the product.

The manufacturer must meet the limits and authorization requirements;
the operator of the device is responsible for harmful interference.

In most of the interference reports I have seen, someone is
interfering with themselves. If the device meets the FCC rules, and it
is likely that it does, the FCC not, of course, offer much sympathy.

The premise that boxes don't radiate much on HF, but wires do, points
to the cure. The easiest thing to try is a common-mode choke on the
wires going to and from the box. Below 30 MHz, ferrite beads usually
don't have enough inductance to function well as ferrites. You will
need to get a toroidal core that is big enough to pass the
connector(s) through, and wrap about 10 turns of wire onto it. For
lower HF, -73, -75, -77 or -J material is probably best. For upper HF,
I would use -43. The nomenclature of most cores sold to hams has the
size in the first part of the part number and the material in the
second part. For example, an F-240-43 core has an outer diameter of
2.4 inches and uses -43 material.

That may solve the problem, but with a switch-mode supply, it is
likely that differntial-mode noise is also being coupled onto the
line. For that, you need a "brute-force" AC-line filter that uses
inductors and capacitors. I don't know if is still a good number, but
Radio Shack's catalog # 15-1111 is a suitable filter. Others are sold
by Industrial Communications Engineers (http://
www.iceradioproducts.com/).

You may need a combination of brute-force and common-mode filtering.

73,
Ed Hare, W1RFI
ARRL Lab