In article .com,
wrote:

I hate switch mode power supplies.
I hate wall warts.
And I really really hate switch mode wall wart power supplies.
Linear wall warts are bad enough.


Found this while looking for a better (as in active) RFI probe.
http://www.arrl.org/tis/info/cable-modems.html

The biggest problem I have with the linked article is that, if I read it
right, he connected the low side of the power supply output to the AC
earth ground. This causes the supply to have a specific relationship to
earth ground and might conflict with other equipment the supply, then
modem, then all else connected to it. If some other piece of connected
equipment did the same thing there would be a ground loop that could
cause equipment malfunction. Highly valuable smoke contained in
semiconductors could be released as a result.

He has the right idea as evidenced by the fact that it greatly lessened
the interference but a .1 or .01 uF cap should suffice instead of the DC
connection and allow the power supply output to remain floating relative
to earth ground.

Another lesson to be learned here is that you must filter differential
noise before trying to filter common mode noise. The author would then
find common mode ferrite chokes to help reduce radiated noise.

The reason switching power supplies generate so much interference in the
short wave bands is the way switch mode power supplies work. First the
AC is rectified to a high voltage DC. Usually a couple of FET
transistors form a high voltage oscillator into a transformer primary
for AC input to DC output isolation. The switching supplies usually
operate at a fixed frequency and regulation is achieved thru pulse width
modulation. The feedback circuit is usually another small transformer or
opto-isolator.

This varying width pulse is usually 100 to 500 KHz square wave with fast
edges so the supply conversion is efficient. This ensures that the power
pulse is rich in odd harmonics that extend into the short wave spectrum.

--
Telamon
Ventura, California

In article ,
"Michael A. Terrell" wrote:

Telamon wrote:

In article .com,
wrote:

I hate switch mode power supplies.
I hate wall warts.
And I really really hate switch mode wall wart power supplies.
Linear wall warts are bad enough.


Found this while looking for a better (as in active) RFI probe.
http://www.arrl.org/tis/info/cable-modems.html

The biggest problem I have with the linked article is that, if I read it
right, he connected the low side of the power supply output to the AC
earth ground. This causes the supply to have a specific relationship to
earth ground and might conflict with other equipment the supply, then
modem, then all else connected to it. If some other piece of connected
equipment did the same thing there would be a ground loop that could
cause equipment malfunction. Highly valuable smoke contained in
semiconductors could be released as a result.


The CATV coax shield is required by the NEC to be bonded to the
electrical system's earth ground.

That's because it is a utility coming into the house. If power supply
outputs are connected to earth ground you need to pay attention to all
other DC supplies connected to the system in other equipment. They no
longer "float" relative to each other but have a specific relationship
to earth ground.

--
Telamon
Ventura, California

Telamon wrote:


That's because it is a utility coming into the house. If power supply
outputs are connected to earth ground you need to pay attention to all
other DC supplies connected to the system in other equipment. They no
longer "float" relative to each other but have a specific relationship
to earth ground.


Technically, its NOT a utility, because they are not regulated by the
PUCO.


Thanks, but I've worked with switchers in radio design for over a
decade. The quoted article referred to a cable modem, which is bonded to
earth ground at the ground block. That means that the power supply is
also bonded, and any RF noise generated by the switcher will be radiated
by the shield, on its way to ground. The higher the frequency, the
worse the problem is.

BTW, Vicor sucks! We were using them in the Microdyne receivers,
till their quality fell through the floor. I tested a half dozen other
brands before we settled on another supplier that was low noise,
acceptable build quality, the switching frequency not near one of the
multiple IF frequencies, and that shut down properly on overload or
short circuit. One defective Vicor supply set a $20,000 radio on fire
and did over $10,000 damage.

Oh, BTW, I worked on the 48 volt powered telemetry radio that's
aboard the ISS.

--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

In article ,
"Michael A. Terrell" wrote:

Telamon wrote:


That's because it is a utility coming into the house. If power supply
outputs are connected to earth ground you need to pay attention to all
other DC supplies connected to the system in other equipment. They no
longer "float" relative to each other but have a specific relationship
to earth ground.


Technically, its NOT a utility, because they are not regulated by the
PUCO.

OK, then it is because it is a service coming into the house.

Thanks, but I've worked with switchers in radio design for over a
decade. The quoted article referred to a cable modem, which is bonded to
earth ground at the ground block. That means that the power supply is
also bonded, and any RF noise generated by the switcher will be radiated
by the shield, on its way to ground. The higher the frequency, the
worse the problem is.

BTW, Vicor sucks! We were using them in the Microdyne receivers,
till their quality fell through the floor. I tested a half dozen other
brands before we settled on another supplier that was low noise,
acceptable build quality, the switching frequency not near one of the
multiple IF frequencies, and that shut down properly on overload or
short circuit. One defective Vicor supply set a $20,000 radio on fire
and did over $10,000 damage.

That would be tough as the switching frequency is not that well
controlled and since they are usually PWM then harmonics would change
depending on the load presented to the supply output.

Oh, BTW, I worked on the 48 volt powered telemetry radio that's
aboard the ISS.

Well, that would make it a floating supply alright.

--
Telamon
Ventura, California