Maker Pro
Maker Pro

ground loop

I'm trying to understand a system with rf circuit.

I found that there is a concept called "groundloop"

I searched it at the Wikipedia, but I don't think I now know it.

How can I explain the concept?

What is floating potential?

below is the original statement I'm trying to understand

"
Ground loops can add excess noise and therefore efforts were made
to avoid these throughout the whole system. As a consequence,
many signals are transmitted differentially, the cable shields are
typically only connected to ground at one end, many of the power
supplies use floating potential, and isolated inputs with optocouplers
are used for most digital signals. These precautions were only
taken in the parts where the DC level is important, and not in the
pure radio frequency (RF) circuitry.

"
 
J

John Popelish

Jan 1, 1970
0
I'm trying to understand a system with rf circuit.

I found that there is a concept called "groundloop"

I searched it at the Wikipedia, but I don't think I now know it.

How can I explain the concept?

What is floating potential?

below is the original statement I'm trying to understand

"
Ground loops can add excess noise and therefore efforts were made
to avoid these throughout the whole system. As a consequence,
many signals are transmitted differentially, the cable shields are
typically only connected to ground at one end, many of the power
supplies use floating potential, and isolated inputs with optocouplers
are used for most digital signals. These precautions were only
taken in the parts where the DC level is important, and not in the
pure radio frequency (RF) circuitry.

"

Floating potential just means that the voltage has no direct
connection to earth. A transformer secondary is a common
source of floating potential. Of course, the static charge
on a balloon that holds it to the ceiling is another.

A ground loop is just a conductive loop that is made up, in
part, by two separated connections to Earth. This is what
you get if you ground both ends of a wire. The Earth
completes a loop that connects the two ends through another
path than through the wire. Any changing magnetic field
that passes through the area enclosed by this loop will
generate circulating current around the loop. The
resistance, inductance and propagation delay of that current
will produce voltage differences between the ends of the
wire. So it you are using that wire as a zero volt
reference for another signal at both ends of the wire, the
measure of that signal with respect to this reference will
be different at the two ends of the wire. The loop will
have coupled energy from the magnetic field that passed
through the loop into the signal, corrupting it to some extent.

If that reference wire were earthed at only one end, there
would be no loop, but at the other end, that "ground"
reference may not have exactly the same voltage as the local
Earth potential. So an isolated receiver that subtracts the
reference voltage of the remote ground from the signal, and
re references the difference to the local ground potential
will break the loop and eliminate the corruption.
 
C

Charlie Siegrist

Jan 1, 1970
0
Circa Wed, 08 Aug 2007 03:43:40 -0000 recorded as
below is the original statement I'm trying to understand

"
Ground loops can add excess noise and therefore efforts were made
to avoid these throughout the whole system. As a consequence,
many signals are transmitted differentially,

Differential transmission is another way of describing a balanced line,
such as a twisted pair cable, where neither polarity of the signal is
referenced to ground. Google "balanced line."
the cable shields are
typically only connected to ground at one end,

This is now describing an unbalanced line, such as a coaxial cable, where
the shield is referenced to ground at only one end of the line, in order to
reduce the possibility of current finding a path through the shield. This
statement could also refer to a technique of grounding chassis in a "star"
pattern, meaning that there is a central point where connection to ground
is made, and all chassis are isolated from each other except at the point
where ground cables from those chassis are tied to the central point.
many of the power
supplies use floating potential,

Describes not grounding either end of a power supply. This way, current
that is produced by the power supply does not return through a ground path,
thus does not add to current through ground connections. The return path
is separate from ground.
and isolated inputs with optocouplers
are used for most digital signals.

An opto-coupler will isolate return paths for small signals, again keeping
the paths for ground current to a minimum.
These precautions were only
taken in the parts where the DC level is important, and not in the
pure radio frequency (RF) circuitry.

I'm not sure what this means. It seems specific to the application. In my
experience, these methods of isolation are used to minimize the effect of
60Hz radiation on RF circuitry.
 
R

Rich Grise

Jan 1, 1970
0
Floating potential just means that the voltage has no direct
connection to earth. A transformer secondary is a common
source of floating potential. Of course, the static charge
on a balloon that holds it to the ceiling is another.

A ground loop is just a conductive loop that is made up, in
part, by two separated connections to Earth. This is what
you get if you ground both ends of a wire. The Earth
completes a loop that connects the two ends through another
path than through the wire. Any changing magnetic field
that passes through the area enclosed by this loop will
generate circulating current around the loop. The
resistance, inductance and propagation delay of that current
will produce voltage differences between the ends of the
wire. So it you are using that wire as a zero volt
reference for another signal at both ends of the wire, the
measure of that signal with respect to this reference will
be different at the two ends of the wire. The loop will
have coupled energy from the magnetic field that passed
through the loop into the signal, corrupting it to some extent.

If that reference wire were earthed at only one end, there
would be no loop, but at the other end, that "ground"
reference may not have exactly the same voltage as the local
Earth potential. So an isolated receiver that subtracts the
reference voltage of the remote ground from the signal, and
re references the difference to the local ground potential
will break the loop and eliminate the corruption.

John, this is awesome! Even _I_ get it now! ;-)

So, when's your book coming out? ;-)

Cheers!
Rich
 

neon

Oct 21, 2006
1,325
Joined
Oct 21, 2006
Messages
1,325
lets start with floating potentials. Look at the sky during a storm now that i what a call floating potential. it is there just floating millions of volts just sitting there. you got now potential difference?. now if the lighting hit your house which is tied to mother earth but when that happens the house will experience a voltage dirfference because of the massive current flowing through it. electrical substation have the same problem. soluton tie a heavy copper cable to earth ground to pass the current without elevating the edifice to 1000 of volts. ground loops are caused by payin attention to where current flow. example you have a 40w amp and you tie the gnd to the power supply then you tie the same gnd to at the end to the preamp. so now whatever current the power amp use it elevate the gnd it becomes a signal for the preamp. that is a loop . best thing to do is tie each comonent path to the source the power supply. this is the basic at hi freq more things gets invoved. and yes a cable shiled should never carry any current for any reasons and should be tied at one end only the lowest gnd potential usualy the very input that you are trying to shield.
 
J

John Popelish

Jan 1, 1970
0
Rich Grise wrote:
(snip)
John, this is awesome! Even _I_ get it now! ;-)

So, when's your book coming out? ;-)

Sorry, I haven't yet read nearly enough questions to force
me to face all the stuff I still don't know.

But there is nothing like trying to stuff more than you know
about something into a single paragraph, to make you think
really hard. That is what I get out of being here. I'm
just getting close to understanding some other people's
books, at this point.
 
I'm trying to understand a system with rf circuit.

I found that there is a concept called "groundloop"

I searched it at the Wikipedia, but I don't think I now know it.

How can I explain the concept?

What is floating potential?

below is the original statement I'm trying to understand

"
Ground loops can add excess noise and therefore efforts were made
to avoid these throughout the whole system. As a consequence,
many signals are transmitted differentially, the cable shields are
typically only connected to ground at one end, many of the power
supplies use floating potential, and isolated inputs with optocouplers
are used for most digital signals. These precautions were only
taken in the parts where the DC level is important, and not in the
pure radio frequency (RF) circuitry.

"

Ground loop: You need to draw a picture and think about it, like this:
Your (ancient) hi-fi amp is plugged into the wall socket. It has an
earth in the mains plug.

Your (ancient) tuner is also plugged into the wall socket. It also has
an earth in the mains plug.

You connect them together using a screened co-axial cable because...
it is screened and you don't want any mains hum... but you get... lots
of mains hum!?

When you listen to the tuner with headphones (only) it is perfect.

When you plug a microphone (only) into the amp it is perfect.

You only get the hum when they are joined together by the coax.

The reason is that both the (ancient) amp and the (ancient) tuner have
an continuous earth between their coax screens, to their wall sockets
so when they are plugged together a loop is created, an earth or
ground loop via the coax, though the amp, through the mains loop,
though the tuner back to the coax screen.

Here is the hard-to-understand-bit: both the amp and the tuner have
huge, bulky, mains transformers in them. Each transformer is radiating
magnetic fields at mains frequency. These changing fields are cutting
the ground loop. The ground loop now has changing currents induced in
it.

Any conductor carrying a current *must* have a voltage across it
driving that current - otherwise there would be no current :)

Therefore the coax screen *must* have a changing voltage (at mains
frequency) between each end.

So even with no signal in the coax's center conductor, there is a
signal (mains hum) in the screen. And it gets amplified and comes out
the speakers.

The problem ceases by breaking the "gound loop" e.g. disconnecting the
earth in the plug of the tuner.

That's were the name "ground loop" came from but it applies to any
loop grounded or not. But evidently it is a common error particularly
in screens and earthing plans.

*******************************************
The twisted pair: suppose you plug a mike into the amp but you *still*
get mains hum?!

How can that be? The coax screen of the mike should prevent mains
radiation getting to the centre conductor! The answer is that is
*does* prevent radiation getting to it and *that* in itself, is the
problem. Like this:

Imagine the coax screen at the amp's imput as being zero (or put it
another way, that you are standing on it so as far as you are
concerned, you and it are always the same voltage i.e. zero) then
mains radiation (and any other nearby source) will be inducing
currents in the coax screen. These currents flow back and forth
between you, and via the mike to the signal wire and thence to the
amp's input - hence the mains pickup.

The screen *is* the problem.

An answer could be to use two plain wires instead of the coax - now
this is the clever bit, any radiation is now picked up on *both* wires
the same and currents flow back and forth in each IN THE SAME
DIRECTION and the result is that they *cancel* each other out (you
might have to draw a picture here).

Evidently if both wires are laying straight then one wire might pick
up a little more than the other if it is closer to the source of
radiation so you avoid that by twisting them together to even out the
likely pickup from all directions (as well as keeping them close).

This twisted pair thing is usually done inside a screened cable. Your
mike cable will have three conductors, i.e. two signal wires, and one
screen.

Robin
 
Top