how is noise cancelled in a balanced signal?

[SouthSIDE Glen]

well, microphones hanging upside down in studios has nothing to do with noise, generally. It's more of a space reason to get the microphone and microphone stand out of the way....or for Tube mics. Our noise enters the microphone cable, not the microphone. So it doesn't matter where the microphone is.

I honestly wasn't sure if he was joking about the mic thing or was serious.

I think the whole mic upside down thing started with the tube mics. The idea was that the tubes generated heat that when rising could warp or otherwise affect the condenser elements. So hanging the tube upside down allowed the heat to rise away from the condenser instead of into it. Though the space thing is also valid.

Nowadays I think there are two reeasons why people hang their mics upside down. one is the space/room thing, the other is because it looks "k3wl" .

 

[Farview]

this i did know but i thought that a balanced cable is called a 2 conductor cable and unbalanced cable is called a single conductor cable.

do you have any idea how a balancing circuit distinguishes between noise and good signal?

Just to be clear. Balancing doesn't get rid of noise in the signal. It gets rid of noise that is picked up by the cable. it is not a noise reduction circuit.

 

[travelin travis]

Just to be clear. Balancing doesn't get rid of noise in the signal. It gets rid of noise that is picked up by the cable. it is not a noise reduction circuit.

Yea, I realize this. The noise is still in the signal but cancelled out by phase cancellation.

 

[easychair]

Ok, I now understand how a transformer works. I read a bit on op amp theory and..........i'll have to get back to that some other time.

The op-amp thing is easy. Op-amps have two inputs. One is normal, the other inverts the polarity of the signal.

Without going into the "whys" of the purpose of the inverting input, just recognize an op-amp's inverting input can be used to invert signal polarity at the output of a piece of gear, giving you your balanced signal, and another used at the input of the next piece, to return it to proper polarity.

So the signal gets split. Half runs through the regular input of an op-amp, the other goes through the inverting input. Both get sent down the cable from the outputs.
At the inputs, the normal polarity signal goes through the normal input of an op-amp, the inverted part goes through the inverting input of an op-amp, which returns it to proper polarity. The two signals are then summed.

Beware, this is pretty simplified. There are designs that use one, two, or three op-amps, but this is the basic idea.

 

[travelin travis]

So, using Bennychico11's ingenious drawing technique :

.....noise is the same phase on each lead when entering the balancing circuit:

+)+)+)+)+)+)+)+)+)+)+)+)
-)-)-)-)-)-)-)-)-)-)-)-)-)-)

.....the balancing circuit simply flips the phase of one of the leads:

+)+)+)+)+)+)+)+)+)+)+)+)
+(+(+(+(+(+(+(+(+(+(+(+(

.....the noise on each lead is now 180 degrees out of phase, which cancel each other out......

Is this right? Now I'm wondering how the noise is the same phase on each lead even though the leads are out of phase, before entering the balancing circuit.......and how does the circuit flip the phase of a signal.

thanks Southside Glen

farview, i think i had a good grip on it here.

 

[travelin travis]

I also found this late last night........

http://www.mediacollege.com/audio/balanced/cable-balanced.html

........which confirms what was said here.

 

[CoolCat]

Are you Balanced or Unbalanced?

yeah i was just joking on the upside down mic thing...whooa..i'm not that far out, i don't think?
i think upside down mics also helped drip off the spittle from overly wetted sssshinger'shhh.

benny, yeah i thought a "third" was created to invert the signal one time.
so my mentioned ABC formula wouldn't work for this.
but the op amp mention clarified it for me.

+)+)+)+)+)+)+)+)+)+)+)+)
-)-)-)-)-)-)-)-)-)-)-)-)-)-)

.....the balancing circuit simply flips the phase of one of the leads:

+)+)+)+)+)+)+)+)+)+)+)+)
+(+(+(+(+(+(+(+(+(+(+(+(

this explanation works for me. i guess the ground, third wire never see's any noise? Some of my units use XLR (mic) but the out put is standard 1/4" out (Joe Meek VC1Q channel strip)..goes into my 2488SIAB combo XLR/1/4" jack!

hell..now i have to go stare at all my wires and wonder if my ve's are plugged into the correct Op Amp input!

 

[SouthSIDE Glen]

My problem is I get too many ampersands and tildes in my signal

G.

 

[bennychico11]

i guess the ground, third wire never see's any noise?

the ground wire is just that....a chasis ground. our signal is not carried on it so any noise picked up on it doesn't matter.

 

[ermghoti]

What happens is the the signal on the hot and cold leads are inversions of each other (they are 180 degrees out of phase, so to speak.) The noise on both leads is of identical phase, however since the noise is basically the same.

The "balancing circuitry" simply flips the phase on one of the leads before summing the signals together. This puts the two signals - which were in inverted phase to begin with - back in phase with each other (doubling the amplitude while it's at it when the two now-in-phase signals are summed.)

But since the noise was in phase on the cable, when one of the leads is inverted, the noise on that lead is then inverted as well, So when the leads are summed, the "regular" noise on the one lead is cancelled by the inverted noise on the other.

In short, the key is that the signal on the two leads is in inverted phase on the cable but the noise isn't, So when the circuitry flips the phase off one of the leads, the signlas are now in phase but the noise isn't. So the noise gets cancelled but the signal doesn't.

G.

AH-ha! Now even I understand! Without the balancing circuit, signal coming from the "source" connector would be 180 out of phase, and the cable noise in phase, right? Then when you reverse 'em, the noise is gone, and the signal is back?

 

[Sonixx]

the ground wire is just that....a chasis ground. our signal is not carried on it so any noise picked up on it doesn't matter.

the ground is not required in a balanced design... the ground may be chassis connected or may not...

 

[CoolCat]

is it safe to say that "balanced" requires the OpAmp for the incoming and a dual wire, at minimum, for +ve and -ve?

chasis or eaarth ground being optional, as was stated.

 

[Sonixx]

butting in late...

there's really a lot of mis-info here...

here's the scoop

in a balanced design two conductors carry the the same signal with one being reversed in polarity...

the impedance of each conductor is assumed to be equivalent so stray fields (electric and magnetic) are induced equally on each conductor

at the receiving end, the DIFFERENCE of the two signals determine the final single ended signal.

the math is:
conductor 1 signal is A
conductor 2 signal is -A

so -> (A - (-A)) = A + A = 2A

since the noise is induced equally on each conductor and is not polarity inverted, noise is (N - N) = 0

remember the receiving circuitry calculates the DIFFERENCE of the two conductors.

the ground (shield) generally only needs to be connected on one end. it's actually best if it does not carry any current whch can happen with ground loops.

-keith-

 

[SouthSIDE Glen]

AH-ha! Now even I understand! Without the balancing circuit, signal coming from the "source" connector would be 180 out of phase, and the cable noise in phase, right? Then when you reverse 'em, the noise is gone, and the signal is back?

Kind of. That's the second half of the story.

Think of the "balanced signal" as being created on the sending end. That is the sorurce of the term "balanced"; it basically means that the sending curcuit creates a balanced signal by splitting the original signal into two paths. It sends half as the original signal down the "hot" lead. It then inverts the other half of it and sends that inverted copy down the "cold" lead. This is the "balance", so to speak. The reason the two leads are 180 degrees out of phase is because it is "encoded" that way on the sending end.

Then once the signal gets to the receiving end, there is a complimentary "decoding" curcuit that "re-inverts" the signal on the cold lead and sums it back with the unprocessed signal from the hot lead to get the original signal back again.

It may be more technically correct to say that the op amp calculates the difference, but that can be very confusing to those who don't understand the basics of balanced design to begin with. Since the taking the inverse difference of a value is mathematically the same as summing the value [(A-(-A)) = A + A], it's a bit easier to conceptually explain that way, I think.

G.

 

[Sonixx]

Kind of. That's the second half of the story.

Think of the "balanced signal" as being created on the sending end. That is the sorurce of the term "balanced"; it basically means that the sending curcuit creates a balanced signal by splitting the original signal into two paths. It sends half as the original signal down the "hot" lead. It then inverts the other half of it and sends that inverted copy down the "cold" lead. This is the "balance", so to speak. The reason the two leads are 180 degrees out of phase is because it is "encoded" that way on the sending end.

Then once the signal gets to the receiving end, there is a complimentary "decoding" curcuit that "re-inverts" the signal on the cold lead and sums it back with the unprocessed signal from the hot lead to get the original signal back again.

G.

sorry, but balancd refers to the fact that the impedance is equivalent or balanced in both conductors and has nothing to do with the fact that one of the signals is reversed polarity. this scheme works because of the equivalent impedance of each conducting line and that noise is induced equally in both conductors. the difference the receiver sees is the signal. really no encoding and decoding per say and the signal is not split up and sent down each line. a balanced line can send signals that are not equal and reversed ploarity and still be noise canceling.

 

[SouthSIDE Glen]

Of COURSE the impedance is the same in both conducters. I have heard and read from various "authorative" sources that Balanced refers to that, that it refers to the polarity balance, *and* it refers to the fact that the ground is isolated, depending on the source. The fact is all three concepts have a symmetry, or a "balance" to them, and none of the three have a monopoly on the definition.

G.

 

[Sonixx]

Of COURSE the impedance is the same in both conducters. I have heard and read from various "authorative" sources that Balanced refers to that, that it refers to the polarity balance, *and* it refers to the fact that the ground is isolated, depending on the source. The fact is all three concepts have a symmetry, or a "balance" to them, and none of the three have a monopoly on the definition.

G.

uhhh...

the ground does not have to be isolated at either end, source or receiver...

not sure about your symmetry concept... if the ground is not lifted, then where's the symmetry. it's in the two signal conducting lines. the design is balanced even in the absence of a signal

-keith-

 

[travelin travis]

butting in late...

there's really a lot of mis-info here...

here's the scoop

in a balanced design two conductors carry the the same signal with one being reversed in polarity...

the impedance of each conductor is assumed to be equivalent so stray fields (electric and magnetic) are induced equally on each conductor

at the receiving end, the DIFFERENCE of the two signals determine the final single ended signal.

the math is:
conductor 1 signal is A
conductor 2 signal is -A

so -> (A - (-A)) = A + A = 2A

since the noise is induced equally on each conductor and is not polarity inverted, noise is (N - N) = 0

remember the receiving circuitry calculates the DIFFERENCE of the two conductors.

the ground (shield) generally only needs to be connected on one end. it's actually best if it does not carry any current whch can happen with ground loops.

-keith-

was'nt this already stated multiple times, only worded different?

 

[Sonixx]

it may appear to you that way, but not really.

I felt that some very important points were missed and the descriptions never really got to the crux of the matter... maybe I missed the post, but the real reason Balanced designs cancel noise is due to the equal impedance of each conducting path and that the receiver calculates a difference and not a sum (not a distinction without a difference) and not really due to the polarity inversion of the signals from the source. Psuedo balanced designs can be created using a single ended source and balanced receiver, but not the other way around.

 

[CoolCat]

Signal Noise Cancelling Devices

Actually, Sonixx explanation seems totally different than the "polarity, OpAmp"
explanation to me?? A "difference" not a "Summing" explanation?

Is the Inverted signal created in the amplifier or a specific Output circuit??

Balanced INPUTS & Balanced OUTPUTS on the equipment.
so a Balanced signal could be sent "from" unit#1 to unit#2 "to" a Balanced Input. OK, this is the ideal setup.

So what if one piece of equipment in the chain isn't Balanced???

Ok, Its Saturday, I'm going in to check my cables....Fire IN the HOLE!!!
If I don't come back out Balanced, well... it was swell.


......and this has nothing to do with Chasis NOISE or AC Line noise...does it?
Which is probably more common a problem.