Why balanced XLR needs an inverted signal?

[batata004]

Hello dear friends,

I understand that the balanced XLR needs an inverted signal wire (pin 3) so at the end, in the mixer, the pin 3 is inverted again and summed with the pin 1. This will make any noise that got picked up in the cable to get subtracted from pin 1 and pin 3. Awesome idea!

But why do I really need to invert the signal on pin 3? Cant I just connect pin 3 to GND and still have a balanced connection? If pin 3 is connected to GND and it gets any noise along the cable, then at the mixer this noise will be inverted and summed with the noise+signal from the pin 1. The result: the signal from pin 1 will go through, the signal from pin 3 will have no effect since it was grounded BUT the noise at both pins will be subtracted because if any noise was generated on pin 1, it will also be generated on pin 3, and since pin 3 will be inverted at the mixer, then this noise will simply disappear when pin 1 and pin 3 get summed.

I know what I am saying above is wrong. I know that to have a balanced connection I really need the inverted phase on pin 3. But why?

One second question: some people say that if I cant invert the phase, I should not connect the pin 3 to GND because it makes noise worse. BUT if I dont connect pin 3 to GND then the signal on the pin 3 will "float", it will not be a stable 0 signal. If the signal on pin 3 floats, then when it gets inverted and summed with pin 1, a lot of "noise" will be heard but this noise will not be due to eletromagnetic field, but purely because pin 3 was floating. Right?

 

[Ultrasonic]

At the receiving end, one signal is subtracted from another to reject noise that is the same on the two lines. If you don't want the actual signal to be cancelled out too you need one of the signal lines to be inverted.

 

[batata004]

At the receiving end, one signal is subtracted from another to reject noise that is the same on the two lines. If you don't want the actual signal to be cancelled out too you need one on the signal lines to be inverted.

I think you didnt understand what I asked. Of course I need a signal on pin 1. But I am questioning why do I also need an inverted signal on pin 3? If it's just because to provide noise subtraction, than it clearly wouldnt be necessary an inverted signal on pin 3 since the noise will automatically be "zeroed" (subtracted) in the mixer, no matter if there is an inverted signal on pin 3. That's why I am asking: why do I need an inveted signal on pin 3 since it looks completely unnecessary if the only reason is to get noise subtracted.

 

[DVDdoug]

You are right and with an unbalanced signal into a balanced (differential) input you retain most of the noise rejection advantages of a fully-balanced connection.

I actually made something for a "special application" with an RCA connection into a differential amplifier (with no ground). It solved my noise problem.

One second question: some people say that if I cant invert the phase, I should not connect the pin 3 to GND because it makes noise worse.

It depends on the circuit. Depending on how it's wired you might defeat the differential input. The shield should be grounded at one end or the other or both.

I know that to have a balanced connection I really need the inverted phase on pin 3. But why?

It's not actually "balanced" without the two opposite signals.

 

[Ultrasonic]

I think you didnt understand what I asked. Of course I need a signal on pin 1. But I am questioning why do I also need an inverted signal on pin 3? If it's just because to provide noise subtraction, than it clearly wouldnt be necessary an inverted signal on pin 3 since the noise will automatically be "zeroed" (subtracted) in the mixer, no matter if there is an inverted signal on pin 3. That's why I am asking: why do I need an inveted signal on pin 3 since it looks completely unnecessary if the only reason is to get noise subtracted.

I did answer your question but maybe adding equations will help.

Say the signal on pin 1 is: P1 = S + n, where S is the signal and n is the (common) noise. Then correspondingly P3 = -S + n. (The '-S' being the inverted signal.)

At the receiver, the pin 3 signal is subtracted from that on pin 1: P1 - P3 = (S + n) - (-S + n) = 2S.

If the P3 signal wasn't inverted then P1 - P3 would equal zero.

 

[TheKoC]

I did answer your question but maybe adding equations will help.

Say the signal on pin 1 is: P1 = S + n, where S is the signal and n is the (common) noise. Then correspondingly P3 = -S + n. (The '-S' being the inverted signal.)

At the receiver, the pin 3 signal is subtracted from that on pin 1: P1 - P3 = (S + n) - (-S + n) = 2S.

If the P3 signal wasn't inverted then P1 - P3 would equal zero.

Wouldn't it be (S +n) - (0 - n) = S?

I don't think a differential/inverted signal on P3 is necessary for the noise cancelation, you can also have no signal on it

 

[Ultrasonic]

Wouldn't it be (S +n) - (0 - n) = S?

No. Firstly because that doesn't have an inverted signal and secondly because the result of that equation is S + 2n, so you'd have reduced the signal to noise ratio, not improved it (which is the goal).

 

[batata004]

I did answer your question but maybe adding equations will help.

Say the signal on pin 1 is: P1 = S + n, where S is the signal and n is the (common) noise. Then correspondingly P3 = -S + n. (The '-S' being the inverted signal.)

At the receiver, the pin 3 signal is subtracted from that on pin 1: P1 - P3 = (S + n) - (-S + n) = 2S.

If the P3 signal wasn't inverted then P1 - P3 would equal zero.

@Ultrasonic Thanks for your equation but I think you are wrong, respectfully.

If P1 = S + n (as you said) and P3 = n (in the question I say I am connecting P3 to GND so there will be absolutelly no signal), then at the mixer the P3 will be inverted, resulting:

P1 - P3 = S + n - n = S
Clearly, you wouldnt need an inverted signal on P3 if the only reason is removing the common noise.

@DVDdoug thanks for your reply! I am glad you fixed the noise problem on your "special application" without connecting PIN 3 to GND! But would you know why it worked, because it should have created a ton of noise, right? See: if you leave a wire (the pin 3 wire) connected to nothing, it will become an antenna and will pick a lot of noise, I mean, a lot! Every nano second the signal will float a lot. This "float" condition will create a lot of "signal" in the pin 3 and when it gets inverted and summed with pin1 it will create a lot of noise. That's why I dont understand how leaving the pin 3 floating (not connected to GND) also works fine - in my opinion it should not work (of course I am wrong because in your "special application" it worked just fine.

 

[TheKoC]

No. Firstly because that doesn't have an inverted signal and secondly because the result of that equation is S + 2n, so you'd have reduced the signal to noise ratio, not improved it (which is the goal).

oh sorry, i meant to write (S +n) - (0 + n) = S

 

[Ultrasonic]

@Ultrasonic (in the question I say I am connecting P3 to GND so there will be absolutelly no signal),

Oh, well that's different but you asked why balanced XLR needs an inverted signal and I answered that.

 

[Ultrasonic]

oh sorry, i meant to write (S +n) - (0 + n) = S

Got you. I think the answer is that where cancellation isn't perfect the fact that using an inverted signal gives 2S rather than S will further help to increase signal-to-noise ratio.

 

[Ultrasonic]

Having a connections between ground and pin 3 will mean there is a path from the cable screen to the pin 3 input too, which would rather defeat the point of the screen...

 

[DonH56]

There is some description and discussion in this thread:

Differential, Single-Ended, Balanced, Unbalanced, and all that jazz...

I posted this picture over on WBF in the middle of a long semi-fruitful discussion on the benefits and drawbacks of differential operation. Ken Newton suggested sticking with the accepted technical definitions and distinction among differential vs. single-ended and balanced vs. unbalanced that I...

www.audiosciencereview.com

Shorting one side of a balanced/differential signal to ground makes it unbalanced or single-ended so you lose the benefits of differential operation (you revert to the single-ended picture in the referenced thread). Shorting an input is usually OK but it depends upon the receiver circuit. Shorting the output to ground is undesirable, but again depends upon the internal circuit. The rule of thumb is that leaving the inverting output source floating and grounding the receiver's inverting input is usually acceptable (circuit-dependent), but you lose common-mode noise rejection, reduce SNR, and increase distortion by operating a balanced circuit single-ended.

 

[Speedskater]

XLR pin 1 is the shield not a 'ground". True in most balanced output stages it has continuity to the audio circuit common.
So in these cases, pin 2 and/or pin 3 to pin 1 may be half the signal level of pin2 to pin 3. But it's not true in some cases.

 

[JIW]

Pin assignment for audio XLR3 is 1 = Ground, 2 = Positive, 3 = Negative. If pin 3 is connected to ground on both devices and there is a difference in ground voltages, there will be ground loop noise added to the signal.

 

[Hayabusa]

Wouldn't it be (S +n) - (0 - n) = S?

I don't think a differential/inverted signal on P3 is necessary for the noise cancelation, you can also have no signal on it

Indeed not needed, but to make the noise reduction work well the noise needs to be equal on both lines.
If the source side is directly connected to ground this will not be the case as the other line sees the actual output impedance of the source.
To overcome this put a series resistor (equal to your source output impedance )to that ground connection so both lines 'see' the same impedance.

 

[Gorgonzola]

The Single-end to Balanced and vice versa tends to be confusing for us non-technical types.

My current issue is connecting my WiiM Ultra to my Purifi 1ET400A amp with only XLR (balanced) inputs. I consulted the long-standing RANE interconnection guide for guidance. Extracting from that comprehensive guide I extracted the following info relevant to my issue ...

For this SE to Balance problem I see the recommended solution is to us a two-wire plus shield cable and

1. Connect the out active (red) wire to the #2 (positive) pin on the male, balanced-end connector;
2. Connect the output neutral (black) wire to the #3 (negative) pin;
3. Connect the shield (green) wire to the #1 (ground) pin.

In practical terms a I bought a SE to XLR male interconnect which, I presume, implements that solution.

 

[DonH56]

The Single-end to Balanced and vice versa tends to be confusing for us non-technical types.

My current issue is connecting my WiiM Ultra to my Purifi 1ET400A amp with only XLR (balanced) inputs. I consulted the long-standing RANE interconnection guide for guidance. Extracting from that comprehensive guide I extracted the following info relevant to my issue ...

View attachment 435252

For this SE to Balance problem I see the recommended solution is to us a two-wire plus shield cable and

1. Connect the out active (red) wire to the #2 (positive) pin on the male, balanced-end connector;
2. Connect the output neutral (black) wire to the #3 (negative) pin;
3. Connect the shield (green) wire to the #1 (ground) pin.

In practical terms a I bought a SE to XLR male interconnect which, I presume, implements that solution.

Note that is from a SE driver (output) on the left to a balanced source input on the right. It is not appropriate going the other way.

 

[somebodyelse]

I'd suggest these as required reading - and for me at least re-reading a few times before the details sink in:
Bruno Putzeys' "The G Word" - it's appeared in various places, including as a 4 part series at EDN and a pdf at Hypex. It's an attempt to dispel the myths and correct common implementation mistakes, finishing with an implementation example that some have built as a basic pre-amp.
Bill Whitlock's application notes at Jensen Transformers - my usual go-to is AN007 as having comprehensive coverage.

If there wasn't scope for confusion there would have been no need for AES48, and even with that people are still making products that have problems.

 

[dfuller]

Well, that's the thing - it actually doesn't. It just needs hot and cold to have equal impedances. There can be nothing at all on the cold pin but noise and it works.