Balanced vs. Unbalanced Inputs

This statement is a common myth, and being a myth, is false. A center tap degrades the Common Mode Rejection Ratio, on account of the simple fact that no center tap is really centered exactly. For this reason, center taps are never used. This is true of dynamic mics, cartridges, tape heads and any other inductive source including input and output line transformers.

I never said that center tap would be better for noise rejection (though it absolutely could be). You appear to have have misinterpreted my use of center tap to imply a single ended ground reference, which is not the same as connecting the shield to a center tap to balance the signal w.r.t. the shield. That it is not used in audio (not needed) does not mean it does not work. It is used in other industries because it does.


https://www.pulseelectronics.com/wp-content/uploads/2016/12/G019.pdf

https://patents.google.com/patent/US20170131427A1/en

Certainly noise rejection is one aspect, but so is the rejection of ground loops (that is why the balanced line system ignores ground).

Grounds loops ARE common mode noise. Hence, the accuracy of my statement, to reduce common mode noise.

The problem here is that in order to obtain the benefit of balanced operation, your equipment must support the standard, AES48. Having introduced balanced line operation to high end audio, I can tell you that hardly any high end audio gear supports the standard and so you hear differences, and often degradation.

We are talking input to phono stages, i.e. cartridges, so not sure where you are going bringing this up. It is not applicable.


At this point we need to be pedantic since terms that are not interchangeable are being used that way, "balanced" refers specifically to the interconnect method, i.e. the signals are impedance balanced to the ground/shield, hence the noise on one is balanced and opposite to the other. That could be implemented with transformer coupling and shielded cable, or differential signalling and a shielded cable.
What is called a "balanced" input is balanced in the sense that it connects to a balanced interconnect, but electrically it is differential. The AES standard is a differential signalling standard coupled with a balanced interconnect.

Since the cartridge is already a balanced source, its a matter of getting the signal to the preamp without involving that signal with the ground system (the tonearm tube and turntable).

Which makes this statement as noted before inaccurate.  A cartridge is not a balanced source. It is a floating differential source.  If you twist the wires and maintain a consistent (w.r.t. each wire) shield to the amplifier then the interconnect would be balanced.

Because the cartridge is floating, and hence its output differential, going into single ended input, assuming you use the practices of balanced interconnection (with shielding), you can still have a low noise floor especially at low frequencies, because the common mode noise at those frequencies will be rejected. However, from a practical standpoint, this does not work well at high frequencies, due to parasitics. This is where the differential input comes in that will reject common mode not just at low frequencies but at high frequencies.

Unfortunately, too often balanced and differential are thrown around interchangeably and they are not and mean different things. This is confusing especially if you are trying for a deeper understanding.

See my previous two posts, clearthinker


The AES standard is balanced interconnect with differential signalling AND the source and load and cable impedance is defined. Really, at least to me, it is that defined impedance that justifies most of atmasphere's argument, with the balanced and differential taking care of the rest.

Defined and low impedance compared to consumer equipment impedance ensures that the relatively low capacitance and inductance of interconnect cables are essentially eliminated from the equation. Matching the cable impedance to the source/load impedance guarantee fast signal transmission with limited ringing and also makes resistance purely a level shift. By fast, we are talking so far beyond audio it is not worth even mentioning.

Balanced interconnection with differential signalling effectively eliminates noise, the only purported benefit of expensive cables.

Ooooh, I can see that one is going to be controversial. You seem to be saying SQ differences between different cables arise from running single-ended, as most audiophile systems do, and not from the cables themselves.
Is it going too far to draw some further conclusions from that statement:
1. All well-designed cables sound (much) the same in a balanced set-up.
2. Money spent on expensive cables is (largely) wasted in a balanced set-up. Therefore:
3. There are SQ differences due to cables in a single-ended set-up so expenditure on cables can be justified. But if you choose a balanced system, this will gain at least the same SQ improvements as buying expensive cables, but without the extra cost. Pace - I accept in principle it is more costly to build balanced amplifiers than single-ended.


C.  "your equipment must support the standard, AES48; I can tell you that hardly any high end audio gear supports the standard "

Are you saying that some amplifier hardware fitted with XLRs and said by the manufacturer to be 'fully balanced', is not fully balanced and does not support AES48? Presumably XLRs could be put on for show without the correct wiring behind? If this is correct, someone suitably qualified should start naming names.

I never said that center tap would be better for noise rejection (though it absolutely could be).

It would only be truly differential if you center tapped it and connected the shield to the center tap, such that

The parenthetical remark renders the first statement false. A center tap decreases noise rejection, at least at audio frequencies.

The second quote is certainly suspect. There is no attempt to make the connection differential, but to receive the signal properly differential techniques should be used if a high CMRR is to be expected.

Ooooh, I can see that one is going to be controversial. You seem to be saying SQ differences between different cables arise from running single-ended, as most audiophile systems do, and not from the cables themselves.
Is it going too far to draw some further conclusions from that statement:
1. All well-designed cables sound (much) the same in a balanced set-up.
2. Money spent on expensive cables is (largely) wasted in a balanced set-up. Therefore:
3. There are SQ differences due to cables in a single-ended set-up so expenditure on cables can be justified. But if you choose a balanced system, this will gain at least the same SQ improvements as buying expensive cables, but without the extra cost. Pace - I accept in principle it is more costly to build balanced amplifiers than single-ended.


C.  "your equipment must support the standard, AES48; I can tell you that hardly any high end audio gear supports the standard "

Are you saying that some amplifier hardware fitted with XLRs and said by the manufacturer to be 'fully balanced', is not fully balanced and does not support AES48? Presumably XLRs could be put on for show without the correct wiring behind? If this is correct, someone suitably qualified should start naming names.

Yes, cable manufacturers and inexplicably, many audiophiles, don't like to hear this stuff. But there is so much product out there that does not support the standard that there will be plenty of need for exotic balanced interconnect for a long time.


To be clear, I am saying that SQ differences between different cables are audible if there is no termination standard and if there are signal return currents present in the shield. So this can mean both single-ended and balanced, if the latter is improperly executed. Since that happens a lot in high end audio, we have a high end cable industry making expensive balanced line cables! Quite the opposite of how it should work.

Regarding your numbered points:
1), 2), 3) all Yes, if AES48 is observed.
C. Quite a lot of high end audio hardware has XLR connectors and are not balanced at all; the connectors are there for convenience only. Some VTLs and Lamm products are set up this way (examples I have had contact with did not have pin 3 connected to anything in the amplifier; this will introduce a buzz if the source driving that amp meets AES48 as the circuit is incomplete). Other high end audio products have XLR connectors and are fully balanced (such as ARC and mbl) but don't support the standard regardless. The mbl amp I had contact with was balanced because when it was monostrapped, you could then use the XLR input (meaning you needed two of these amps for stereo). This approach causes signal return currents to be present in the shield and results in a very low CMRR, making the connection susceptible to noise pickup and cable construction. The ARC stuff I've had contact with has had fairly high output impedances, preventing the preamp from driving anything less than about 30,000 ohms. The output was two single-ended stages that were driven out of phase with each other. This causes signal return currents to be present in the shield, and the relatively high impedance of the resulting system causes the setup to be sensitive to cable construction.

Thank you again atmasphere.

Given the haphazard wiring behind the XLRs and non-adherence to AES48 of many amps, I am indeed lucky (just that) to find my system is free from hum and other obvious non-matching artifacts and appears to have a very low noise floor - I hear nothing on no-signal until I turn the volume pot up a long way.

For reference, I (believe) I run fully balanced throughout.  Ortofon A90 on a Simon Yorke Aeroarm though a bespoke SY connection box into a van den Hul The Grail SB phono amp, Audio Research Ref 6, vintage Krell KRS200 monoblocks.

It seems wrong that getting the benefits of running fully balanced should be such an uncertain adventure.

And if nothing else, these exchanges have revealed the truth about the alleged benefits of high-end cabling.

It seems wrong that getting the benefits of running fully balanced should be such an uncertain adventure.

+1 Sure got that right. The point of the standard was/is plug and play.