I am curious to know why won’t manufacturers follow the AES48 standard. Do you think Audio Research, Mark Levinson, Cary Audio, Krell, and most of BAT all (Balanced Audio technology) don’t follow this standard? For the kind of money these guys charge, I really doubt they will not adhere to the standard. But then how do you find out whether a particular manufacturer adhere to the standard of not?
@pwerahera , @inna got that partially right; you can build a balanced output that does not conform to AES48 easily and it can perform quite well.
To support the standard, there are 3 ways I know of. The first is to use an output transformer and that’s how a great deal of studio equipment does it to this day because transformers do that really well and at the levels required can also have good bandwidth.
The second way, if you are going solid state (or if you don’t mind a hybrid circuit) is to use an IC like this one.
There is a third method that involves the use of a balanced circuit known as a Circlotron.
Each of these techniques have their strengths and weaknesses. Transformers are expensive and you have to pay attention to loading them properly to prevent ringing or bandwidth issues. The IC chip isn’t going to do it if you want to have an all-tube embodiment. The third way is also more expensive and to do it right probably means the manufacturer would have to pay a royalty.
So, since cable manufacturers don’t mind making expensive balanced cables, most of high end audio has chosen to ignore the standard.
With all respect to Ralph, that’s a circular argument and the claim not quite exactly true. The AES48 standard he touts was only adopted in 2019, and balanced audio circuits predate that by many decades. To suggest that only those complying with the standard are capable of such feats as eliminating ground loops is just not accurate.
@cleeds This statement is incorrect. AES48 was updated in 2019 but existed long before that and all it did was codify the existing practice. One of those practices is to ignore ground when an input or output is used (IOW the ground is only for shielding and not to complete the circuit, unlike single-ended connections). That practice goes back to the early 1950s. If you take a look at this Ampex 351 schematic you’ll see that the microphone input uses a transformer and neither side is grounded, nor is there a grounded center tap.
I would be interested to know exactly how my argument you quoted might be considered ’circular’.
In case it wasn’t clear, sure, you can run equipment that is balanced and does not support AES48, and it can sound fine. But you’ll have to audition the interconnect cable to really winnow out the best performance from the equipment. As I pointed out yesterday, you’ll never succeed at that last bit, because the best cable you were able to find will still not be right and I explained why. The equipment will also be susceptible to ground loops (which should never happen with balanced equipment...), and if you’ve wondered why some people say single-ended connections sound as good or better, its because the balanced standard is being ignored.
The balanced line standard, as it existed in 1958, made possible the golden age of stereo, along with the Westerex 3d stereo LP cutter head, and is this (long before AES48 codified it):
1. Pin 1 is ground, pin 2 inverting, pin 3 inverting (in Europe pins 2 and 3 are reversed)
2. Pin 1, ground, carries no signal circuit and is ignored by the signal circuitry. This means that the pin 2 signal is generated with respect to pin 3 and vice versa.
3. For each side of the signal, pin 2 and 3, if there is an impedance to ground, it will be an equal impedance on each side.
4. There is a low impedance aspect; in 1958 the termination standard was 600 Ohms for line level (150 Ohms for microphones).
IMO/IME if you really want cable immunity, your equipment should be able to drive loads as little as 1000 Ohms. This last bit requires a low output impedance from the circuit driving the cable. The chip I linked above has no problem doing this and you’ll notice that balanced line transformers are often designed for low impedance operation. This impedance helps swamp artifacts caused by inductance or capacitance in the cable.
BTW, this standard is used to prevent cables having a ’sound’ so you don’t have to compare cables to get things to work- its meant to allow plug and play. It has nothing whatsoever to do with how that equipment itself actually ’sounds’; your allusion to some of it sounding like ’drek’ is a red herring.