Baldo wrote:I recently saw a microphone lead in a cheap shop in the east end of London. At one end was a female xlr jack and at the other end was a male 1/4inch jack plug, Mono.
Yes, it's a widely available adaptor cable intended for connecting a balanced dynamic mic to an unbalanced input, such as on a guitar amp or low-end mixer amp. It's not a balanced cable and the wiring intentionally unbalances the source.
Now my question is, should the jack plug not have had 3 connectors on it?
If it was intended to connect a balanced output to a balanced input then that would be more conventional, yes... XLR to TRS plug balanced adaptor cables are widely available too. And in general, as Mike says, proper balanced audio cables normally have three terminals.
My understanding of balanced leads is that there are two signals being transmitted, one is the phase inverted version of the other.
That's a common, but completely erroneous, (mis)understanding!
A 'balanced' connection comprises two signal wires (often referred to as 'hot' and 'cold'), neither of which are connected directly to ground. Often a protective, grounded electrostatic screen or shield encases those two signal wires, but it plays no part in the signal transfer and is frequently omitted, particularly when the signals in question are line level. (Most traditional professional patch-bay plug cords (the PO316 and TT types) aren't screened and only have two internal signal wires).
But assuming the screen is present, then the connectors require three terminals for the hot, cold and screen connections.
However, a 'balanced connection' is only 'balanced' if the impedance to ground for each signal wire (at each end) is identical. It has absolutely nothing whatsoever to do with the inverted symmetrical signals!
A balanced line works as a form of Wheatstone Bridge, and that impedance balance is the critical part that makes it work, not whether the audio is passed as a symmetrical inverted form or not!
The key benefit of a balanced line is that it rejects interference, but it can only do that if the interference signal generates what is called a 'common-mode' signal in the cable. As the name implies, a common-mode signal means that there is exactly the same voltage on each signal wire -- and you can only get the same voltage on each wire if the impedance to ground is identical for each wire too -- because of Ohms Law (V=IR).
The balanced input receiver works as a differential receiver, which means it looks for the difference voltage between the hot and cold wires. With a common-mode signal, the voltage on each wire is engineered to be the same because the impedance to ground for each wire is the same, so there is no voltage difference between the two signal wires. Consequently, there is no output signal from the differential receiver.
And that's how the infamous interference suppression of a balanced line works.
The cable screen -- if present -- helps to reduce any electrostatically-coupled interference from reaching the signal wires, but whatever interference does get through is forced to create a common-mode voltage because the signal wires have balanced impedances to ground. Electromagnetically-coupled interference is reduced by twisting the two core wires around each other, so that the magnetic field induces the same current into both wires equally. The tighter the twist, the better the interference rejection, and in extreme cases, the star-quad format works even better...
So the cable screen and the core twisting help to reduce external interference, but it's the impedance balancing that forces any successful interference to generate a common-mode signal which can then be rejected by the differential receiver!
As for the transfer of wanted audio signal, that has to be present across the two wires as a differential voltage signal. The original way to do that was by using a (floating) transformer output, and that does indeed produce the symmetrical and inverted signal format that you're thinking of.
It's still only one signal though (not two), effectively with half of it carried on the hot wire, and the other half on the cold wire. As far as the differential receiver is concerned, it looks for the voltage difference between the two wires which is the full signal. The simple maths is: Output = +0.5 - (-0.5) = +0.5+0.5 = 1.
Most active balanced outputs replicated the symmetrical inverted signal format too, mostly because it has some advantages in terms of headroom and signal-to-noise. But it doesn't have to work that way... and this is where the so-called 'impedance-balanced' interface format comes in. This type of interface is extremely common in budget and mid-level mixers, interfaces, and microphones (including some very high-end ones), and it is designed to send the full signal voltage on the hot wire, with (apparently) nothing on the cold -- although the connection is still required, of course to maintain common-mode signals!
As far as the differential receiver is concerned, it still sees a differential signal of the full amplitude between the two signal wires, so all is well. And, because the impedance to ground for each signal wire is still properly balanced, the common-mode interference rejection still works as perfectly as any old-school balanced interface.
But the benefit is that only one active output driver is required, instead of two, and if the output happens to be connected to an unbalanced input the full signal level is still transferred. This is particularly helpful in semi-pro situations where the equipment could be connected to either balanced or unbalanced equipment, or with balanced or unbalanced cables -- as is common in semipro applications.
In contrast, if a symmetrical inverted balanced output is connected to an unbalanced destination only half the signal will be received*, and so the output will be 6dB quieter than expected. (*There are ways around that using floating transformer outputs or cross-coupled active outputs... but these are both relatively rare these days).
I hope that clears up the 'it's only balanced if carrying symmetrical and inverted signals' myth!
Going back to the required number of terminals, yes most balanced audio interfaces use three... but the screen connection is an optional extra nicety in the majority of cases and not mandatory (although it is required for the phantom power ground reference with power mics, of course).
In other (non-audio) applications it is perfectly acceptable and very common to pass a balanced signal with just two terminals -- as each data pair in a Cat5 cable does, for example, or the plain old telephone wiring....