XLR Cables - Technical Discussion

From Bose Portable PA Encyclopedia
Revision as of 13:51, 4 January 2007 by DanCornett (talk | contribs)
Jump to: navigation, search

What does it mean to have "balanced" vs. "unbalanced cables and connections?

Balanced: What this really means is that there are TWO wires carrying a varying signal, not just one. There has to be a "reference", which is provided by the 'ground' -- hence a third wire. The advantage of a "balanced" signal is that one can get a better "signal-to-noise" ratio because the receiving amplifier is looking at the difference between the two signal wires, rather than just the difference between one wire and "ground". That's why a "balanced" system (balanced output, 3-wire cable, balanced receiving input) can be better at rejecting any "noise" which might exist between the 'sender' and the 'receiver' -- and the longer the cables, the more of a problem the "induced noise" might be.

All those 'features' cease to exist when one end or the other is 'unbalanced' -- when only one wire (signal) is compared to the reference (ground).

Unbalanced is NOT bad ... balanced signals are just one way to help deal with "difficult" situations.

By the way: with balanced cabling, doing a "ground lift" means that the "third wire" is removed, and thus the 'sender' and 'receiver' are comparing the two "signal wires" to their separate local "ground" reference, rather than sharing a common "ground" reference. If the two local "ground" references are too different, that can cause safety and/or signal quality problems -- which is why 'ground lift' has to be done with care, and is typically a "last resort" solution.

Shielding: Shielding is a means of reducing, or "side-tracking", or "shunting aside", any unwanted signals (induced noise) BEFORE it has a chance to affect the signal-carrying wire(s). The shield, thus, tries to act like an "antenna" for any external electrical signals, but it's "output" is shunted aside to the "ground" before it has a chance to affect the signal wires.

For this "antenna effect" to work, the cable must only have its shield connected at one end, not both. This *is* common practice with most good XLR cables, but it is far less common with 1/4" cables. That's why XLR cables are generally preferred for long distance runs ... even if being used for an "unbalanced" signal. Unless you take apart the 1/4" connectors at both ends to see how the shielding is connected (or not), you really can't reliably depend on mixing-and-matching 1/4" cables ... but you can (generally) depend on good XLR cables to be properly shielded.

===========

There are a lot more subtleties about cabling, but that's the (non-pictorial) essence of the wiring terms "balanced/unbalanced" and "shielded".

===========

If the ground is common to both pieces of connected equipment, why would it matter if both ends of a shield were connected? Now I'm confused.

You're not alone ... the difference between shield and ground can seem rather esoteric, but -- if there is a lot of "electrical noise" in the vicinity of the cables, it's an important factor.

The basic problem is that the "power ground" (also know as "chassis ground") can fluctuate (vary) -- even within a single piece of equipment ... but with microphone-level signals, that variation can show up as 'noise'. Therefore, circuit designers often put some effort into ensuring that the "signal ground" used by the electronics is stable and "clean".

When it comes to connecting the shield and signal ground, there are 4 variations -- all with advantages and disadvantages, as succintly explained here. Note that the primary "concerns" of that author seem to be frequencies ABOVE audio frequencies (100kHz and up).

Even manufacturer's literature is not always consistent about this. For example, this otherwise good Rane article on Grounding and Shielding has some sections where it seem rather confusing whether the reference is to the cable shield or the equipment shield. However, Figure 5 in that article is somewhat useful, showing all the grounding permutations and which are helpful for audio. Notice, however, that Rane article is refering to a TWO WIRE cable, not a three-wire cable! Two wire cables are common with TRS (e.g.: 1/4" or 1/8") cables, but some XLR (mic) cables are three-wire cables. Remember that TRS cables can be used to carry either stereo signals (as in most home audio equipment) or a mono-but-balanced signal (as found, for example, with some electronic keyboards) -- that can further confuse the issues of the "ground" vs. "shield" and "balanced" vs. "unbalanced".

As an aside: VIDEO cabling is almost always a single conductor with one (or more!) shield(s). The shield(s) in these cables act as both the signal reference/ground and as an interference shield.

So, an XLR-type connector and a TRS connector can both carry a balanced signal: two "hot" wires and a ground/reference connection. Where they differ is in how they can handle any shielding around the wire.

Physically, the XLR connector can provide FOUR different connections: the 3 pins and the external metal "shell". A TRS connector can only provide 3 connections. That's where differences in handling the ground vs the shield happen.

So, with XLR connectors, the issues are: what, if anything, is connected to the external "shell"? Is there a separate wire for the third pin (pin 1)? Should the cable shield (or chassis ground) also be connected to that third pin?

To further confuse things, the Wikipedia reference to EIA Standard RS-297-A says that third pin (actually, numbered pin 1) is used for chassis ground -- but doesn't reference the SIGNAL ground at all -- but one of the referenced links points to a manufacturer's diagram which shows pin 1 as a SIGNAL connection, and the shell is not connected to chassis ground!

Ideally, the external shell is connected to the cable shielding at both ends, and is connected to an independent "shield ground path" on the equipment at both ends, so that the "antenna" (induced) currents from that cable shield never enter the "reference ground" (signal ground) used by the circuitry at each end of the cable.

However, as a user -- do you know what the manufacturer has done with the "shell/chassis" ground and the "signal ground"??? Not knowing that ... and knowing that manufacturers disagree on the best practices for doing so ... one can get all sorts of variations of grounding, as in Figure 5 in that Rane article.

===========

So, other than throwing up one's hands in frustration, what can we non-wiring-technician end-users do?

My recommendations are to use XLR-to-XLR for balanced signals whenever radio noise or hum are a problem. If you need to use TRS connectors for the equipment connections at either end, ... and a simple TS line-level cable isn't working adequately ... then use a TRS-to-XLR (or TS-to-XLR) device (a.k.a. a "direct box) along with XLR cables. If that TRS-to-XLR device has a "ground lift", then you should be able to deal with any 'hum'/ground-loop situations.

Ideally, use XLR cables which have 3 wires connected to the three pins at each end, and the cable shield is connected to the shell at one end. If it is connected to the shell at both ends, that's o.k. -- but preferably not also connected to "pin 1" at both ends. You can't change what the manufacturers do inside the box, but you can control the type of cable used.

The best mic / XLR cabling I've used actually has 8 internal wires, configured as 4 twisted pairs with those 4 pairs ALSO twisted into pairs, plus a shield. Twisted pairs of wires are an additional means of rejecting induced noise (which is why most cables with 2 or more wires are typically "twisted" together). The XLR connector has each signal pin (2 & 3) connected to both wires of a twisted pair, rather than a single wire. Pin 1 (signal ground) is connected to the remaining 4 wires (the other pair of pairs) and the shield is only connected to the shell. Relatively expensive to make, but utterly reliable.

Here is a pointer to a commentary on the standards -- even the professionals don't agree on the best and safest way to handle XLR connections, grounding, and shielding. (Comments on shielding/grounding standards.)