Welcome to No Limit Sound Productions

Company Founded
2005
Overview

Our services include Sound Engineering, Audio Post-Production, System Upgrades and Equipment Consulting.
Mission
Our mission is to provide excellent quality and service to our customers. We do customized service.

Friday, February 28, 2020

The Low-down On Analogue Interfacing

By Hugh Robjohns
The Low-down On Analogue Interfacing
With so much focus on computer workstations these days, wiring up analogue gear is becoming a bit of a lost art, but don't fear: all you need to know is here.
Despite the prevalence of digital audio in today's home studios, we still rely on analogue connections at some points in our signal chains. Analogue interfacing formats are stable and mature, but nevertheless it's surprising how often I am asked questions about connecting. This article is therefore intended to serve as a definitive guide to the basics.

Levels

The first thing to deal with is the topic of analogue signal levels. There are, fundamentally, five distinct nominal signal levels to be aware of. Starting at the lowest, we have 'microphone level' — the typical level you could expect to come from a typical microphone placed in a typical position in front of a typical instrument. Of course, there is no such thing: the actual level can vary wildly from microvolts (if you put a ribbon microphone at the back of a hall while someone plays quietly on a piano), to whole volts (from a capacitor mic inside a kick drum)! In general, though, most people would consider something like -50dBu to be a typical starting point for a nominal microphone level, and that equates to about 2.5mV rms. The 'rms' stands for 'root mean square,' a shorthand way of conveying the slightly involved method of describing an 'average signal voltage,' which is necessary because we are talking about varying AC signals here, rather than a steady-state DC voltage.

Hopefully, the use of the decibel is already familiar to most readers. The decibel is simply a logarithmic ratio of two signal amplitudes: one is the signal we are interested in, and the other is a stated reference value. The standard reference in professional audio is 0.775V rms, and we use the term dBu to denote it. The 'u' means 'unterminated' and refers to the absence of 600Ω terminations that used to be mandatory back when dinosaurs ruled the audio industry!

In broadcasting circles, the nominal 'line level' is called the standard 'alignment level,' and is defined as 0dBu, or a signal of 0.775V rms. However, most professional audio equipment actually employs a slightly higher nominal line-level of +4dBu, or 1.223V rms. You will often see and hear references to professional equipment as having "+4" interfaces.

Semi-professional and domestic equipment is generally designed to use a lower standard line-level, which is described as "-10." However, there is a trap here: a different level reference is used — this is -10dBV. The 'V' denotes a reference of one volt instead of 0.775V, and that means that the semi-pro line level of -10dBV equates to 0.316V.

The important practical point of this is that a -10dBV signal is almost exactly 12dB lower than the +4dBu professional line level. Which explains why, if you connect a semi-pro device to a professional one, the average signal level is about 12dB quieter than expected and, going the other way, a professional signal peaks roughly 12dB higher than a semi-pro device expects!
Another important signal level to be aware of is the maximum 'headroom'. Well designed professional equipment will run out of headroom and start to clip at about +28dBu, which means a whopping signal level of about 20V rms — although I've seen some really impressive equipment that can cope with more than +36dBu! Less well designed audio devices usually have lower headroom margins. For example, +22dB is fairly typical, and semi-pro devices will have even lower margins because of their lower nominal line level.

Balanced Or Unbalanced?

The next key concept is that of the balanced interface, which is probably the most misunderstood concept in the entire world of audio! Most people know that a balanced interface involves three wires: hot and cold signal wires, plus a screen connection; and an unbalanced interface has just two: signal and screen/ground. Most people also know that whereas the screen/ground is essential in an unbalanced connection because it provides the signal return path, it plays no part in the signal path for a balanced interface, and is purely a screen to reduce interference. That's why you can remove the screen connection of a balanced interface to prevent a ground loop, without losing the signal.
The familiar XLR, used extensively in professional audio. It offers a sturdy connection, and because it locks in place it is difficult to disconnect accidentally, which makes it perfect for connecting devices that use phantom power.The familiar XLR, used extensively in professional audio. It offers a sturdy connection, and because it locks in place it is difficult to disconnect accidentally, which makes it perfect for connecting devices that use phantom power.

Most textbook explanations claim that a balanced interface requires two symmetrical signals of opposite polarity, passed over a pair of screened, twisted wires. These two signals are received by a differential input device (transformer or electronic input) which subtracts one signal from the other to extract the wanted audio, while also cancelling any unwanted interference picked up on the lines in the process.

There is an element of truth in there somewhere, but it is not the whole story and it is rather misleading.
In fact, the operation of the balanced line relates directly to the Wheatsone Bridge... I expect you wished you had paid more attention in school physics lessons now! The critical aspect of any balanced interface is that its design forces any interfering signals to have as near as possible the same amplitude on both wires. That is the essential aspect that allows the differential input to cancel out the interference — and that condition is ensured by arranging for each signal wire to have an equal impedance to ground.

So it doesn't actually matter to the operation of the balanced interface whether there is a pair of symmetrical and inverted signals or not. The point is that given equal impedances to ground from each signal wire, any interference will be presented to the differential receiver as a 'common mode' signal, which means equal in polarity and amplitude on the two signal wires. The differential receiver (whether a passive transformer or an active electronic input) inherently rejects common-mode signals — removing the interference — while passing any differential signal, which should be the wanted audio.
A TRS B-gauge connector. Though the diameter is the same as instrument jacks it has a different tip, so don't mix and match!A TRS B-gauge connector. Though the diameter is the same as instrument jacks it has a different tip, so don't mix and match!

In the case of a fully symmetrical signal, where both wires carry identical but inverted versions of the same thing, the output will obviously be twice the amplitude of either individual signal — and this is the traditional way of working a balanced interface. However, the system still works perfectly well if the signal is passed only via one of the two wires, and this is a very common modern arrangement employed in budget mixers and other audio equipment.

This kind of configuration is often referred to as an 'impedance-balanced' or 'ground compensated output' and, typically, the hot side of the balanced interface is driven to carry the required output signal, while the cold side is connected to ground at the source end via a resistor which is carefully chosen to make sure that both wires have the same impedance to ground. (A true 'ground-compensated' output also monitors the voltage on the cold side and feeds that into the driving amp to cancel out any ground-loop noise.) I know many people look at impedance-balanced and ground-compensated outputs with scepticism, and don't believe they are a true balanced output at all — but they really are perfectly functional balanced interfaces, with exactly the same interference-rejecting properties as the more familiar, symmetrically driven system.

The only practical difference is in terms of the signal amplitude on the wires. The hot side of an impedance-balanced or ground-compensated output typically carries a signal which is 6dB larger than the level carried on each wire of a fully symmetrical balanced output. If you think about it, it has to be that way to ensure compatible output levels from a differential line receiver. The fact that only one line is driven in a ground-compensated output is also advantageous in situations where the output is likely to be connected to an unbalanced input, since the signal level remains as expected. If you connect a symmetrical output to an unbalanced input you normally lose the cold side contribution, and the signal level ends up 6dB lower than it should be!

Patchbay Wiring

Once the heart of any studio, patchbays, or jackfields, aren't as popular or necessary in computer-based facilities. However, where you need to be able to patch analogue outboard equipment into different signal paths, patchbays remain the most cost-effective solution.
A common wiring convention for patchbays, with inputs arranged beneath the corresponding outputs.A common wiring convention for patchbays, with inputs arranged beneath the corresponding outputs.Inserting the plug into the top socket of a half-normalled patchbay allows the signal to be heard without breaking the direct signal path, which can only be broken by inserting a jack into the lower socket.Inserting the plug into the top socket of a half-normalled patchbay allows the signal to be heard without breaking the direct signal path, which can only be broken by inserting a jack into the lower socket.Stereo half-normalling allows mono signals to be patched into stereo inputs.Stereo half-normalling allows mono signals to be patched into stereo inputs.In a single-normalled patchbay, plugging a jack into either socket breaks the signal path. In double-normalled patchbays, the signal path is not broken unless both sockets are filled.In a single-normalled patchbay, plugging a jack into either socket breaks the signal path. In double-normalled patchbays, the signal path is not broken unless both sockets are filled.Most commercial home studio patch-panels are designed for A-gauge plugs (see main text) and have some form of user-adjustable 'normalling', either via switches or by removing and re-orientating PCB cards holding the socketry. The normalling in professional B-gauge and Bantam jackfields is configured by wiring the appropriate terminals and switch contacts as necessary on each socket — not a job for a soldering novice!

Configurations vary, but generally patchbays are configured with two rows of jack sockets per rack strip. The upper row (as seen from the front) is normally used for outputs, while the lower row is used for inputs. For example, the upper row might carry the outputs from a multi-channel recorder, while the lower row provides the channel inputs to a mixing console.
To make a connection, you simply insert a suitable patch cord to link the appropriate output to the required input destination — giving each plug a couple of twists as you insert them to make sure the contacts are wiped clean.

In most cases, there will be a standard default set of connections, such as the recorder outputs to the desk channel inputs. It is obviously tedious to have to plug patch cords for this kind of thing, so it is common practice to 'normal' the connections instead. This means arranging for the output signal on the top row of jacks to be wired inside the patchbay to a set of break contacts on the corresponding input sockets directly below.

In this way, with no patch cords plugged in, the recorder outputs are passed automatically to the corresponding desk channel inputs. This is called a 'half-normalling' arrangement. If you need a second output from a recorder channel, you can simply plug a patch cord into the appropriate output socket on the patchbay without breaking the signal to the desk channel.

However, if you need to route something else into a particular desk channel, instead of the normalled recorder output, you can simply plug a patch cord into the required channel input socket. Inserting a plug forces open the break contacts, interrupting the signal flow from the normalled output jack above it. For this reason the input socket is sometimes called a 'break jack.' The diagram shows the wiring arrangement, and exactly the same configuration is used for A- and B-gauge and Bantam jackfields.

There are a couple of alternative forms of normalling, although these are less commonly used and rarely available on commercial patchbay panels. Single Normalling is where the breaking contacts of both the output and input socket are wired to each other, so that inserting a single plug into either the output or input socket breaks the normalled link. Double Normalling is effectively doubled-up half normalling, and to break the normalled signal path it is necessary to insert plugs into both the output and input sockets. I can't think of an installation in which I've used single normalling, but double normalling is often used for main outputs and other mission-critical connections. The advantage over half or single normalling is that there are two sets of break contacts wired in parallel to pass the normalled signal, which improves reliability significantly.

Obviously, you can't see the normalling arrangement behind the sockets of a patchbay, so to indicate which sockets are normalled it is usual to colour code the labels above each socket. The BBC convention is to show half normalling with red above the input (break) sockets, while double normalling is shown with yellow above both output and input socket — but this is only one convention and different manufacturers and organisations use alternative colour schemes.

When designing a new patchbay layout, it is important to plan the location of each group of signals carefully and to make sure the normalling is logical. The classic mistake is to have the output of an effects unit normalled to its own input, which isn't a good idea! Usual practice for a small installation is to have the signal flow zig-zag across the patchbay from left to right: source outputs to channel inputs, group outputs to recorder inputs, recorder outputs to monitor inputs, desk outputs to master recorder inputs, speaker outputs to amplifier inputs... and so on.

Connectors

There are five commonly used general-purpose analogue audio connectors: the XLR, the B-gauge jack plug, the Bantam/TT plug, the A-gauge jack plug, and the RCA phono plug.
The three-pin XLR is probably the connector most strongly associated with professional audio equipment, used for both mic and line-level interfaces. The name actually refers to the original Cannon X-series connector, which was then upgraded with a Latch and a Rubber insulator — hence the XLR. Although now made by many alternative suppliers, the XLR-format connector has been the standard for most microphones since the late 1970s, and the AES organisation eventually managed to standardise the wiring convention worldwide in 1992, in a document called AES14.
The Bantam TT jack is a miniature version of the B-gauge jack. Though they are more delicate, and a little more fiddly to solder, you can fit more connections in a 19-inch patchbay, which has made them common for applications such as built-in patchbays for large consoles.The Bantam TT jack is a miniature version of the B-gauge jack. Though they are more delicate, and a little more fiddly to solder, you can fit more connections in a 19-inch patchbay, which has made them common for applications such as built-in patchbays for large consoles.

This standard demands that the screen connection is carried on pin 1, the hot side of the balanced interface on pin 2, and the cold side on pin 3 (often remembered by the mnemonic Xcreen, Live, Return). Outputs are always presented on male connectors (with pins) and inputs are always on female connectors (with sockets) — easily remembered, as the pins point in the direction of the signal flow. In general, if you see an XLR being used to carry a line-level signal, it is reasonable to assume the nominal level will be +4dBu.

Balanced, stereo microphone and line-level signals can also be conveyed using a five-pin version of the XLR. Again, the screen is on pin 1, with the left channel on 2/3, and the right channel on 4/5. The even pin-numbers carry the hot side of the balanced interface and the odd numbers carry the cold side. The same connector convention is used, with female for inputs and male for outputs.

XLRs are very rugged connectors, with self-cleaning and gas-tight contacts in the best designs, and provide excellent screening properties. However, they are quite bulky and impractical where many connectors have to be squeezed into a small space. The original solution to this problem was the Post Office 316 plug, known generically as the B-gauge plug. Developed for manual telephone switchboards, this brass quarter-inch plug has three contacts arranged as a rounded tip, a ring and a longer sleeve. The tip normally carries the hot side of a balanced signal, the ring carries the cold side, and the sleeve carries the screen.
Quarter-inch, A-gauge jacks generally come in two flavours: TS (Tip, Sleeve) and TRS (Tip, Ring, Sleeve). You can identify these easily, as there are two plastic bands on the tip of the latter and only one on the former.Quarter-inch, A-gauge jacks generally come in two flavours: TS (Tip, Sleeve) and TRS (Tip, Ring, Sleeve). You can identify these easily, as there are two plastic bands on the tip of the latter and only one on the former.

Most professional patchbays (jackfields) use B-gauge sockets, typically squeezing 20-26 sockets into each 19-inch frame row. The use of B-gauge plugs and sockets is generally associated with +4dBu line levels, although some installations do route microphone level signals via jack plugs. However, there are numerous caveats to this practice, not the least of which is that phantom power gets shorted out when inserting or removing a patch cord, and that can damage both mic and console, so I wouldn't recommend patching mics this way — an XLR patchbay is a much safer solution!

For even greater socket density — such as might be required on built-in console patchbays — there is a miniature version of the PO316, called the Bantam or TT (Tiny Telephone) jack plug. Essentially, the Bantam is a scaled-down version of the PO316 with a 4.4mm diameter instead of the latter's 6.35mm diameter. That allows 48 sockets per row, doubling the socket density compared with a standard PO316 patchbay. Largely because of the smaller size, Bantam patchbays don't tend to be quite as reliable as the full size B-gauge version, but in most cases this shortcoming is outweighed by the space-saving convenience. Bantam plugs are usually employed for line levels at +4dBu.
Phono connectors are used extensively in domestic hi-fi applications, but generally are only found in professional studios for digital S/PDIF connections.Phono connectors are used extensively in domestic hi-fi applications, but generally are only found in professional studios for digital S/PDIF connections.In the semi-professional world, the A-gauge jack plug is often used in an equivalent way to the B-gauge for balanced circuits. Again, the TRS (tip, ring, sleeve) version carries hot, cold and screen connections, respectively. When used for balanced line-level signals, A-gauge plugs and sockets could work at either +4dBu or -10dBV, depending on the intended market for the equipment. Some equipment even allows the level to be switched to suit the application.

A-gauge jacks are also used for unbalanced circuits, but in a simpler TS (tip-sleeve) version, typically used for guitars and keyboards. The TRS version is also used for unbalanced stereo connections, most commonly on headphones. In this case, the wiring convention is left channel to tip, right channel to ring, and the commoned returns to the sleeve — and that applies to the miniature 3.5mm TRS plugs on stereo earpieces and headphones too.

Although the A-gauge and B-gauge plugs have the same overall shaft diameter, they have completely different tip shapes and the dimensions and positions of the insulators between conductors is also different, which makes them completely incompatible with each other. The most significant difference is that the B-gauge has a small, rounded tip, while the A-gauge has a larger, diamond-shaped tip. If a B-gauge plug is inserted in an A-gauge socket, it is unlikely to make reliable contact because of the smaller tip, whereas if an A-gauge plug is inserted into a B-gauge socket its oversized tip is likely to strain and bend the contacts, such that they then become unreliable with proper B-gauge plugs. So the golden rule is to never insert the incorrect type of plug for the socket: a handy rule of thumb for life, I find!
As well as the basic connectors described in this article, a number of others are sometimes used in commercial studios. The D-sub snake pictured here, for example, is often used to provide an eight-channel breakout cable from rack equipment. Others, such as EDAC, are often used in fixed installations where cables are not likely to need plugging and unplugging on a regular basis.As well as the basic connectors described in this article, a number of others are sometimes used in commercial studios. The D-sub snake pictured here, for example, is often used to provide an eight-channel breakout cable from rack equipment. Others, such as EDAC, are often used in fixed installations where cables are not likely to need plugging and unplugging on a regular basis.

The last common audio interface I'm going to mention here is the RCA phono plug and socket. This is the standard unbalanced connector found on most hi-fi systems, and was very commonly used in semi-pro equipment, such as analogue tape recorders and mixers. This connector does make an appearance on professional equipment, but normally only in its digital guise, for conveying S/PDIF signals.

The outer ring of the RCA phono plug and socket are used for the screen connection, while the central pin/socket is used for the hot signal connection. The spring action of the outer contact is all that holds the RCA connector in place, and cheap phono plugs have a nasty habit of losing their grip over time. This makes the screen connection unreliable, and sometimes the plugs can even fall out of the socket!

I've only described the five most common connectors here, but there are many other types used for specific applications. For example, vintage mics and some older audio equipment use variations on the multi-pin DIN connector (Tuchel, Preh and Binder), while loudspeakers are usually connected via 4mm banana plugs, or Speakon connectors. Multicore cables are typically equipped with Plusconnection, Contact or MIL connectors, and multi-channel interfaces to tape recorders and DAWs are often via D-sub (see photo above) or EDAC connectors. There are also numerous variations on wiring standards and pin-outs, just to add to the confusion. Hopefully, though, this article will have clarified the most important aspects of analogue interconnection. Happy plugging! 



Published May 2007

No comments:

Post a Comment