Every console has linear faders of some kind and these are particularly vulnerable to dirt, dust and general abuse. There are three basic types: carbon track, conductive plastic track and voltage controlled. The latter type is used extensively in so-called computerised mixing consoles and has built-in electronics (VCA, Sample & Hold, etc). As might be surmised from the name, the resistive element feeds a DC voltage to control the amount of attenuation required. A disadvantage of many of these faders is that when control is handed over to the computer the fader knobs fail to move, so one has no clear idea of where the electrical setting is. VCA fader manufacturers are looking at ways of overcoming this with various types of displays.

Another type of automated fader uses a small servo motor mounted in the fader so that the knob actually moves in the automatic mode. Apart from the obvious visual advantage, the system also offers the facility of being able to grab hold of the knob, which disconnects the servo drive, and set the fader to any desired point. This is very useful in mix-down when quite often you spontaneously decide to adjust a setting.

The chances are that your console will have standard faders with either carbon or plastic resistance elements. Generally speaking, the longer the faders, the more possible it is to exercise control over the signal. Faders which can be taken apart for servicing are a big bonus. Since they are basically mechanical devices, naturally parts wear out. It is, therefore, cheaper in the long run, if you intend keeping a console for any length of time, to have removable, easy to disassemble, faders. Carbon track faders can be cleaned as outlined in part 3 with contact cleaner. Conductive plastic faders are better washed in slightly soapy water, rinsed and left to dry. Any residue can be wiped off with a dry lint free cloth.

Level Matching

Most consoles have inserts or points in the signal line (see Figure 3, part 2) where auxiliary equipment may be used (eg. parametric equalisers, compressors, digital delay line, etc). Since these have to match to a wide range of units, a few parameters are important. In general, one feeds from a lower impedance to a higher impedance (15 to 20 times), although maximum efficiency is obtained when the impedances are equal. This latter point is only of real consequence in AC power distribution systems so as such it has little bearing on our subject matter. However, a quick look at the ideas involved should enable you to resolve any doubts you may have about compatibility in your system.

The circuit in Figure 6a shows a theoretical output ie. an AC voltage source in series with a fixed impedance Z₁. It follows that if we were to terminate it with an impedance equal to Z₁ (Figure 6b) we would make a potential divider where half the original EMF (6dB) is dropped across Z₁ and the other half is dropped across Z₂. So if the EMF across points A, B was 10V AC without Z₂ (unloaded) it would drop 5V AC when we connected Z₂. If we added another load, ie. Z₃, the EMF would drop still further. This is obviously not the best situation.

In most sound systems the signal is considered to be an alternating EMF. It is for this reason that audio console manufacturers and others mutually agreed to introduce the dBV (decibel Volt). Before that the standard was the dBm (decibel meter) which is referenced to 600 ohm. Since different manufacturers provided different input and output impedances, confusion on specifications abounded. The main requirement though is that several inputs can be connected to a single output without a significant drop in signal level. This is why input impedances are generally between 10k ohm to 100k ohm and output impedances lower than 50 ohm.

Coupling with Transformers

There is still an abundance of secondhand equipment available at reasonable cost, so it is worth briefly looking into the history of the 600 ohm termination. This is still used by the BBC and others as far as I know. The need for a 600 ohm balanced termination arose when audio signals had to travel down long lengths of cable without degraded frequency response or unwanted electro-magnetic interference. Figure 7 shows the basic scheme, which you would do well to make a note of for those odd occasions when this is necessary.

The advantage of this system is that since the impedance is low on the long section, the signal is not unduly affected by the shunting effect of the cable capacitance at high frequencies. Also any stray alternating electromagnetic fields will impinge equally on both conductors, inducing the same EMF in both conductors. This will not form a potential difference across the windings of the transformer, so it will not be coupled to the secondary of transformer. Hence this balanced, floating configuration is largely immune to interference. Some equipment will have transformer coupled outputs which have to be correctly terminated, usually with a 600 ohm load, to prevent the transformer from ringing (ie. making unwanted harmonics, see Figure 8b).

Correct termination is mandatory with transformers. The best way to check this is by running a 1kHz square wave through the equipment and observe the output on an oscilloscope. Any deviation from the input waveform should be thoroughly checked out. Don't expect too much from your tape machine in this respect. Recent advances in technology have enabled manufacturers to provide machines with virtually ideal transient response. However, most ordinary machines are quite poor (we will be looking at tape machines in greater detail later on). All other ancillary equipment and the mixing console should be able to handle a square wave without any side effects.

An output obtained similar to Figures 8c, d, e or f would show up as poor frequency response on a sine wave test. However, the chances are that 8b would go unnoticed because of its dynamic nature. Many synthesiser sounds have a high transient content so the ability to handle these is very important. Figure 7b shows how the situation arises in the first place and how it can be resolved. An unterminated 600 ohm transformer feeds a 10k ohm fader. Since the transformer should see a lower impedance ie. 600 ohm it produces a waveform similar to Figure 8b. If the unit is provided with a switch to drop a 600 ohm resistor across the output all well and good. Otherwise a resistor and capacitor in series across the output (0.1uF and 470 ohm) should do the trick. The precise combination of resistor and capacitor will depend on the characteristics of the transformer, which you probably won't have, so some experimentation may be required. Once you have obtained a good square wave (Figure 8a) check for flat frequency response using sine waves, from 30Hz to 15kHz, as outlined in the next section.

Checking Frequency Response

Flat frequency response, or the ability for your system to handle all audible frequencies equally, is vitally important for correct reproduction. To check this you will need: a sine wave generator, an AC millivoltmeter, and if possible, an oscilloscope. You may be able to borrow or hire some or all of these. If you are in any doubt about using them it is best to seek advice from someone who has done it before.

There are several waveform generator chips available for less than £4 so a complete oscillator with buffered output could be built for about £10. Similarly it is possible to use a VU meter in place of a millivoltmeter provided you know the calibration is good, but beware of multimeters with decibel scales. Most of these are only accurate up to about 200Hz so you will certainly get misleading results if you try to measure 1 Volt rms or so at 15kHz. On the other hand you may have access to a high quality DVM. Providing the instrument has adequate frequency response (check the specification) you should get accurate results. The difference between two voltages can be expressed in decibels by:

Diff. in dB = 20 log (Measured Voltage) / (Reference Voltage)

So if for instance we apply 1.5V RMS at 1kHz to the line input of our mixing console or auxiliary unit, and set the output for any convenient reference point eg. 0.775V RMS or 0dBV or 1.228V RMS for 0VU (ie. +4dBV), then by adjusting the input frequency over the desired bandwidth ie. 30Hz to 15kHz in octave steps (2kHz, 4kHz etc, 500Hz, 250Hz etc) we can ensure the output stays within certain limits. Under normal circumstances it is only the extremes that droop, hence the common way of specifying equipment is "plus 0dB, -1dB: 20Hz to 20kHz".

In our example let us assume we've set the output for 1.228V RMS at 1kHz (0VU) and at 15kHz we measure 1.094 V RMS, then by our equation:

The negative sign indicates that the measured voltage is less than the reference voltage.

Similarly at 30Hz we measure 0.9209 V RMS then:

You can see now why meters in audio are scaled in decibels rather than volts! The important thing to remember is that a rise or drop of 'xdB's' is the same relative amount (ratio) irrespective of the actual level. Twice or half as loud is equivalent to saying +6dB or -6dB (remember part 2?). Hence it is a relative rather than an absolute description. Later on we'll be talking about adding or subtracting 'xdBs' at 'xkHz' or, losing or gaining 'xdBs' overall etc, so make sure you've got this relative aspect clear in your own mind.

The DI box

Many electric instruments have high output impedances and relatively low level. This is where the Direct Inject box (DI box) comes in. Most DI boxes are, in fact, only a transformer with a 50k ohm input impedance and a 600 ohm output. By connecting the 50k input to the guitar and the 600 ohm to the mic input, the problem of interfacing is solved. The transformer turns ratio will reduce the EMF from the guitar though this is not normally a problem because of the high gain available in the mic amp.