PA Systems (Part 1)
An explanation of PA complexities.
Looking back at band gear over the last ten or fifteen years, it is difficult to ignore the feverish activity that went into the early development of instrument amplification as compared to the very slow rate of progress in band PA's. By the time a four channel 100 watt amp and a pair of the proverbial 4x12 columns was the standard issue in PA's, the instrument amp had all but reached its current state-of-the-art. That was back in the early sixties!
As well as taking its name from conventional Public Address trade, it took the basic form of equipment that was in use by the trade and covered it in black leather cloth. The poor old 4x8 and 4x10 column never knew what hit it, as it distorted and groaned and blew out cone after cone! On almost every concert hall platform were rows of immaculate, black Vox AC30's, proudly sporting their brown, diamond-tracked, grill cloth, or the distinctive Selmer Thunderbird 50's clad in crocodile skin with their green eyes winking mysteriously in the dark. The sounds they produced were as impressive as their appearance. But, alas, the two PA columns, hardly noticeable at the sides of the stage, and the poor little 50 watt amp driving them, were almost driven to destruction and the sound was atrocious.
Did no one understand PA in those days, or was it quite simply that Public Address equipment already existed for other purposes? It was so much more simple to apply existing design practice than to start afresh.
It is amusing to record that the first steps in the development of improved PA loudspeakers were prompted by the unlikely office of Her Majesty's Customs and Excise, in their great wisdom, they declared that any loudspeaker system that employed units of under 12in. diameter would be liable to purchase tax, but if units of 12in. or over were used, no purchase tax need be paid. For obvious reasons manufacturers hastily re-designed their products and the 4x12 PA column had arrived. Almost without exception, those early 4x12's used the tried and proven Rola G12 unit, similar to the standard speaker fitted to the Vox AC30 Combo. Judging from the vast numbers of PA and instrument cabs that still use a version of the G12, it must be the best selling loudspeaker unit of all time!
There can be little doubt that the use of the 12in. loudspeaker was a vast improvement on both the quality and power handling capability of PA columns. For many years after, the 4x12 column was the PA cabinet, if more power was wanted, all that was necessary was to add more 4x12 columns and amps to drive them. The absurd stage was reached where it was not unusual to find bands carrying up to fifty 4x12 columns into a gig!
At this point, with WEM as the big name in a booming market, bands began to discover that there were many new problems associated with a multi-speaker, multiamp rig and began to wonder if this really was the answer to the PA problem, it seemed that no matter how many slaves or columns might be added, the sound that came out was always harsh and distorted when the gain was turned up. It seemed impossible to achieve that really full-bodied, clean power that slapped you in the face whenever the guy in the mix-down studio replayed your tape through his monitors. That was the sound that was wanted, but the PA just could not be coaxed into giving it. A few horns began to appear perched on the top of a band of columns. This certainly improved the crispness at the treble end, but it did not solve the main problems. What then was the answer?
The answer was found by reverting back to the type of loudspeaker systems and equipment configuration that was used in cinemas back in the 1930's. A glance behind the screen of almost any cinema will reveal an array of bass bins and horns that very closely resemble those in use as band PA's today.
There were many, very good reasons why the multiple 4x12 column and stack of 100 watt slaves did not give the power or quality that was expected. Understanding these reasons will form the main part of this series of articles over the next few months.
The workings of low impedance, balanced line mics. and condenser mics. will be covered; the problems associated with the distribution of loudspeaker feeds; the effects of stacking multiple speaker units together; the mysteries of crossovers — passive and electronic — leading into biamplification and the advantages to be gained from this approach; the workings of mixing desks, horn loaded bass bins and treble horns; the use of graphic equalisers; the intelligent use of foldback; understanding in real terms the meaning of amplifier power output figures, will be among the PA topics discussed. At least one article will be devoted entirely to the understanding of room acoustics. The aim of the series is to give a basic working knowledge of modern PA's and, hopefully, to dispel some existing misconceptions that seem to surround this subject.
In the introduction, the development of PA's up to the advent of the multiple 4x12 column rig was very briefly and sometimes, nostalgically discussed. It was realised that this approach was not likely to provide either the quality of power that bands were seeking in a PA system. The probable answer was to be found behind cinema screens, in the form of horn loaded bass "bins", as they are termed today, and arrays of treble horns. As the series progresses, the shortcomings of the old approaches will be identified and solutions described until the basic system evolves into a modern PA rig.
One of the earliest problems that confronted bands as power levels began to increase, was the need to control the PA from a vantage point out in the auditorium, so that the guy at the controls could hear what the audience was hearing. In the first instance, the old PA amp was moved out front in its entirety, and the microphone and loudspeaker leads lengthened to reach.
Whilst this arrangement certainly served its purpose in getting the control point out front, it suffered many more serious disadvantages. The sound quality became muddy and generally lost its clarity. This was due to the fact that almost all straight PA amps have high impedance inputs, and the maximum length of lead that should be used with a high impedance mic. is about 5 meters (approx. 20 feet). As this length is exceded, the treble response of the microphone is progressively reduced due to the effects of the capacitance of the cable itself. The example given of 5 meters is for a good cable, and will be considerably less for an inferior cable.
An additional problem was the increased tendency of the PA to pick up radio transmissions from passing police cars or taxis. The long capacitive leads made excellent aerials and turned some of the less elaborate amplifier input stages into simple radio receivers.
The solution to this was to change over to low impedance microphones, preferably of the balanced line type that had been in use by the general PA trade for many years. The amazing thing here is that a high impedance microphone consists of a low impedance diaphragm assembly, connected to a transformer in the microphone case, to convert it to high impedance! All that should be necessary is to move the transformer from the microphone case and wire it into the amplifier input jacks instead to produce a low impedance system. In practice it is not that simple due to the rather complex mechanical construction of most microphones.
It will generally be necessary to purchase a suitable transformer for wiring into the amplifier, and to use the low impedance connections provided on most good quality high impedance microphones, as an alternative. Fig. 1 shows some typical wiring arrangements for microphone circuits. 1(a) shows the normal circuit for a high impedance mic. connected directly to the input stage of an amplifier. This arrangement will be typical of most straight PA amps. The fact that a valve may be used instead of the transistor shown will only effect the ratio of the transformer needed to obtain a good 'match' between the mic. and the amplifier. 1(b) shows the situation described earlier, with the transformer moved to the amplifier input, so that the signal can travel to the amplifier at low impedance. Circuit (c) shows a low impedance balanced line system as used on the more sophisticated PA mixers and in studio systems.
The correct matching between microphones and the input stage of the amplifier or mixer is a matter often overlooked by entertainers, and one which can make a significant difference in system performance. Most microphones like to 'see' an amplifier input impedance higher than the impedance of the microphone itself and will usually perform better under these conditions. For example, the actual impedance of most high impedance dynamic mics lies around the 20K ohm to 40K ohm mark, while the input impedance of most straight PA amps is about 50K ohm to 100K ohm. There are several impedances in common use under the low impedance banner, ranging from 25 ohms up to 600 ohms. Any of these will work into an input designed for any of the others, but there will be a marked improvement when a microphone is fed into a properly matching input stage. The impedances in general use seem to be 50 ohms, and these are best fed into an input with an actual impedance of between 300 and 500 ohms. Particular makes and models of mics. behave differently under varying loads, so the information given here should be taken as a general guide only.
It will be seen from Fig. 1(a) & 1(b) that the only difference between a high impedance and low impedance un-balanced system lies in the impedance that is presented to the screened cable. It can be shown by formulae and such, that the rate of attenuation (or reduction) of treble frequencies due to the effects of capacitance referred to earlier, will be much on a high impedance circuit than on a low impedance circuit. So one advantage of using low impedance is that the quality of the microphone response is preserved when fed down long mic. lines.
A study of 1(a) will show that with the high impedance system, the screened cable is in effect carrying the input circuit of the amplifier right up to the microphone as well as performing the more obvious task of carrying the microphone signal to the amplifier. Therefore, any electrical characteristics of the cable, such as capacitance and inductance are added to the input circuit and can easily modify the amplifiers performance. In extreme cases of very long lengths of poor quality cable, the amplifier will become unstable and oscillate of its own accord or become tuned and pick up radio signals. The inclusion of a transformer in the input stage as shown in 1(b) has the effect of isolating the amplifier input and therefore reducing the modifying influence of external circuitry on the amplifier. The balanced line circuit shown in fig. 1(c) has several further advantages over the unbalanced line, the main ones being improved amplifier stability and a cancelling out of extraneous clicks and other noises that are often picked up by the cable itself. With a low impedance balanced line system, there should be virtually no limit to the cable lengths that can be used on microphone circuits, provided that the amplifier or mixer has been properly designed for balanced line working.
Another problem that became apparent as the PA amp. was moved out front, was a noticeable drop in power output, and this was especially so with the new solid-state amplifiers. The problem here was again cables, but this time it was due to a general overlooking of the fact that loudspeaker cables have their own impedance. This is small enough to be ignored when cable lengths are short, say up to about 10 meters, but it can have a considerable effect on output power when long cables are used.
Consider as an example a 100 watt amplifier connected to two 4x12 8 ohm cabinets by a 50 meter lead. It would not be unusual for such a lead to have a self-impedance of 4 ohms. It can immediately be seen that we have a situation where a speaker load of 4 ohms is connected to an amplifier by a lead of a further 4 ohms, so that only half of the amplifier power ever reaches the cabinet, the rest being dissipated as heat in the lead. Now if the amplifier happened to be of the solid state type whose output power decreases in direct proportion to the increase in load impedance, the situation is worsened still further. If we suppose that our solid-state amp. was designed to give 100 watts into a 4 ohm load, then it follows that with a load of cabinets and cable totalling 8 ohms connected to it, it will at best only give 50 watts into the total load, and as half of that will be lost in the 50 meter lead, we will have only 25 watts actually reaching the two cabinets! An absurd situation you may think, but one that is encountered only too often in practice. There are two answers here, and either will provide a sensible solution. The more obvious is to use a thicker speaker lead so that the impedance of the cable represents a smaller proportion of the load impedance that is presented to the amplifier. It must be realised that some loss will be inevitable, and if this can be reduced to below 20%, a sensible compromise will have been reached. A better solution is to reduce the length of the speaker leads until they are so short that the cable impedance is negligible anyway, and this means lead lengths of only 1 or 2 meters.
This, however, is completely at variance with our purposes as such short leads will not enable the amp to be located out front. We arrive at the point when it is necessary to split the amplifier into two parts. The input stages are separated into a separate mixer that can remain out front connected to the stage by low impedance, balanced line inputs and outputs, and the output stages separated into a series of slave amps that can now be placed right behind the speaker cabs and connected directly to them with very short leads.
By considering a few of the problems and discussing some solutions, we have now evolved the basic system into its current form employing mixers and slave amps to provide the flexibility required for proper control of the PA. Next month's article will deal similarly with the development of loudspeakers systems.
Feature by Ken Dibble
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!