Chapter 7

T9 PHASING AND MATCHING PROBLEMS IN SOUND REINFORCEMENT SYSTEMS

In certain situations, phasing of loudspeakers is unimportant, but most often it is at least noticeable in its effect and sometimes it assumes serious importance. In any system, where the acoustic waves generated by individual loudspeakers should be "parallel" (loudspeakers in same plane) their phasing is extremely important. Where the sound comes from overhead or in front of the audience, reversal of phase of any one loudspeaker will result in a serious "hole" in coverage. Where the intensity from the reversal loudspeaker is equal to that from others in the system, the combined effect is a transverse wave movement that create the impression that the local loudspeakers are not working at all, and actually what the listener " thinks" that he hears, is confused sound from more distant loudspeakers. This will be the complaint usually received under these circumstances and it should be checked (a) by verifying that the speakers in question are connected and working, and (b) that their phasing is correct. But in many places, such as long halls, successive rather than parallel reinforcement is needed, even if delay in system. In such a system it is difficult to avoid atleast small "dead spots" where hearing is difficult, because sound from two speakers at different distances, almost inevitably causes some confusion of sound. Often reversal of phase of one of such a pair, compared to the other, will make the "transition zone" smoother. The required phase may be "in" or "out" in the electrical sense. The only way to find which is best, is to try each of them and explore the transition zone quite thoroughly. Changing phase will almost invariably move the transmission zone at which most difficulty in hearing intelligibly is experienced, so do not just check at the listening position while phase is reversed. Choose the connection that gives best overall listening throughout the whole transition area from where the sound appears to come exclusively from one unit, to where it appears to come exclusively from the other. One connection may shift the "difficult" spot into a less important area, or somewhere else not occupies by audience for one reason or another. In such cases, this is the obvious choice of phasing. In making checks, use voice intelligibility tests, because these are the most critical. Also use different voices (either live or recorded) because the frequency range in which important intelligibility components lie can affect the apparent result. When setting up an installation, such as a large hall, where both parallel and successive waves occur one must work successively. First connect only the first group and check for correctness of phasing. Then connect only the second group and so on, checking each group for phasing within the group, separately from other groups. Next check for the best phasing between successive groups, first and second, then with both these connected, find the best phase for the third group, and so on, until the entire system is connected. Some loudspeaker manufacturers mark on winding post of the speaker for phasing purposes, so that when a positive voltage is connected to the marked post (usually a red dot), and a negative voltage to the other one the diaphragm moves forward. Where wiring is identified in polarity, by

black/red red/white or black/white insulation colouring, it makes the ensuring of correct electrical phasing simple. If all speakers are connected in parallel to the amplifier, it is a simple matter of connecting all marked posts to one colour of wire and all unmarked posts to the other colour. However, maintaining correct colour continuity should be maintained throughout the system. The connections to the voice coil whether in series or in parallel, must be made in such a manner that in any one instant all diaphrams must be moving outward or inward in unison. When the voice coil terminals are not marked, the simplest way of determining the correct phasing is by doing as follows. Take1.5V or 3V batteries (dry cell) with the polarity marking of each observed, connect them momentarily to the voice coil of the speaker. The cones or diaphrams should move in the same direction for correct phasing. The voice coil connections of those level speaker cones that moves in opposite directions to the majority should be changed. Phasing is of least importance where two loud speakers are at a good distance apart (as a thumb rule if the distance is more than 5 meters) or pointing in opposite directions. In system where the entire acoustic field is a parallel wave, it is safe and quick to connect up on this basis. If the speaker should happen to be wrongly phased (which is quite unlikely, but nevertheless possible), it will be easy to spot, because it will produce "dead spots" surrounding itself on each side, where it meets the sound field from its neighbours, in some instances the speaker itself will seem dead too. Between any pair of speakers intended to produce parallel wave development, phasing can be checked by listening critically, at a point equidistant from both. Moving a little to the left should make the sound appear to come from the left speaker, while moving to the right should make it appear to come from the right speaker. In between, there should be a point where the sound seems to shift from one to the other, remaining quite definitely in front of the listener. Wrong phasing destroys this continuity. Where successive wave development is used, and phasing proves necessary, connect the system successively and check at each point, as described earlier, before proceeding to the next. This care is time well spent, because you can waste a lot of time trying all the permutations and combinations of phase changes, with all speakers connected, and still finish up missing the correct one. Another factor they control effective coverage is the relative power fed to different units in a system. In a parallel wave system, if each unit feeds a virtually identical audience area, each should receive the same audio power, as well as being connected in the same electrical phase. But if areas fed are unequal in such a system, or the building being served locks symmetry in one way or another, different power distribution may be needed. The system still uses parallel wave distributions on phasing is maintained electrically correct, but the power to the units serving the larger areas is stepped up relative to those serving smaller areas. Where successive wave distribution is used, the relative power fed to successive units also affects the phasing which will seem best, because it moves the zone, where confusion is likely to occur. In such cases, the relative power can probably be decided by simple arithmetical comparison of areas served twice the area, twice the power. This will simplify the phasing procedure, because only one "variable", the phasing has to be checked experimentally. But sometimes the calculated relative power will be found incorrect when the job is complete, because some factor was either not considered, or was incorrectly assessed. Seldom is out of

phase operation employed but where reverberation problems exist, many difficult cases have been over come by judicious use of a combination of connections. For example, it may be that, as well as having to service a larger area, one speaker also "looses" sound because its area is not enclosed in the same way as adjoining areas are. This will call for a greater proportion of power than mere increase in service area will. After stepping up in power to correct a weak spot, you will need to recheck phasing, if successive wave distribution is involved. The above deals mainly with phasing and variation of electrical power purely from the view point of the acoustic effect i.e., the overall result on the effective coverage of the system. But the basic problem is how to get the right amount of power in the right place. The normal choices are to use series or parallel feed, with mixtures of series/parallel sometimes serving a useful purpose. Where a number of speakers are used, the disadvantage of parallel feed is that impedance gets very low, and consequent line losses become a considerable portion of the available output. The system gets quite inefficient. The argument against series feed is that speakers "work best" with parallel feed. The usual reason given is that they get better amplifier damping. A group of speakers all in parallel results in each speaker being directly connected to the low amplifier output impedance as damping. The same group all in series results in each speaker having all the others connected as series damping, so the amplifier cannot damp each individual speaker unit at all. There is a better alternative available, that avoids both problems by using constant voltage lines. In high fidility and similar audio applications, the reference is to impedance matching. If multiple units are used, the connection must be that the total load is matched to the correct impedance of the amplifier. Constant voltage lines involved a change in concept. The system is regarded as analogous to a power distribution system, where the lines carry constant voltage, and the power taken by any appliance connected is controlled by the impedance of the appliance. So, in speaker systems, the relative power controlled by the impedance presented by the speaker, through its constant voltage line matching transformer. Any system of connecting a number of speakers to an amplifier has to serve two purposes:a) deliver the full available output of the amplifier, in correct proportions, to the individual units of the system. b) do so efficiently. The "correct proportions" is a matter of distribution, the realisation of full available output, efficiently, is a matter of matching and impedance. Reconsidering the series versus parallel question first if we have to fed 16 ohm units, in parallel, the resultant impedance is 1.6 Ohms. If the wiring has a resistance of 1 Ohm (half ohm in each log, which is small for a big system) almost 40% of the available power will be used to warm the wire. Such small audio power will not appreciably heat the wire , but it is a serious waste, because it means that only 6o percent of power produced by amplifiers can reach the speakers. If we consider the series connection, the total impedance is 160 ohms, of which 1 ohm is less than 1 percent more than 99 percent of the available power reaches the speakers. But the damping question has not been considered. For high fidelity purposes, damping can be important. But when dealing with commercial sound installations, where voice reinforcement is the only function of the system, transient distortion would probably be unnoticed.

However, the fact that individual units are fed virtually constant current, rather than constant voltage, means that serious frequency colouration can occur, as well as the less noticeable effect on transients. While this is a more valid argument, the fallacy lies in the implied assumption that only the loudspeaker whose response we are considering (just one, not all of them together) has all impedance characteristic and all the others are resistances. Actually, all the speakers have impedance characteristics, which can mutually affect one together. This could aggravate the situation or minimise it, in different circumstances. If all the units are of the same type, then their impedance characteristics will be reasonably similar, so 10 of them in parallel will result in substantially identical impedance characteristic, of one tenth the value, while in series it will have ten times the value. Matched correctly to the amplifier (with nominal output impedance if 1.6 ohms or 160 ohms, respectively) the frequency response will be identical neither will be fully constant voltage or constant current, and each method of connection will receive substantially the same frequency colouration. Only if individual units have radically different impedance characteristics, as one unit having a high value at a frequency where the others have low values, or vice versa, will colouration at an individual unit appear. The overall energy fed into the areas may be less coloured than with identical units, but the colour differences between units will get exaggerated, with possible unpleasant effects in parts of the audience area. So, for assessing performance we have a relatively simple guide line viz., if the units are for the same type series is as good as parallel, and the circuit is much more efficient. If the units are different, parallel, will avoid possible. One of these is safety. Any individual unit may develop an open circuits or may develop a short circuit. In loudspeakers, open circuits are the more common fault. In parallel connection, a short circuit would 'kill' the whole line, while an open circuit 'kills' only the defective unit. In series connection, an open circuit 'kills' the whole line, while a short circuit only 'kills' the defective unit. Thus, as opens are more common, the parallel connection has a slight edge from the safety factor viewpoint. Another factor is matching the amplifier. Normally the output impedance is neither 1.6 ohms nor 160 Ohms. Hence, a matching transformer is required to produce the impedance match. The alternative is to juggle series/parallel combinations, in an endeavour to get combined impedance for which the amplifier does provide an output. But whatever combinations we consider, in however large quantities of speaker units, the steps from one arrangement to another always seems rather drastic and it is different to find a combination to give the overall power distribution required and come fairly cose to an impedance the amplifier happens, to provide wiring it may get even more involved. The solution, therefore, revolves around a design 'voltage' rather than a design 'impedance'. For matching a number of loudspeakers to an amplifier particularly for use in large multi speaker installations. This method is based on "constant voltage" system which simplifies to a great degree the computation of the proper transformer taps when varying sound levels are required. It also permits the addition of speaker to an existing system, without the recalculation of the load and source impedances so long as the total power consumed by the loud speaker is less than or equal to the amplifier rating, a favourable load condition is always exist. It permits loudspeakers to be connected a cross a transmission line with the same case that electric lights are loaded on a power line up to the capacity of that circuit. To use this method requires that a power amplifier

incorporate an output transformer tap which will deliver 70.7 volts at the rated output of the Amplifier. Amplifier over 100 watts power capacity employ a 141 volt tap under this system. The choice of 70.7 volts (referred to hereafter as 70 volts) took into consideration under writer requirements which in many locations limit loudspeaker circuit voltage to a specified maximum. The Railways in India prefer the 100 volts constant voltage system, originally introduced by the Philips Electrical Co. Reference to a "70 volt loudspeaker distribution line" does not mean that the voltage on this circuit will always be 70 volts. It may of course be less when the system is operated at levels well below the rating of the amplifier. The 70 volts is the maximum voltage on a sine wave test signal for a given amount of distortion. This expresses the standard for rating amplifier power. Standardising this voltage means that the voltage is the same for a low power amplifier as for a high power amplifier if rated under 100 watts. The condition of 70 volts at the output will exist when the amplifier is terminated in its rated load impedance and is supplying power to the load equal to the rating of the amplifier. However, to simplify calculations, it is considered at the rated output. How to use the Constant Voltage System. When transformer taps are marked directly in "watts" no mathematic are required. Simply choose a transformer with the correct power tap and connect to the desired terminals. For transformers marked in impedance, follow the procedure below. Actual calculations are simple. One basic formula is used: Z = = E2 P of required impedance

(Output Voltage) 2 desired power so that for an amplifier employing a 100 volt output tap, the formula is reduced to required impedance = 10000 desired power

The figure to bear in mind are 10000 (roughly the square of 100) and 5000 (square of 70.7). In actual practice, wiring for a quantity of loudspeakers becomes very simple. First, decide on the power required for each loudspeaker. Add up the power required for all the speakers and choose an amplifier capable of supplying at least this much power. Using the formula above, choose a transformer which will give this power on a 100 volt line. All transformer primaries may then be connected in parallel across the 100 volt line without any further consideration of impedance. Amplifier power does not enter into any of the calculations.

FIG 60 to be patched
Fig. 60 Constant Voltage line set up to deliver 5 W to each of 10 speakers from a 50W Amplifier For example, the typical industrial sound installation discussed under "Impedance Matching" may be laid out on a 100 volt constant voltage basis (assuming the amplifier incorporates a 100

volt output tap on the transformer). impedance values are: For each 10 watt speaker: Z = E2 = P For each 5 watt speaker: Z = E2 = P For each 1 watt speaker: Z = E2 = P

Using the basic formula, primary speaker transformer

10000 10 10000 5 10000 1

= =

1000 ohms 2000 ohms

= 10000 ohms

From the above, it may be readily seen that many conventional line matching transformers may be used with equal effect on constant voltage as well as constant impedance system as shown in figures 60. Where the derived values differ slightly from the actual transformer taps. It should be remembered that mismatches up to 25% are permissible. However, the following must be observed:1. The total power to be supplied to all speakers in the system must not exceed the amplifier rating. 2. The power rating of the matching transformer must be adequate for the power consumed by its associated speaker load. 3. Matching transformer secondaries must be terminated by an equivalent speaker load.

Line (Impedance) Matching Transformers: Line (Impedance) matching transformers are commercially available in a wide variety of impedance ratings. Four factors must be considered in the selection of the proper transformer: 1. Primary impedance (determined by the requirement of a given installation). 2. Secondary impedance (to match load impedance presented by a speaker or speakers connected to the transformer secondary). 3. The transformer must be capable of handling and transmitting an amount of power at least equal to the power consumed by its associated speakers. 4. The transformer must not discriminate unduly against any of the audio frequencies which are to be transmitted. The ability of the transformer in this respect is usually reflected in the price. A poor or improperly selected transformer may result in loss of tone quality, loss of power and an increase in distortion. What it does mean is that the available or peak signal voltage is regarded as a constant figure, at which value various currents, or wattages, may be drawn, for the determination of relative power throughout the system. Thus if ten 5 watt speakers are connected to a 50 watt amplifier, correct matching will be achieved (if the same 'constant voltage' design figure is used throughout) and each unit will receive one tenth of the total power.

Although for convenience in application, amplifiers, speakers and matching transformers are given a voltage and wattage rating, rather than their operating impedance, they are still designed on an impedance basis. Example: Problem: In a factory to connect 8 speakers of different voltages to an amplifier, so that two speakers in the factory yard require 25 watts each. Two speakers in the factory work shop 20watts each. Two speakers in the canteen to require 10 watts each. Two speakers in the supervisors room to receive 5 watts each.

Solution: For the above installation the required material is 1. 2. 3. Amplifier Loudspeaker Line matching transformers 1No. 8 Nos. 8 Nos.

The amplifier can be capable of delivering at least the required wattage i.e., 25 w x 2 20 w x 2 10 w x 2 5wx2 Total : = = = = 50 w 40 w 20 w 10 w 120 watts.

CONDITION: The power handling capacity of the loud speakers must be greater than or equal to the required wattages if they are with out line matching transformer if the loud speakers used are with line matching transformers the wattage rating must be equal to the required wattage. The line matching transformer should be capable of handling the required power.

Fig. 61

These installation is shown in the figure 61 using the impedance type line matching transformers. Calculation for primary impedance of the LMT. Amplifier Line Voltage is 100V. V2 P Where V is the line voltage and 'P' is the required wattage. Primary impedance of LMT ZP = Therefore for 25 watts ZP for 20 watts for 10 watts for 5 watts ZP ZP ZP = 100 x 100 25 = 100 x 100 20 = 100 x 100 10 = 100 x 100 5 = 400 ohms. = 500 ohms. = 1000 Ohms. = 2000 Ohms.

Directly loud speakers of required wattages can be connected if the loudspeakers are with built in line matching transformers as shown in the figure 62

Fig.62

A typical P.A. System for VIP functions is as shown in the figure 63. In this care should be taken for connection of the loud speakers. There are five Amplifiers used two main amplifiers, two amplifiers for standby and one amplifier is used for monitoring. Four microphones are connected to a mixer pre-amplifier. One tape recorder is connected to the same mixer for playing and recording the programme. A common output is taken and is given to all five amplifiers. Some mixers have more number of output terminals. Loud speakers of column type are placed at about 8 metres apart both left and right hand sides of the auditorium. Alternative loud speakers are connected to the Amplifier 1 and Amplifier 2 in both the columns of loud speakers. Care should be taken to avoid any coiling of the cable used for loud speaker connections. Power supply arrangements must provide a 100% standby with the Batteries. A proper neat cabling should be done in such a way that in any case it should not be an obstruction for the moving persons in and around the stage. Loud speakers should not be driven more than seventy five percent of its rated power. All speakers must be in parallel and are connected in constant voltage matching system only.