line array wiring question

[mcargill]

Wiring a number of similar drivers, as in a line array, is usually done by combining series and parallel groups to produce a tolerable combined impedence load for the amplifier.

But this puts the inductance of one driver's voice coil in the signal path of other drivers, which would attenuate the high frequencies, would it not? If the array were crossed over to a tweeter array, this might not be a problem, but in a full-range array, I would expect this would be detrimental to the high frequency response of the system.

If so (or even if not), I would like to know if wiring all of the drivers in parallel could be an alternative? This would, I am guessing, call for a high-power, high-quality, 8-ohm resister (or resister network) in series, to keep the amplifier happy. Would there be a problem with that?

Is there another solution? (An amplifier that produces 2 watts into twelve channels, perhaps, in a dodeca-amped system?)

While on the subject, I would also like to know if the inductances of the drivers in the array would have any other effects on the behavior of the system, so that the purity of sound achieved with a single full-range driver would be compromised by wiring it together with several others. For example, would the line array exhibit phase shifts or impedence peaks not evident in the drivers taken singly?

[Bill Fitzpatrick]

Originally posted by mcargill

1. But this puts the inductance of one driver's voice coil in the signal path of other drivers, which would attenuate the high frequencies, would it not?

2. If so (or even if not), I would like to know if wiring all of the drivers in parallel could be an alternative? This would, I am guessing, call for a high-power, high-quality, 8-ohm resister (or resister network) in series, to keep the amplifier happy. Would there be a problem with that?

3. For example, would the line array exhibit phase shifts or impedence peaks not evident in the drivers taken singly?

1. It would not.
2. Do NOT parallel drivers and add a series R to regain an 8 ohm impedance.
3. No

[Eccu]

Originally posted by Bill Fitzpatrick


1. It would not.
2. Do NOT parallel drivers and add a series R to regain an 8 ohm impedance.
3. No

And why not?
I was thinking these same aspects also.

[Bill Fitzpatrick]

I'm not telling.

[Eccu]

Originally posted by Bill Fitzpatrick
I'm not telling.

If you cannot give any explanation about your previous statements, we have some credibility gap

[Bill Fitzpatrick]

[Eccu]

Originally posted by Bill Fitzpatrick

Thanks

[mcargill]

So far, I think we have learned more about Bill Fitzpatrick than about wiring line arrays. Here is my thinking (and if mcargill had thought some more I imagine he could have come up with this):

A single speaker exhibits the electrical characteristics of a resistance in series with an inductance, and the effects of these two factors are implicit in the behavior of that speaker when driven alone.

If we wired, say, ten such speakers together in series, the inductance in the circuit would increase proportionately. But so would the resistance. And since the resistance and inductance are proportionate quantities in the formula for frequency attentuation, the net effect on the frequency response of the circuit would be nil.

What is more, if the speakers were wired in parallel, there would be the complementary problem of how the multiple inductances in parallel would affect the circuit. Each of the ten drivers would appear to be wired in parallel with an inductor 1/9 the value of the voice coils taken individually. This appears to threaten to attentuate the low frequency response of the system. But once again, this combined impedence value is in parallel with the combined resistances of those drivers, and this effectively counteracts the suspected effect of the multiple inductances. Again, nil effect.

So one reason why one should not bother with all-parallel wiring is that there is no need for it. So save the trouble of having to cope with the resultant ultra-low impedence, save the cost of the hefty resistor, save your amplifier power that would otherwise be just generating heat in that resistor, and avoid introducing new distortions due to that added piece in the circuit.

But then, what if an array happens to exhibit some wild impedence behaviors (for whatever reason, probably unrelated to their series-parallel grouping)? Why not tame that impedence variation by wiring the drivers mostly in parallel, producing a very low impedence circuit, and a correspondingly low impedence variation, then adding a simple resistor in series to bring the load up to a level that does not threaten the amplifier? If a driver swings from 2 ohms to 30 ohms, and is driven full-range so that every bit of this impedence variation is relevant, wiring ten in parallel would reduce that swing to 0.2 - 3.0 ohms. Adding an 8-ohm resistance in series would produce a more manageable 8.2 - 11 ohms load. Would this ever be worth it?

[Bill Fitzpatrick]

The man answers most of his own questions. Similar reasoning will answer the last.

[dnsey]

One way of avoiding this problem (if it exists) is to use higher impedence drivers!
I have some PA arrays (Bose clones, sort of, - not line arrays) which I made up using ten 35 Ohm 6.5" drivers: I picked up 100 of these in a clearance sale, but similar types are fairly easy to obtain.