line array, mulitple drivers, dipole .. questions that need answers!

[JinMTVT]

I have a few direct questions, i've done researches on this board and didn't find what i was looking for.

- Is it true that by using multiple drivers ( sould it be in line array, or side by side or whatever config that gives good results)
we distribute the work to attain the same SPL and thus reduce distortion made from drivers at higher spls ?

-then, is this profitable VS the compromises of a big line array?
or should we be pointing toward using better lower distortion drivers?

- i've read more than one in here that most drivers suffer from 1 to 3% distortion ??? OMG that is a huge number compared to what we are working for with amplifiers?? is this less revelant that one might think ? how does it affect sound fidelity VS amplifier distortion?

- when designing a dipole ( open baffle) loudspeaker system,
what are the important factors to look for in drivers?
i've read that we need higher Qts to compensate for what would be generated in an enclosure ?? what is the point of using higher Qts value ? what will it bring ?

- does it apply to tweeters? are common tweeters affected by dipole operation ?

- what are the negative sides of using dipole configuration
for a loudspeaker system ?

i guess it all comes down to compromises, usually we can't get 100% of everything, so we need to make choices and compromise elsewhere ... does it applies well to accoustics ?

thanks again very much for your time and efforts

sorry that my ??? are sooo noobs, i strive to learn as much as possible from this site!!

[Pan]

Yes, multiple drivers sharing a load will give less dist. Optimal is of course to use plenty low distortion drivers.

I´d say stay away from high Q drivers. Use EQ instead. A high Q driver will always have worse transient response than a well damped one. Check out SL as a refference, he use low to normal Q drivers. What you should look for is noise/turbulence and also the driver should have decent linear displacement.

Tweeters are influenced by dipole operation "exactly" as bass/mids.

One clear drawback with dipoles is the strong backwave that mess up the sound if the speaker is to close to the wall behind it. Diffusion and/or absorbtion is worthwhile here.

/Peter

[FrankWW]

Hi,

General comments from a non techie who's been to the school of hard knocks.

With speakers everything's a tradeoff.

The best speaker is a full range one element speaker but such a creature doesn't exist with satisfactory sound pressure levels or low distortion at the frequency extremes.

So we have to use multiple drivers of one kind or another.

Disadvantages

The outputs of the multiple drivers can interfere with each other causing distortion.

(This is especially the case with linear arrays in residential setting. Linkwitz mentions in the quote you posted in the other thread that they're suitable for public venues, not home use -he doesn't make it clear -at least in that quotation - that you can't get far enough away from them in small rooms).

Have to have a crossover which, especially if passive, can introduce more distortion.

Advantages

Multiple speaker elements means smaller excursion for producing high frequencies and larger excursions for drivers producing lower frequencies -you've got to move lots of air for good LF response.

Probably the most satisfactory and easiest to implement compromise is a two way speaker or three way depending on your "box" design and your room.

Regarding dipoles.

They give a nice, natural sort of sound. But you cannot get really loud low frequencies from them because they can't produce a static pressure wave. Sealed box speakers can.

Linkwitz recommends tall speakers because, I believe, he says they're more "natural" or something like that. (He's right because the bounce off the floor happens later and is not as loud. I'll get to that below). Linkwitz's speakers are tall but they're not line arrays. If I remember correctly, he uses a wide range mid and high and fills in the very bottom with the dipole woofer. Essentially, what he's got is a two way dipole arrangement with a dipole sub on the floor.

Most of us listen in small rooms and a lot of speakers are not suitable for small rooms which present different requirements from large ones. Omni directional speakers create psychoacoustical problems in small rooms (especially rectangular ones) because of reflections off the walls and floor. If these relections arrive too soon and too loud we don't get the perception of spaciousness because much of the ambient information that might be in the recording is masked.

Dipoles have an advantage here because there is a 'null' in their radiation pattern at the sides, and therefore much less sound bounces off the sidewalls. You can take advantage of this characteristic by making sure the driver you choose radiates a bit narrowly itself

.Make sure thedges of the baffle are rounded and the distance from the speaker to the edge of the baffle varies to cut down the difractive effect.

Put a carpet on the floor in front of the speakers.

[KaffiMann]

Sorry about reviving an old thread, but if this is what it takes, so be it.

I am wondering about the total q of the system.
I know people use resistors to trim the q value, but what about the difference of wiring drivers in series/parallel?
Say, if i want to use some specific drivers in a dipole setup, these drivers have a low fs, and a low qts. If i can make up for it in the amp section, will i benefit from driving a higher impedance load (like 16 ohms or maybe even higher, vs 4 ohms)?

I do not want to have a lot of junk in my signal path, i want a neat and tidy circuit without a lot of active and passive elements, just barely enough to make the sound right. If i can wire it differently to get a similar result to eq i will prefer that over adding another thingy to my system.

[LineArray]

The question whether series or parallel wiring will affect Qts has
been asked more than once in here ...

Assuming the drivers to be equal there is no difference in Qts,
whether you wire in series or parallel.

When drivers differ due to tolerances the resulting impedance
curve will look more "gaussian" like if wired in parallel.

When wiring differing drivers in series you may observe more then
one peak in the impedance when you spread the curve widely.

So all in all i would say when combining series and parellel wiring,
that series wiring should not be dominant.

If you are a perfectionist, you can select those drivers to be
connected in series due to resonant frequency being as close
as possible.

The differing "series groups" can then be wired in parallel.

Whether you choose a higher or lower impedance will depend on
the behaviour of your ampifier. There are amplifiers getting along
with low impedance very well. Then it dependes mainly on the
desired voltage sensitivity of the final speaker which impedance
to choose.