[brj]

The use of so many identical drivers to cover a given frequency band in a line-array speaker provides a real benefit by reducing individual driver excursion levels significantly, thus allowing the drivers to easily remain within their regions of linear performance and providing very low levels of distortion.

So, my first question (to provide background to my primary question) is: What do you give up in a speaker driver to get increased excursion and thus, hopefully, a greater region of linear response?  New driver designs seem to offer ever increasing levels of low distortion/linear performance, but is there a performance tradeoff required to achieve this?  (Sensitivity?)


Moving forward to my primary question...

Despite the benefits of using many low excursion drivers in a line array concept, I see one potentially significant difficulty - you now have to match the level and frequency response of a lot of drivers.  Great imaging is usually attributed to extremely accurate level matching between drivers in the left and right channels.  The driver compliment of your average line array would seem to complicate this a great deal.

So, for those among us that have designed and/or built line-arrays, to what extreme do you attempt to match drivers?  Do you simply rely on the quality control of modern manufacturing processes?  Do you still check drivers by hand for matching responses?  Is there some other feature of the line array concept that makes driver matching for great imaging a non-issue?

Thanks for any thoughts!


(TCG, I considered posting this question in "The Lab", but ended up leaving it here with the thought that it could end up as a rather free-ranging topic.  If you want to move it, please feel free!)

[JLM]

While level matching of randomly selected drivers of the same model is a valid concern (and rarely done for any speaker), you've eluded to the far bigger imaging issue with line arrays, that being the inherent inability to image along the axis of the line (normally vertical).  The theory behind line arrays is to produce the same sound wave, floor to ceiling and for that wave to only travel horizontally, therefore image height cannot be defined.  And until the sound waves from each driver coalesces with adjacent sound waves there can be no single image either.  So why worry about the relative subtly of driver matching?  (This two dimensional propagation provides increased overall efficiency, OTOH smaller drivers are less efficient than larger ones.) 

To produce a given spl at a defined distance in a defined room and at a given frequency you must displace a fixed amount of air.  Except for horn loaded compression drivers and bass tuning of the rear waves found in horns, quarter wave pipes, and transmission lines you are looking at driver area and travel as the involved factors.  Smaller drivers must travel farther to displace the same amount of air.  As the time needed for travel distance increases dynamic response goes down.  This is less of the problem as frequency increases, the number of drivers is increased,  and/or as spl's decrease.

Quality control of drivers is a hit or miss proposition by brand, cost, or type.

Sorry, but I can't answer how modern designs allow for increase excursion.  Good question.  Hopefully one or more of the guys from AV123, DIY Cable, GR Research, etc. will respond.

[Danny Richie]

What do you give up in a speaker driver to get increased excursion and thus, hopefully, a greater region of linear response?  New driver designs seem to offer ever increasing levels of low distortion/linear performance, but is there a performance tradeoff required to achieve this?  (Sensitivity?)

With any woofer there is always a trade off of sensitivity for low end extension or visa versa.

So, for those among us that have designed and/or built line-arrays, to what extreme do you attempt to match drivers?

Actually it is less of an issue. Instead of having to match a pair of drivers, your are matching the average of a group of drivers.

Do you still check drivers by hand for matching responses?

With the planar magnetic drivers used in the Alpha LS or the new LS series designs, I check the impedance of each driver and match them in that way. Matching our woofers has been a waste of time. The only real variance is in the top of their range (well beyond the crossover point).