Tapered array crossovers in Klipsch Speakers

[Raider]

I have seen tapered array crossovers referenced in some of the Reference line literature. Crossovers in this configuration minimize lobing at frequencies shorter than the spacing between multiple drivers, and allow a better defined midrange while maintaining the advantage of the cumulative bass response of multiple lighter-coned drivers . Such a configuration is apparently used in the RC64. Do the RF83, RF63, RF82, RF62, RF52 also use this configuration? I surmise from the limited language in the literature that this is probably the case.

I find it interesting that the sales info pretty much glosses over this aspect of the design, or at least fails to develop enough detail to make the significance of this design technology evident.

For instance, multiple drivers with lighter cones operating in unison allow the bass drivers to better match the exceptional efficiency of the horns, while simultaneously allowing exceptional transient response over that of a single, heavier cone. By using this crossover design, the mid range interference between drivers that would result normally is eliminated or mitigated. Drivers in this configuration also limit vertical dispersion, which allows the dispersion pattern to better match that of the horns. This results in a dispersion pattern that is inherently applicable to many room environments and HT applications. So i think the work that Klipsch engineers have done to optimize the overall design and emphasize the inherent strengths of their heritage go unfortunately unnoticed at worst, or minimized at best. It's really remarkably ingenous work.

[Jay481985]

I believe due to the horizontal nature the comb filter effects come to play and the tapered array is used while vertical mounted speakers do not exibit such effects and do not need tapered arrays.

[Raider]

I believe due to the horizontal nature the comb filter effects come to play and the tapered array is used while vertical mounted speakers do not exibit such effects and do not need tapered arrays.

Parts of this white paper on line arrays by Jim Griffin would suggest that it can be perceived in both vertical and horizontal orientations....

http://www.audiodiycentral.com/awpapers.shtml

[Jay481985]

less so in the vertical plane.

[Al Klappenberger]

If I understand what you are talking about here, there are multiple lowpass filters connected in parallel such that some drivers are cut out above a certain frequency leaving others still operating. The bug here for me is that the impedance seen by the amp jumps up as each lowpass switches off above it's cutoff. If the filters are singly terminated (as they should be), they don't care, but a cheap amp might loose it when the load impedance suddenly jumps up by 2:1! This might be fixed by "diplexing' the switched-out driver's filter into a dummy load to hold the impedance constant, but it becomes a lot of parts and a big price tag to do it right. I never thought about how to actually arange the filters to do this though. It might wind up a real mess!

Al K.

[DrWho]

Dang Raider, you're really doing your homework [Y] To answer your

question, the floorstanding speakers do NOT have a taperred array

crossover - it is something only used in the center speakers.

I believe the reason for this is that the woofers are going to have a

naturally decreasing polar as frequency goes up, which will in turn

induce less overlap. I suppose the same could be argued for the center

channel, but the difference between the two designs is the targeted

off-axis response and slight differences in the driver spacing.

A center channel by its nature needs to be a wide dispersion design,

but the horizontal array significantly narrows the horizontal polars.

In the case of the floorstanders, a narrow vertical is often a good

thing and the listening position is rarely very far off-axis from the

speaker - so there is no need for increasing the vertical dispersion of

the mains.

In addition to all that you have the issue of matching the polar

response of the woofers to that of the HF section to allow for a smooth

crossover transition. The tweeters implemented on all of the reference

lineup are 90x60 nominal. Assuming the vertical array of speakers

introduces a 60 degree nominal for the woofer section, then the

taperred array is working to extend that natural dispersion to 90

degrees on the horizontal array.

As far as the impedance response shown to the amplifier - I don't think

it will be that bad. Every direct-radiating ported driver configuration

has a similar impedance response with peaks on either side of the

tuning frequency, which is going to be the region that the secondary

woofers are being slowly brought into effect. That and the minimum

impedance usually occurs at a much higher frequency...so in the end the

crazy bumps inherent in ported designs at the lower frequencies will be

rounded out by the parallel loading of the drivers with little change

at the higher frequencies. And you most certainly don't need a ruler

flat impedance response for a speaker to sound good. I think I might

need to draw a picture to make the point a bit more clear...

[colterphoto1]

Isn't it mostly to eliminate timing differences for off-center listeners in the critical vocal range of the dialog (center) channel speaker?

Doc- you got my head spinning with that last one.. []

[Al Klappenberger]

Doc,

I wasn't aware that the transition from multiple drivers to one was done so low in frequency. If that's the case passive networks would have HUGE values and would not be practical to make any more agressive than first or second order. An actual "absorbative" design having energy routed to an internal dummy load above the transiton would be the size of Texas!

Al K.

[DrWho]

I'm not very well versed on line-array theory and crossovers, but I do know that phasing differences between drivers can be beneficial in arrays and that different kinds of filters affect the phase response differently - I wouldn't doubt that a higher taper crossover point could be achieved by taking this into account. Though I suppose that makes the impedance issue a bigger one [] (but it certainly improves the efficiency of the system - so is the amp really working harder?)

Ultimately I think it comes down to choosing the compromises and putting the money where the most sonic return exists. I certainly wouldn't want to spend a lot of money on a load that's easier to drive...that kind of money could be put into an amplifier than can drive the hard load and yield other sonic benefits at the same time.

Or you could just screw the flawed horiontal array concept in the first place and just put another identical floorstanding speaker in the middle. Now you've got perfect timbre match and no crazy polar responses to deal with. I suppose in Raider's situation I would highly recommend this approach as it would be easily implemented by mounting the flat screen TV on the wall directly above the center floorstanding speaker. And due to the insane height of the RF83 it might help to make his final speaker decision a bit easier []

(In my limited experience it is far better to have perfect timbre matching than the ultimate in mains and a slightly lesser center - our ears are cool in their ability to filter out flaws and by going to timbre matching out ears only have to do one set of filtering across the front array which makes it less fatiguing).