Settings for tri-amping an '82 Belle with an active digital crossover

[Chris A]

12 minutes ago, Khornukopia said:

Using a two way bi-amp system for example, crossed over at 500 Hz, both slopes at 6 db/octave and the drivers time aligned, what happens if I change just the high pass filter to 48 db/octave? Did I introduce a one wavelength delay to the tweeter signal, the equivalent of physically moving the tweeter backwards 2.26 ft.?

 

Yes.  This is assuming you're talking about Butterworth or Linkwitz-Riley filters.  I haven't done the same experiment using Bessel filters.  Bessel filters have much lower phase shifts than Linkwitz-Riley or Butterworth.

[Khornukopia]

Then inversely, one could apply a steep slope to the horn loaded bass driver and a shallow slope to the tweeter to time align the loudspeaker? 

[Chris A]

Well, no not really.  However, the resulting phase shift will be lower than using steep slope for the high pass and low pass.

 

Perhaps this presentation (see slide 22 for the description of the filters used and the time delay) will help you eliminate much of the crossover-induced phase lag using a high pass Bessel and a low pass Butterworth: https://community.klipsch.com/applications/core/interface/file/attachment.php?id=148433

 

Also see this diyAudio thread: http://www.diyaudio.com/forums/multi-way/277691-harsch-xo.html

[Khornukopia]

Where would I use an all pass filter?

[Chris A]

An all-pass would be used to correct for driver-induced phase errors--like the type of added "real-world response" seen in the diyAudio thread.  Other than those kind of issues, I see no need for an all-pass filter to implement the Harsch crossover.

 

Note also that @Edgar(Greg Berchin) has mentioned that a Bessel-only on both high-pass and low-pass would probably be superior in terms of summed frequency and phase response.  Since I haven't implemented either filter schema yet (Harsch or Bessel-only), I can't give you any more specific advice.  When I do, however, I'll be sure to report back my findings (and there are always a lot of those...).  I'm actually kind of excited to do this, but events always conspire to delay the work.

 

Chris

[Chris A]

By the way, if you're thinking about using an all-pass filter instead of a delay of 180 degrees of the higher frequency channel in the case of the Harsch crossover, I'd try a first order all-pass at the center crossover frequency.

 

Chris

[Edgar]

17 hours ago, Chris A said:

Note also that @Edgar(Greg Berchin) has mentioned that a Bessel-only on both high-pass and low-pass would probably be superior in terms of summed frequency and phase response. 

 

 

 

Chris, when you get around to experimentation with the matched-delay crossovers, let me know. There is at least one variation that might interest you -- uses a Butterworth instead of a Bessel. Not quite as good with the phase response, but the slope of the derived highpass is increased by approximately one order. I've never actually tried this configuration, but the simulation looks good.

 

Greg

[Khornukopia]

Using a digital active crossover allows you to experiment in a controlled manner.

 

[MEH Synergy]

Damn those are beautiful! 

[AHall]

43 minutes ago, Khornukopia said:

Using a digital active crossover allows you to experiment in a controlled manner.

 

 

Whats the story on that beauty?

[Khornukopia]

Thanks to @jwgorman for letting the forum know about the sale price on the ESS AMT tweeters, and thanks to Chris A for testing the AMT, which helped persuade me to buy a pair of the tweeters that have intrigued me for a long time. Another thing I have wanted since my very first K-horns is tweeters that can be aimed, and since the AMTs don't have a typical mounting flange, I assembled some tilting brackets. My plan is to install these inside the top hats. They work really good on these tri-amped Klipschorns. Compared to other tweeters, the sounds of cymbals and chimes tend to shimmer for a moment longer, with very fine detail.