A K-402-Based Full-Range Multiple-Entry Horn

[Chief bonehead]

 

In the meantime, am wondering why straight axis bass horns of the past needed to be so very much larger? What is it about the Unity/Synergy style and your Klipsch K-402 based full range horn that defies past limitations? Is it the slight flare at the end of a short cone? If one added this flare extension to the exit of a La Scala, K-horn or Jubilee, will it transform into a profoundly superior speaker?

 

 

I don't wish to give anyone the impression that "new physics" is involved -- it isn't.  The issue has always been the so called "mid-bass problem".  PWK struggled with it in the Khorn design, changing the compression slot in its bass bin to be smaller in order to to improve its well known issue to reach 400 Hz.  One of the reasons why I believe that some people prefer the La Scala or perhaps Belle over the Khorns' bass bin (also perhaps using a subwoofer with the La Scalas/Belles) is that the Khorn performance over its decade of usable bandwidth (~32-400 Hz, nominally) is about at the limit that one can achieve with a bass horn--even folded. 

 

The Jubilee bass bin has the same issue, albeit with much better on-axis higher frequency performance than the Khorn bass bin (designed as the Khorn II toward the end of PWK's life with Roy D. assisting in the specifics of design and testing).  The result was the "Golden Jubilee" of the Khorn, renamed because its performance was that much better than the Khorn's bass bin that it deserved a new designation. 

 

The remaining issue with the Jubilee bass bin is its very narrow polars above 220 Hz due to its dual mouths interacting to form, in effect, an array of two emitters (horn mouths), which narrows its HF polar performance significantly, just like in the Khorn bass bin. (This also applies to the Belle at a higher frequency--about an octave, but not so much the La Scala) 

 

Also there has always been the issue of vertical separation of the horn mouths in the Khorn and in the Jubilee - more than 1/4 wavelength at the crossover frequency.  These vertical separations cause disturbances in horizontal and vertical coverage at the crossover frequencies (also between the tweeter and midrange, too), as well as the psychoacoustic effect of a non-point source loudspeaker over a point source one.  The effects of this single design detail was noted in the Altec 604 and later horn-loaded and reflex loudspeaker designs.

 

These really aren't revelations, only design trade-offs that may or may not be well known. All loudspeaker designs have design trade offs and strengths/weaknesses--that isn't really a revelation, either.

 

So the current horn design makes use of a different set of design trade offs than Khorns or even the Jubilee, recognizing the strengths of the K-402 horn design by Roy and the multi-entry horn design (i.e., Unity horn) of Danley.  So the subtleties of these two designs are put together into a new design that hopefully makes use of the strengths of those other designs and not their weaknesses. 

 

So the current design (New Center) makes use of the ideas of a point source design, multiple entry horn, and conical horn design below its Fc to yield a design with different performance and physical size characteristics that address the design weaknesses of the prior designs.  It also introduces it's own design weaknesses, of which I've identified a few in the above postings.  The idea is to produce a design that addresses customer needs/wants that aren't currently being addressed by the prior designs.

 

Some of those needs include things such as: better midbass performance in terms of polars, single point source loudspeaker (better imaging/presence), much smaller package, not having to use a room corners if they aren't available, lower cost, and perhaps a couple more customer needs/wants that haven't been addressed by the existing products (Klipsch and Danley, etc.).  The current New Center design has the potential to do that, IMO. 

 

So the current horn uses the real strength of the K-402's conical horn and tractrix horn properties, but simply extends that performance to its limit, using the same design trade that Danley uses with its full-range multiple entry horn designs.  Nothing really new by itself, but the combination IS new. 

 

Also note that the current design really doesn't make use of the Khorn, La Scala, Belle, or Cornwall designs because those designs don't have the design characteristics of the K-402 or the Danley Unity horn designs that are needed. Otherwise, it would look like a kit to update those designs.  That's really not the case in this instance.  It would be nice if it could.

 

So, in a nutshell, the current design combines the strength of the K-402 horn profile with the conical multiple-entry horn into a single aperture full-range horn loudspeaker design.   It recognizes some design trades that others might have missed.  Hopefully, the weaknesses of this design are minimized to create a significantly better design for a large customer segment whose needs aren't currently being addressed through current products: Jubilee or Danley Unity/Synergy designs.

 

Chris

 

just a small detail....the 402 is not a conical expansion.......

[Chief bonehead]

chris.  i am impressed with the way you pursue things.....passion, data and being honest with your ears.  keep it up!!  

[wvu80]

One day......maybe.....I will show you guys things I have in the vault........

 

For the present, we'll settle for pictures from the Klipsch Book of Secrets.   

[ClaudeJ1]

just a small detail....the 402 is not a conical expansion.......

 

Well, we don't need to know the recipe for the "secret sauce." Like my favorite Marinara on Pasta, I just enjoy eating it. As long as we can put good drivers on it and have it sound great, that's all that matters. No need to reveal the recipe.

 

In the case of the K-402, it's just the best sounding horn mid tweeter I've ever owned. Although I do like the Danleys too. It's like all the super cars out there to compete with Ferrari and Porsche, they all go over 200 miles per hour.

 

I lived with 402's and 1133's for over two years. My "new ones" have TAD 4002's and the "dirty curves" look great. So I'm working on a woofer section while I ponder how to build some out of Rosewood veneer and some well placed slots. It will be a while since I have other projects. As an easy start, I will try double 15's in a U or H Open Baffle, since the bass I have heard from these is pretty detailed when done right, a la Martin King and/or Sigfried Linkwitz.

Edited January 20, 2016 by ClaudeJ1

[Chris A]

just a small detail....the 402 is not a conical expansion...

 

The way that it is constructed--I agree with you, but for its effective properties related to flare rate expansion vs. off-axis port location (the low-pass acoustic filter), it behaves just like the horns that Danley uses. 

 

So for the simplicity of calling it something other than "the straight-sided walls of the modified tractrix horn", I'll just refer to that area as a "conical", even though it doesn't expand on an x^2 profile.

 

Chris

Edited January 20, 2016 by Chris A

[Chris A]

so there is nothing that is being compromised? nothing that is being minimized?...

 

 

Well, certainly there is, but to show the relative trade, I'd have to show detailed harmonic distortion plots for performance below the 1/4 wavelength frequency (174 Hz) relative to the Jubilee bass bin harmonics for those same regions, and those are sensitive to SPL.  I didn't want to get into that here due to the more complicated nature of how those harmonics show up as modulation distortion sidebands that are either masked (lower order) or unmasked (higher order) in corresponding dual-tone bode plots.

 

At the end of the day, it gets down to "can I hear those sidebands or not?", and that is much less easy to show and talk about, especially with regard to Jubilee bass bin performance as the baseline.  I'd have to show those detailed sidebands for both, in gory detail.  Right now, I've decided to try to keep it a bit simpler for the readers here.  Note that I don't have to suppress anything--but I think that isn't going to make it any clearer for the readers if I do present everything, but rather provide fodder for misinformation campaigns for those with an axe to grind--solving nothing.

 

If you want that data, I can get into that.  The fact remains that simple harmonic distortion plots don't tell you what you need to know to make a decision with regard to "what's audible and what's not".  Geddes would have a field day in claiming that none of it is audible, because he's not paying attention to levels above 80 dB, for which he limited all his data collection on subjective headphone listening testing with college students.  No one that I know has posted side band reference levels that are and aren't audible vs. frequency and harmonic number--vs. SPL.

 

I've decided to go with my ears and my wife's ears, with later harmonic distortion plots that can be made available--but privately to anyone that requests it, along with the dual-tone plots.  And because harmonic/modulation distortion isn't the lion's share of what's really important at those frequencies -- point source (among other related topics) DOES make a big difference.

 

Bottom line: it isn't simple enough to post single number measures of merit scores, because it isn't a single measure score that corresponds with what I'm hearing.  It's a lot more complicated than that, unfortunately.  Even if I had a full-up Klippel R&D rig, it still wouldn't be simple.

 

One question to ask yourself: why don't any of those Danley Synergy owners complain about bass distortion or flabbiness below ~200 Hz?  They're reporting instead that they are the best sounding speakers that they have ever heard.  There's a clue there to how complicated this is for data presentation vs. what does it actually sound like.

 

I'm sticking more to the latter, because that's what I believe is important for this frequency band (40-175 Hz)--and not so much a single figure of merit.

 

Chris

Edited January 20, 2016 by Chris A

[Chris A]

One thing that I can say that makes all this much easier to understand:  if it sounded like a KPT-942 (i.e., a K-402 sitting on top of a dual 15" woofer reflex box)--or even not as good as it--it would be easy to abandon all this and just use a 942 configuration. 

 

 

But it doesn't--not even close.  When I say that the imaging and overall sound is better than a Jubilee (IMO), that's a significant statement that I don't take lightly.   I've lived with Jubilees virtually every day for the last 7 years, and I can tell you that I know what they sound like. 

 

This one is different.  My wife commented on how much better it sounds...in a way that it makes it clear that she doesn't want to change it out.  It's here to stay.

 

Chris

[ClaudeJ1]

One question to ask yourself: why don't any of those Danley Synergy owners complain about bass distortion or flabbiness below ~200 Hz?  They're reporting instead that they are the best sounding speakers that they have ever heard.  There's a clue there to how complicated this is for data presentation vs. what does it actually sound like.   I'm sticking more to the latter, because that's what I believe is important for this frequency band (40-175 Hz)--and not so much a single figure of merit.

 

As best as I can recall from 2007, Roy has always commented there is still output from a horn below it's cutoff frequency, obviously at lower intensity, but with some EQ, it still sounds good and the distortion is still relatively low.

 

How else could we explain the sonic popularity of LaScalas, even though the horn loaded part of the bass is anemic as compared to the Khorn, Jubilee, and MWM. 

 

I'm amazed at how tight and "micro-defined" the bass is from my SH-50's in such a small (but heavy) triangular cabinet (about 4 cubic feet).

 

Hearing is believing. I'm glad you are reporting a similar experience to my own in this type of configuration in your Center.

Edited January 20, 2016 by ClaudeJ1

[Chris A]

Perhaps another way to look at the situation: we all tend to want to pull everything into separate factors or problem parts, and deal with each part separately.   Unfortunately it isn't always that simple.  Sometimes we underestimate the overriding effect that one factor, or set of factors, might have on another set of factors. 

 

In this case, I realize that the bass distortion performance (modulation distortion, and to a more limited degree harmonic distortion) has been essentially the sole subject area of interest to date.  PWK gave us that discussion and that focus, but to my knowledge, he didn't write on the current subjects that I believe override the decision making when you correct for those factors:

 

1) the importance of getting a single point source of sound with all sources within 1/4 wavelength of each other, time aligned (i.e., within 1/4 wavelength in the z-axis), even at the lowest frequencies that the human hearing system can detect--below 400 Hz as it turns out, and,

 

2) the importance of smoothly controlled polars below 400 Hz - even down to 100-200 Hz (subject to approximately the same 1/4 wavelength constraints as above).  This may be just as important as the point source effect, in "1)" above.

 

 

Once you do those two things, then the trade off vs. a loudspeaker design that doesn't respect those above two factors, even if it has significantly lower bass modulation distortion at higher SPLs, just cannot override the listening advantages that the loudspeaker that has marginally higher bass modulation distortion and perhaps even higher phase/time domain distortion of a reflex bass bin. 

 

This is much closer to what I'm finding in my extended listening trials. Note that all this is apparently governed by threshold performance levels, below which further improvement in a particular performance area is no longer clearly audible. 

 

Chris

Edited January 20, 2016 by Chris A

[Chris A]

Note that I personally do not wish to add the reflex bass distortion (phase/group delay and impulse distortion) of reflex ports to the current design, instead preferring the tighter bass of a closed box system over a vented one, despite having its harmonic distortion rise due to EQing the bass performance to be flat again. 

 

Since the current dual 15" woofers themselves have sufficient piston area to do this more easily vs. the existing Klipsch Heritage designs, the effects of the resulting modulation and harmonic distortion increases are minimized by using a larger total woofer piston area behind the horn in compression chambers to control piston movements to a minimum, providing better loading to the drivers. 

 

Bass modulation distortion is still important to minimize--just not the only important factor to pay attention to.

 

Chris

Edited February 12, 2016 by Chris A

[ClaudeJ1]

Chris, my own listening experiences over the last year and a half, in different rooms and configurations of point source horns, mirror you most recent experiences. After discovering, back then, all the things you mention, and still amazed by them today, I totally agree with your last 2 posts.

Edited January 20, 2016 by ClaudeJ1

[mark1101]

I would venture to say that you don't hear complaints from many Danley owners because there aren't very many.  They are very expensive and not designed for home use.  His customers are stadiums, churches, hall, arenas, some cinema.........and a few groupy audiophiles that wouldn't dare complain after spending that much money.  

 

I'm not saying they are not very good sounding...........I just don't think there are very many conventional owners.

[ClaudeJ1]

I would venture to say that you don't hear complaints from many Danley owners because there aren't very many. They are very expensive and not designed for home use. His customers are stadiums, churches, hall, arenas, some cinema.........and a few groupy audiophiles that wouldn't dare complain after spending that much money.

I'm not saying they are not very good sounding...........I just don't think there are very many conventional owners.

This is the funniest thing I have read in a while! Thanks for that one.

Like your Avatar which screams "not designed for home use! Since we both sold our Khorns in 2007, I also had a 5-way MWM Stack, as subtle as a pair of Hummers in a living room, like yours, but the poor man's version.

It appears that Chris have had the same goals. Improve sound at home and reduce the space while increasing placement options for better imaging.

On two occasions, I heard the very best sound I have ever heard from a fellow Forum member. His SH-50s and 4 Epic subs driven by a NAD 390 DD and fed from a $400 laptop with JRiver software. He is an Architect, Bass Player, and has done his own recordings. His room size and room treatments mean he needs no EQ to get 3D sound. The speakers just disappear into a deep soundstage. On some recordings, the sound even comes forward as well as behind but never FROM the Danleys.

My next house will have similar dimensions and room treatments for sure, now that I own some of the best transducers, it will be time for the most important upgrade.

Edited January 20, 2016 by ClaudeJ1

[DrWho]

I would venture to say that you don't hear complaints from many Danley owners because there aren't very many.  They are very expensive and not designed for home use.  His customers are stadiums, churches, hall, arenas, some cinema.........and a few groupy audiophiles that wouldn't dare complain after spending that much money.   

Since you bring up the intended audience....the most common complaint I've heard is SPL handling - especially at high frequencies. Even when not pushing the levels, several complain that the highs are attenuated a lot more than other cabinet designs when listening in the far-field.

[ClaudeJ1]

I would venture to say that you don't hear complaints from many Danley owners because there aren't very many. They are very expensive and not designed for home use. His customers are stadiums, churches, hall, arenas, some cinema.........and a few groupy audiophiles that wouldn't dare complain after spending that much money.

Since you bring up the intended audience....the most common complaint I've heard is SPL handling - especially at high frequencies. Even when not pushing the levels, several complain that the highs are attenuated a lot more than other cabinet designs when listening in the far-field.

Not true at 10-25 feet at home. Still rediscovering more musical detail than ever. Edited January 20, 2016 by ClaudeJ1

[DrWho]

I just ran WinISD on a ported box configuration.  If the box volume were about what the current box volume size is, and the ports were tuned for 40 Hz (four 2" ports about 15 inches long), there would be about a:

 

0 dB gain at 34 Hz

7 dB gain at 40 Hz,

3 dB at 50 Hz,

1 dB gain at 60 Hz.

 

40 Hz ported cabinet gain.GIF

 

This would add a rather large phase shift starting at 50 Hz (about 360 degrees), and group delay of well over 30 ms (i.e., not a very attractive approach). but I'm sure that harmonic distortion, modulation distortion, and required amplifier gain would be lower.

 

Perhaps there is a way to adjust the "Q" a bit, and control the phase shift, which is massive in the current example.

 

I still think that the current closed box configuration (acoustic suspension), using more amplifier power is a more "hi-fi" approach and the harmonic distortion levels would be lower than trying to use any direct radiating subwoofers, and cleaner than using TH subs, especially below 40 Hz, and still give significant sub-34 Hz performance. 

 

Perhaps a little more thinking on this subject is prudent before cutting holes and attaching ports.

 

Chris

 

A couple thoughts here:

 

1) It's possible to model a vented rear chamber in hornresp to see how the front wave and rear wave couple. The loading in front of the driver definitely affects the diaphragm excursion, which in turn affects how energized the port gets. I've had decent correlation between the model and real world when playing with this in the past, but that might be a bit harder here.

2) Keeping group delay under 1.5 cycles (or better yet 1 cycle) is one of my design targets - but it needs to be looked at after the total system EQ. Even with the model you're showing, a peak like that is going to need some EQ, which in turn will reduce the phase rotation.

3) When comparing phase response against a sealed system, it's important to have things EQ'd to the same final shape within the passband.

4) In my experience the Q of the port for a target bandwidth is mostly dominated by the driver. The rear volume mostly affects how much gain you get at the port frequency, which I suppose is technically affecting the Q too, but that's secondary to the driver selection. If instead you go to a passive radiator, then you can modify the Q separately from the active driver. I'm not recommending going to a passive or to change the active driver, just trying to share some info about modifying the Q. Actually, now that I'm thinking about it....I remember reading a long time ago that you can have a massively flared port and that can modify the Q. I've also read about having multiple different sized ports in a design too.

5) I actually did a full on port design the other day when working through your numbers, but didn't publish it because I wasn't sure about the accuracy of the model. I was able to find a port alignment where the frequency response gain matched the diaphragm excursion - which means if it were EQ'd to be the same response as the sealed alignment, then the cone excursion would be roughly one fourth that of the sealed alignment. Then where the system is rolling off, EQ could be applied to increase the low frequency extension while also keeping the cone excursion less than 1/4 the sealed alignment. Then at way lower frequencies a high pass filter would be required to keep the cone excursion down to the same level as the sealed enclosure. Basically you end up with the same final frequency response with a quarter of the cone excursion. Yes, there would be some extra group delay in the final alignment, but it wouldn't be huge.

[DrWho]

Note that I personally do not wish to add the reflex bass distortion (phase/group delay and impulse distortion) of reflex ports to the current design, instead preferring the tighter bass of a closed box system over a vented one, despite having its harmonic distortion rise due to EQing the bass performance to be flat again. 

 

Have you seen this article?

http://www.data-bass.com/data?page=content&id=82

[DrWho]

Chris, have you measured the response of the TAD driver on your multiple entry horn versus the stock K402? Are you seeing any tradeoffs to the compression driver performance?

[Chief bonehead]

 

just a small detail....the 402 is not a conical expansion...

 

The way that it is constructed--I agree with you, but for its effective properties related to flare rate expansion vs. off-axis port location (the low-pass acoustic filter), it behaves just like the horns that Danley uses. 

 

So for the simplicity of calling it something other than "the straight-sided walls of the modified tractrix horn", I'll just refer to that area as a "conical", even though it doesn't expand on an x^2 profile.

 

Chris

 

to make it even more simple, just call it a megaphone.......

[jason str]

 

Note that I personally do not wish to add the reflex bass distortion (phase/group delay and impulse distortion) of reflex ports to the current design, instead preferring the tighter bass of a closed box system over a vented one, despite having its harmonic distortion rise due to EQing the bass performance to be flat again. 

 

Have you seen this article?

http://www.data-bass.com/data?page=content&id=82

 

 

 

After seeing the drivers they used in the DTS-10 tests Data-Bass lost all credibility in their testing methods.

 

You cannot just throw any driver in any style cabinet and expect it to work properly.

 

Two thumbs down. :emotion-45: