Jump to content

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


Chris A

Recommended Posts

By the way, some of the discussion in another thread (Heresy IV measurements at AudioScienceReview.com) on the value of a bass reflex design vs. acoustic suspension (a closed box design using the air inside the box as part of the restoring mechanism of the woofer's surround and spider) in this age of inexpensive and powerful DSP crossovers and lots of good, clean power is one that I think needs some more discussion.  Most people that buy bass reflex loudspeakers are not thinking about the loss of deep bass extension that one gets using bass reflex ports, i.e.,

 

WinISD_comparison.png

 

Note the difference in woofer output below ~17 Hz (in this particular example, but usually the break frequency is at significantly higher frequency) using acoustic suspension instead of reflex ports.  This is the source of the extended bass response that one can tap into using DSP and a little more power, and actually can be used to avoid having to use subwoofers, which often have terrible performance relative to loudspeaker woofers.  The trade-off of course is increased AM distortion sidebands for the acoustic suspension loudspeaker and not having as much "raw SPL" (and I mean "raw" here, not refined) as typical subwoofers.  As long as this AM distortion is still fairly low, the trade off to not use a subwoofer may be a good one.   That's a reasonable trade, but you can't have bass reflex ports (or open baffle dipoles) if you want that extended deep bass response.  I've done this with my K-40-MEH prototype and I enjoy significant infrasonic bass down to 18 Hz.  I highly recommend this approach, and it is something that K-402-MEH builders can try for themselves.

 

For people that say they don't use their loudspeakers for LFE and movie reproduction, and that say that they don't need that deep bass extension, I find when they are in the room with loudspeaker systems having really deep bass response (below 30 Hz), they are startled by the increase in subjective listening involvement with better recordings that do not attenuate this portion of the spectrum--playing music only.  Many or most recordings made within the past 29 years still retain a significant amount of deep bass--since 1991 when multiband compressors started to be widely used for mastering popular music recordings (but generally that compression was not used on most classical or jazz recordings until the last 10-15 years or so). 

 

I've thought about plugging the ports in my Cornwalls (1979 versions) and using DSP to flatten and extend the deep bass response to hear the differences and to get perhaps another half an octave of deep bass extension. 

 

Chris

  • Like 1
Link to comment
Share on other sites

16 minutes ago, Chris A said:

Note the difference in woofer output below ~17 Hz (in this particular example, but usually the break frequency is at significantly higher frequency) using acoustic suspension instead of reflex ports.  This is the source of the extended bass response that one can tap into using DSP and a little more power, and actually can be used to avoid having to use subwoofers, which often have terrible performance relative to loudspeaker woofers.  The trade-off of course is increased AM distortion sidebands for the acoustic suspension loudspeaker and not having as much "raw SPL" (and I mean "raw" here, not refined) as typical subwoofers.  As long as this AM distortion is still fairly low, the trade off to not use a subwoofer may be a good one.   That's a reasonable trade, but you can't have bass reflex ports (or open baffle dipoles) if you want that extended deep bass response. 

 

It's a delicate tradeoff. Putting transient response aside for the moment, the price that one pays when equalizing a closed-box woofer is increased cone excursion. The excursion always increases as frequency is reduced. The rate at which it increases is predictable and constant until physical limits are reached.

 

By comparison, the excursion of a bass-reflex woofer actually decreases as frequency is reduced, to near-zero at the box tuning frequency. But as frequency is further reduced below the box tuning frequency, the excursion increases rapidly because the driver and the port begin to radiate out-of-phase. At a point not far below the box tuning frequency, the excursion of the bass-reflex woofer becomes larger -- much larger -- than that of an equivalent closed-box woofer. That is why trying to EQ extra output from a bass-reflex woofer, below the box tuning frequency, is a waste of effort.

Link to comment
Share on other sites

Yes--on the bass reflex explanation.

 

If you've got two 15" woofers that the K-402-MEH has, you have a bit more "trade off" region than is the case normally with a closed box "acoustic suspension" bass bin (like my K-402-MEH is).  That's one reason why I decided two 15" woofers was the better option. 

 

The Danley SH-96 uses four 15" woofers in its design, which I decided was a more than a little excessive, especially since my loudspeakers are in room boundary coupling below ~50 Hz. 

 

Most people (including some otherwise smart guys in the horn world) typically forget about that boundary loading, and in fact I see many trying to move their loudspeakers away from the room's walls.  It's a much better trade to use the boundary loading and to cover the little area that you need to cover for early midrange reflections due to the closer walls.  That's a much better trade acoustically.

 

Chris

  • Like 1
Link to comment
Share on other sites

On 11/26/2020 at 12:19 PM, Chris A said:

I've thought about plugging the ports in my Cornwalls (1979 versions) and using DSP to flatten and extend the deep bass response to hear the differences and to get perhaps another half an octave of deep bass extension. 

 

Chris

I was headed down this road myself at one time. If you do, you might want to use the woofers I used for my abandoned Super Cornwall project since it has about 2-3 times the Xmax of the K33, which is what you will be stressing with a closed box, along with higher IM distortion, but you know this already...............Better yet,  you can go to an 18" woofer since it will still be reasonable in the midrange at 500 Hz. where the mid horn starts to take over. But that would require making sawdust.

Link to comment
Share on other sites

How does the compression ratio of the ports affect the sound...is it efficiency? What happens if you have less compression ratio vs higher compression ratio vs 10:1?

Sent from my SM-T830 using Tapatalk

Link to comment
Share on other sites

12 hours ago, NBPK402 said:

How does the compression ratio of the ports affect the sound...is it efficiency? What happens if you have less compression ratio vs higher compression ratio vs 10:1?

This is not an easy answer for me to give you first-hand experience in context of MEHs.  I would have to build a succession of MEHs having varying compression ratios (which I have historically defined as the ratio of the driver's diameter to the port's diameter--which is not area).  This would costly, time-consuming, and difficult to sit and listen to the trade-off in subjective listening performance without introducing other unwanted factors into the listening trials. 

 

Changing that ratio (i.e., making the off-axis ports either smaller or larger) has competing tradeoffs:

  1. On one hand, it increases the efficiency of the woofers as the woofer off-axis ports get larger.
  2. On on the other hand, the off-axis polar coverage of the horn suffers in the region around the first bounce frequency of the off-axis ports relative to the apex compression driver port (the "first notch frequency"). There is a dropout in off-axis horizontal polar coverage around 400-500 Hz as the off-axis port area increases (as a function of the total sidewall area on the horn where the off-axis ports are located).  This turns out to be the most limiting factor in my experience.

 

So the alternative is to lengthen the ports along the crease in the horn to increase the port area, but this has limitations, too.  If the port length becomes significant relative to the highest wavelengths of sound that are produced by the woofers (i.e., 1/4 wavelength at 500 Hz is 6.8 inches), diffraction effects begin to show up at the highest woofer frequencies.  So if the port length gets longer than 6.8 inches, you start to have issues with self-diffraction of the sound waves emitted at the highest woofer frequencies. (This is the same problem that driver phase plugs are there to fix.) It also begins to lower and deepen the first notch frequency of the woofers relative to the horn apex or throat.

 

pst003.jpg

The long ports on the larger sidewalls of this MEH are the off-axis woofer ports that have been elongated in order to increase their area for one-woofer/one-port operation

 

In summary, the answer is a bit complicated, and there are some other woofer interactions (loading the woofers at low frequencies vs. the exact horn shape and off-axis port locations) that turn out to be trade-offs that I haven't talked about yet.  But in general, I believe that the relative compression ratios and T/S parameters of the woofers, the port placements, and their length/width ratio is proven by looking at what Danley has done.  Unless you've got a lot of time and a fairly significant pile of money to build several horns of differing port areas/geometries, and use several different woofers having differing T/S parameters, the answer to your seemingly simple question shall largely remain obscure to generalizations on their effects. 

 

You could try asking Tom Danley himself (whom I'm reasonably sure has done some of these trades), but I'm not sure that he would be willing to divulge the generalizations of what his company has found..."the hard way". 

 

Chris

  • Like 2
Link to comment
Share on other sites

This is not an easy answer for me to give you first-hand experience in context of MEHs.  I would have to build a succession of MEHs having varying compression ratios (which I have historically defined as the ratio of the driver's diameter to the port's diameter--which is not area).  This would costly, time-consuming, and difficult to sit and listen to the trade-off in subjective listening performance without introducing other unwanted factors into the listening trials. 
 
Changing that ratio (i.e., making the off-axis ports either smaller or larger) has competing tradeoffs:
  1. On one hand, it increases the efficiency of the woofers as the woofer off-axis ports get larger.
  2. On on the other hand, the off-axis polar coverage of the horn suffers in the region around the first bounce frequency of the off-axis ports relative to the apex compression driver port (the "first notch frequency"). There is a dropout in off-axis horizontal polar coverage around 400-500 Hz as the off-axis port area increases (as a function of the total sidewall area on the horn where the off-axis ports are located).  This turns out to be the most limiting factor in my experience.
 
So the alternative is to lengthen the ports along the crease in the horn to increase the port area, but this has limitations, too.  If the port length becomes significant relative to the highest wavelengths of sound that are produced by the woofers (i.e., 1/4 wavelength at 500 Hz is 6.8 inches), diffraction effects begin to show up at the highest woofer frequencies.  So if the port length gets longer than 6.8 inches, you start to have issues with self-diffraction of the sound waves emitted at the highest woofer frequencies. (This is the same problem that driver phase plugs are there to fix.) It also begins to lower and deepen the first notch frequency of the woofers relative to the horn apex or throat.
 
post-26262-0-30020000-1452805715.jpg
The long ports on the larger sidewalls of this MEH are the off-axis woofer ports that have been elongated in order to increase their area for one-woofer/one-port operation
 
In summary, the answer is a bit complicated, and there are some other woofer interactions (loading the woofers at low frequencies vs. the exact horn shape and off-axis port locations) that turn out to be trade-offs that I haven't talked about yet.  But in general, I believe that the relative compression ratios and T/S parameters of the woofers, the port placements, and their length/width ratio is proven by looking at what Danley has done.  Unless you've got a lot of time and a fairly significant pile of money to build several horns of differing port areas/geometries, and use several different woofers having differing T/S parameters, the answer to your seemingly simple question shall largely remain obscure to generalizations on their effects. 
 
You could try asking Tom Danley himself (whom I'm reasonably sure has done some of these trades), but I'm not sure that he would be willing to divulge the generalizations of what his company has found..."the hard way". 
 
Chris
Thanks Chris. I was wondering because I am putting together a wood/MDF horn that is modeled off the short side of the k402 and will be square. I intend for this to be a surround version that will be smaller, and was not sure if I should go with 2 15" woofers or 2 or 4 smaller woofers with the cabinet being almost half the size. I did not want to have to change the ports to change to a different size woofer, after the horn was setup for 15" woofers. Any recommendations on woofers for a 25" x 25" horn?

Sent from my SM-T830 using Tapatalk

  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...

I have the center channel completed, and installed on its new angled stand, and my left and right channes are setup now in their positions.@Chris A do you want to work on the tune now or do you want to wait until I get the wood versions built?

Sent from my SM-T830 using Tapatalk

Link to comment
Share on other sites

We can start on this.  Do you need a starting presets file for your DSP crossover.  Send me a PM or email.

 

Chris

Link to comment
Share on other sites

  • 2 weeks later...

If you have not ever heard a pair of [mention=62555]Chris[/mention] A MEHs... I seriously suggest you do! I am very impressed with how mine turned out, and there is no way I could buy anything that would sound as good as these are for anywhere near what I have invested. Look for a used pair of k402s, and save some money too. I went the budget route and got EV DH1As, and then a new pair of Emminence kappa 15Cs...DH1As go for about $100 each used. They are not that hard to build either.

 

Sent from my SM-T830 using Tapatalk

 

 

 

 

 

 

  • Like 2
Link to comment
Share on other sites

  • 2 weeks later...

I placed the following in another thread--quite off topic for that thread.  The discussion got around to digital vs. physical alignment of the drivers.  In MEHs each pair of "ways" (higher and lower frequency drivers) are, by definition, within 90 degrees of phase (1/4 wavelength) of each other, and since there is a single horn aperture (mouth), phase/time alignment is assured...if the users don't try to insert crossover filters, especially ones with higher order filters (higher than first order):

____________________________________________________________________________________________

MEHs apparently solve the problems:

 

Ron Ellis MEH center channel SPL and Phase Response.jpg

 

The group delay effects around 1 kHz are compression driver effects alone.  I'd call that success.  It sounds really, really good, too.

 

The loudspeaker actually needs no crossover filters. There are no crossover filters or FIR filters being used above--only PEQs to flatten response and to notch out the unwanted SPL response outside of the desired pass bands.  It's also very easy to dial-in, since there is also no delay applied to any channel. 

 

1785130915_RonEllisLCRPhaseResponse.jpg.cb06577f7a11aeb6b413a69fcb174a05.jpg

 

Chris

  • Like 2
Link to comment
Share on other sites

6 minutes ago, Chris A said:

... phase/time alignment is assured...if the users don't try to insert crossover filters, especially ones with higher order filters (higher than first order):

 

I don't understand your aversion to higher-order crossover filters, Chris. There are a number of crossover topologies that either sum-to-allpass (e.g. Linkwitz-Riley) or sum-to-delay (any of the perfect-reconstruction digital structures). Personally I don't care for the sound of sum-to-allpass networks, but sum-to-delay is just fine.

Link to comment
Share on other sites

6 minutes ago, Chris A said:

Have you heard an MEH?

 

No. Why is that important to the point? If the drivers are in-phase at all frequencies in a traditional enclosure, then they're in-phase at all frequencies in a MEH.

Link to comment
Share on other sites

MEHs don't behave like multi-horn loudspeakers...in several ways.  The reason why I asked is it's difficult to express how they sound subjectively (especially when dialed in like the SH-50), and how they behave to changes in their settings.  The "rules" of multi-aperture loudspeakers (multiple horns, multiple direct radiating drivers) don't really apply to MEHs. 

 

The most surprising thing that I noticed is that there is summation within the aperture that you don't get with other multiple-aperture loudspeakers.  The way that you think about using higher order crossover filters to minimize the interference bands (which is the sole reason that I see to try to use higher order filters)--that thinking just doesn't apply to MEHs.

 

I got a clue from reading Tom D.'s comments very closely on another forum, and reading about their full transfer function response--which he was touting as a unique capability not found in other loudspeakers (good thing the controlling patent ran out, because he'd be sitting on top of a gold mine if it didn't).  He was showing linear phase using IIR filtering alone (passive), and was saying that brings a new capability to sound reproduction that one could only partially achieve with "full range" drivers, except with none of the disadvantages. 

 

That's probably enough to lay on you for the moment.  If you think about the implications of what I just said above and why it is significant (assuming that you can realize that you've probably never heard what he was talking about before), I think it will begin to answer your posed question (however far away from it that you might think it is--but it isn't), and many others that you haven't asked yet.

 

Here's another answer:  https://community.klipsch.com/index.php?/topic/182419-subconscious-auditory-effects-of-quasi-linear-phase-loudspeakers/page/3/&tab=comments#comment-2379562

 

Chris

  • Like 1
Link to comment
Share on other sites

18 minutes ago, Chris A said:

The way that you think about using higher order crossover filters to minimize the interference bands (which is the sole reason that I see to try to use higher order filters)--that thinking just doesn't apply to MEHs.

 

The main reason in my mind for using higher-order filters in crossovers is to minimize out-of-band excursion at low frequencies, and to minimize break-up at high frequencies. I can't think of any reason that those problems, and their solutions, wouldn't also apply to MEHs.

 

18 minutes ago, Chris A said:

I got a clue from reading Tom D.'s comments very closely on another forum, and reading about their full transfer function response--which he was touting as a unique capability not found in other loudspeakers (good thing the controlling patent ran out, because he'd be sitting on top of a gold mine if it didn't).  He was showing linear phase using IIR filtering alone (passive), and was saying that brings a new capability to sound reproduction that one could only partially achieve with "full range" drivers, except with none of the disadvantages. 

 

I sent you my AES paper showing how to create linear phase system response using sum-to-delay IIR filters. If Danley did it passively, then his solution has to take into account delay between drivers due to physical positioning in space. Pure delay is very difficult to achieve with passive electrical devices.

 

18 minutes ago, Chris A said:

 

If you think about the implications of what I just said above and why it is significant (assuming that you can realize that you've probably never heard what he was talking about before), I think it will begin to answer your posed question (however far away from it that you might think it is--but it isn't), and many others that you haven't asked yet.

 

Forgive me, but I have no idea what that sentence means.

Link to comment
Share on other sites

30 minutes ago, Edgar said:

The main reason in my mind for using higher-order filters in crossovers is to minimize out-of-band excursion at low frequencies, and to minimize break-up at high frequencies. I can't think of any reason that those problems, and their solutions, wouldn't also apply to MEHs.

How many horn-loaded loudspeakers have you dialed in using DSP?  How loud do you listen to them?  Your characterization of problems--don't exist in my experience, and the actual techniques that I'm applying to "zero phase growth crossovers".  That's about the only way that I can respond to your statement.

 

Additionally, my take on higher order crossovers initially came from Roy one weekend in Hope...if that means anything. I can remember the conversation (Q&A with a few guys in the room). 

 

30 minutes ago, Edgar said:

I sent you my AES paper showing how to create linear phase system response using sum-to-delay IIR filters. If Danley did it passively, then his solution has to take into account delay between drivers due to physical positioning in space. Pure delay is very difficult to achieve with passive electrical devices.

I think you are thinking like a DSP engineer here, and not so much a fully horn-loaded loudspeaker engineer.  It's the resulting sound that I'm after, as well as a few key performance measures that I can get from REW and DATS (and I'd take others if I could get them), which, taken together, are important...not how I did the DSP filtering. 

 

I believe that I just answered your original question above...when combined with the statement: "there is summation within the aperture that you don't get with multiple aperture loudspeakers". 

 

Of course there is physical time/phase alignment within the aperture (within 90 degrees)--that's the whole point of the Danley-style MEH, and is in US6411718.

 

Chris

Link to comment
Share on other sites

3 minutes ago, Chris A said:

How many horn-loaded loudspeakers have you dialed in using DSP? 

 

Irrelevant.

 

Quote

How loud do you listen to them?  Your characterization of problems--don't exist in my experience,

 

Perhaps not in your living room. My experience comes from a sound reinforcement context, where power levels are high enough that excurson and breakup are very real problems.

 

Quote

Additionally, my take on higher order crossovers initially came from Roy one weekend in Hope...if that means anything.

 

Without context, I have no way of knowing what it means.

 

Quote

I think you are thinking like a DSP engineer here, and not so much a fully horn-loaded loudspeaker engineer.

 

Why are you making it personal? I'm thinking like an engineer, and an engineer knows that the Laws of Physics aren't repealed just because you happen to be operating inside a horn. There are near-field effects and there are far-field effects, but since we generally listen in the far field, that would seem to be the place that we should focus our attention.

 

Quote

"there is summation within the aperture that you don't get with multiple aperture loudspeakers"

 

If superposition does not apply inside the aperture like it does outside the aperture (or vice-versa), then the system is nonlinear. Nonlinearity is the last thing we want in a loudspeaker.

Link to comment
Share on other sites

25 minutes ago, Edgar said:

Why are you making it personal?

I'm not.  Sorry if it came out like that...that wasn't my intent.

 

25 minutes ago, Edgar said:

Without context, I have no way of knowing what it means.

"What is the purpose of using higher order crossovers...?" Was the question posed, which Roy answered, the first part of which includes the answer I gave.  The second part was the answer you gave--which has no value in my world, since commercial cinema/PA isn't of interest to me in what I do.

 

25 minutes ago, Edgar said:

If superposition does not apply inside the aperture like it does outside the aperture (or vice-versa), then the system is nonlinear. Nonlinearity is the last thing we want in a loudspeaker.

It appears you need to listen to and play with an MEH yourself--because I believe we're not actually communicating here. I'm not sure how to achieve that communication at this point.

 

I do recommend a build (by you) and playing with all the settings to see what occurs--I believe that it will really surprise you what happens (and doesn't happen, too). 

 

I have to confess that I am not revealing all the information that I gleaned from looking at the SH-50 very closely, and what Danley did.  Believe it or not, I do think it's wise not to disclose all that I found, because Tom worked hard, and I don't think it's fair to broadcast the "inner secrets" of an SH-50 design to the world.  That wasn't the reason why I looked into the SH-50 to begin with.  I don't wish to compete with DSL, and I think they are more than worthy of reaping the benefits of Tom's efforts.  Tom is an innovator, and I'd like to see both Klipsch and DSL be successful--because everyone will be the beneficiaries of the results--their products.  But if you understood what's in the design of an SH-50, I believe that my explanations would make a lot more sense.  I choose not to be too verbose, however.  Anyone else can buy an SH-50 and do what I did--so I'm not protecting a lot that can't be "borrowed" (plagiarized).  I choose to go my own way, however.

 

I looked at the SH-50 more closely to understand what Tom was saying (and believe me, I understand a LOT more than I did initially).  I was trying to improve upon the design of the K-402-MEH design so that it could be an improved five-channel MEH surround system in my listening room--that I can't buy, so I have to make them.  I shared what I found in my design with the world, because I'm not making a living off of the design. 

 

Chris

  • Like 2
Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...