Jump to content

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


Chris A

Recommended Posts

By the way, the reason why I made the comments in the last post is to communicate to others (perhaps beside yourself) that are either currently working on or planning to work on K-402-MEHs.  It just isn't that complicated.  I don't want others to think that it is complicated.  It's not.

 

2 hours ago, rickmcinnis said:

...I wondered if one could take some of the load off of the woofers in the area where they are least effective...

Here is a raw output from the prototype K-402-MEH that I've used as my center channel loudspeaker since January 2016.  The fall-off in output above 450 Hz is due to the reflection wave from the throat of the horn (the returning acoustic energy that is emitted from the off-axis woofer ports that cancels the acoustic energy being emitted toward the mouth of the horn).  I'm struggling to understand what you mean by the above quoted text:

 

New Center - Woofers Only RAW No EQ.png

 

Note that there are two 15" woofers in the cabinet, and these were sized to provide plenty of capability to the bass response--even down to very low frequencies.  You can read the -3 dB point of the following SPL response plot (16.5 Hz) of the K-402-MEH in the center position (EQed flat), elevated above floor level by over a metre, back touching the front wall...as well as the phase response:

 

K-402-MEHwithDanley-StyleCrossoverFilters.jpg

 

Chris

  • Like 1
Link to comment
Share on other sites

3 hours ago, rickmcinnis said:

One has to find out for themselves.

Can I make a suggestion:  perhaps you can find out and then report the results.  You can include the response and harmonic distortion plots, too, and perhaps a picture of the amplifier innards with external capacitors.  I believe that your comments before you actually try it with your (admittedly esoteric) amplifier configuration are a bit like "making the sausage" instead of "tasting sausage".  If I'm not conversant with what you're currently doing (and why), then I'd also fear that most others are not following along, too.

 

It helps to keep things as simple as possible at first, then layer on the more complex cases later.  In that type of course of action, I'd also recommend also trying a more garden variety SS amplifier first that does not require such provisions (and issues).  KISS works really well, I've found.  After establishing a working baseline with run-of-the-mill configurations, then reporting on more esoteric component use (i.e., your First Watt SIT-1 amplifier that requires a very large capacitor on its output channels) afterwards.

 

3 hours ago, rickmcinnis said:

Still would like to know if you are using the first order slopes within the XILICA or only PEQs.

I've used it both ways.  Surprisingly, it doesn't make any difference in the output SPL/phase response, harmonic distortion, as well as the off-axis polars, etc. with the MEH design (which is not true with a multiple-horn loudspeaker like the Jubilees on either side).  The difference is in the acoustic summing that takes place within the horn/waveguide itself within the MEH.

 

3 hours ago, rickmcinnis said:

If I set the crossover higher (obviously) I could get a less extreme acoustic slope order and that is something I will try before I assemble the MEH.

Trades on the crossover point center frequency are okay.  I think you have yet to acknowledge that there isn't any audible difference if using Danley-style crossovers in an MEH.  There is no lobing, so increasing the  overlap of drivers in the "interference band" has no effect in the output sound quality or measured distortion.

 

3 hours ago, rickmcinnis said:

...There are lots of pieces!...

I find that there are very few pieces, actually, much fewer than trying to use multiple horns and passive crossovers.

 

3 hours ago, rickmcinnis said:

Since my temporary setup with the Edgar Seismics as lower midranges/mid bass are much happier with the 300 hz crossover it was good that the CELESTIONs allowed this.

I can see that--but I will recommend listening to those Celestion compression drivers at that lower crossover point.  My experience has not been good trying to use metal or ring radiator diaphragm (poly material) compression drivers of any make or type below ~400-450 Hz.  I find that it's easy to hear the difference during the upsweep measurements using REW.  The human hearing system can pick up very small differences in distortion levels under upsweep conditions.

 

3 hours ago, rickmcinnis said:

Getting those 18 inches JBLs to go higher up in frequency requires lots of boost.

Using 18 drivers above ~300 Hz will probably require phase plugs inside the horn-loaded cabinets in order to avoid driver-based cancellations due to the diameter of the cones themselves.

 

3 hours ago, rickmcinnis said:

I know you posted a raw response curve but i cannot find it.  Would you mind posting it again along with your current XILICA settings for the CRITES?  I would be very grateful, AGAIN!

I've posted the raw response of the K-402-MEH woofers, above. 

 

I'll have to look around for the Xilica settings that I used.  I experienced a hard drive failure this spring that wiped out my measurements from the past three years (which I apparently failed to back up to my backup external hard drive.  However, I did manage to back up all the music tracks that I demastered--yea! 

 

I've also got a USB port issue with the Xilica (yet again). I have to again replace the USB connector in the Xilica. (Good thing I bought a bag of those--I keep inadvertently running into the cable when taking measurements and damaging the cable and USB connector in the Xilica...bummer.)  I have to take the time to put the system offline to do that--and I find that I'm spending all my available time listening to the system since I made the latest crossover changes in April 2019.  I just finished demastering the first 6 The Alan Parsons Project albums over the past couple of days and the sound is too addictive to stop listening to them presently.  Note that I've found that those bass bin settings are so sensitive to room placement, etc. that they're not transferable to other installations, etc., so I don't bother to post them for that reason.

 

Chris

  • Like 1
Link to comment
Share on other sites

@rickmcinnis Did you make your ports 10 to 1 compression...if so I do not see how you could change the size now without patching. On mine I used a 45 degree router bit to chamfer the woofer side of the woofer mounting plate. We chamfered down to about 1/8".

 

Sent from my SM-T830 using Tapatalk

 

 

 

 

Link to comment
Share on other sites

I can post these recently demastered albums (i.e., I redid my demastering from 5 years ago when I first was learning to demaster--in "whole cloth" fashion) on the Google drive account.  They're really nice to listen to now-especially since I got the system dialed in so much better in terms of phase response.

 

I don't have any of these albums on multichannel.  I checked the DR Database on the Eye in the Sky album, and it was more than 4 dB lower in dynamic range than the stereo versions that I've acquired over time (i.e.,  the first CD versions from the 1980s have significantly higher dynamic range).  So I'll probably pass on the multichannel version since 4 dB of dynamic range is a lot of difference.  I know that the Eye in the Sky multichannel disc won some sort of award last year.

 

Chris

  • Like 1
Link to comment
Share on other sites

Pm sent... I have Ammonia Ave, and Eye in the sky in 5.1. One of my Elvis Costello albums shocked me at the dynamics one night. I was just using my phone to measure SPL.

Sent from my SM-T830 using Tapatalk

Link to comment
Share on other sites

On 8/21/2020 at 4:05 PM, NBPK402 said:

One of my Elvis Costello albums shocked me at the dynamics one night. I was just using my phone to measure SPL.

 

Which one?  I have listened to the 24/96 versions and they are all much better than the standard ones.  Speaking of the "classic era" from MY AIM IS TRUE to TRUST.  After that it got to the point the records were very uneven - a couple of great songs and the remainder McManus's throwaways.

 

My favorite EC is GET HAPPY which is probably the worst recording but it sounds very good in 24/96.  

 

As far as dynamics go - NEW LACE SLEEVES on TRUST is pretty impressive.

 

To add: NORTH is a very well recorded record and I have a soft spot for PAINTED FROM MEMORY.


 

 

Edited by rickmcinnis
addition
  • Like 1
Link to comment
Share on other sites

On 8/14/2020 at 8:29 PM, Chris A said:

Well, the answer may surprise you...

 

I run the natural response of the woofers and then compression driver, after flattening the SPL response of both sections individually.  Then I look at the natural crossing point of the two.  In the case of the BMS 4592ND, I believe that was 550-600 Hz. 

 

I use first-order crossover filters (effectively) to take out the 90 degree lead of the woofers to the compression driver.  Voila! Time alignment.  When I did this (and stopped trying to tell the compression driver and the woofers where to cross using the crossover and PEQs), all the performance measurements got better and it sounded better.  I simply trim off the response of drivers that is not needed using attenuating PEQs in order to keep the crossover interference band SPL response flat. 

 

Chris

'THE FORMULA' for JUBILEE tuning's : (May I add, after TUNING the room first !)

 

…run the natural response of the woofers and then compression driver, after flattening the SPL response of both sections individually .

 

...take a look at the natural crossing point of the two. 

 

use first-order crossover filters (effectively) to take out the 90 degree lead of the woofers to the compression driver.

      

                                         Voila! = Time alignment. 

 

stop trying to tell the compression driver and the woofers where to cross using the crossover and PEQs.

  

trim off the response of drivers that is not needed using attenuating PEQs in order to keep the crossover interference band SPL response flat. 

 

Thank You, brotha CHRIS !

  • Like 1
Link to comment
Share on other sites

20 hours ago, rickmcinnis said:

I know you are correct that I am making this to be far more complicated than it should but then I have two horns to cut into and I think this is what makes me overly cautious.  I know it would not be that hard to fix an error but I think I hate the thought of having to take the things apart. And times two

Well, this thread serves two purposes (at least): the first is to document the experiences of building and dialing in/measuring the MEHs, and the second is to convey to others considering building this amazing sounding loudspeaker--that it isn't really as difficult to build and dial in (using a DSP crossover) as a typical box-type loudspeaker. 

 

So my comments above on the difficulties encountered were to clarify that those that are considering their own MEHs that it is really simple--assuming that you own a $95 calibration microphone and a computer.  I dare say that if a pair of these were put into the same room as other loudspeakers (and I'm not going to "go there" by identifying any others), that the K-402-MEHs would, well, let's just say..."do really well". 

 

______________________________________________________

 

I would hope that Klipsch (consumer division) itself would try out this configuration and listen in a home hi-fi--sized listening room. I think the configuration would blow away any competitors in the market.  And at the prices that Klipsch pays its supplier of horns (probably significantly less than $290 per horn), this is a no-brainer for those looking for Jubilee sound quality (at least) in a package that is 1/3 the volume and 1/4 the cost (and price to the consumer).  The price of good DSP crossovers that can do the job is now at $200 (USD) for a pair of MEHs, and the net performance of them after dialing in would eclipse the in-room performance of anything that Klipsch currently produces exclusively for the home hi-fi market (without exception).

 

The K-402-MEH works well elevated above floor in half space, i.e., touching the wall behind it.  It also picks up 15-18 dB of low bass gain if placed in a room corner...like a Jubilee or a Khorn.  So it has the flexibility of placement that the other two loudspeakers really don't have (unless large false corners or bass bin wings are added and taking up a great deal of space).

 

I've found that for my center channel needs, the K-402-MEH also serves as its own subwoofer due to the two 15" woofers enclosed, and has consistent polars all the way up and down the audible band, and flat phase response (within ±90 degrees) from 100 Hz to 20 kHz.  I've posted those in-room REW measurements on this forum. This is a good deal.

 

Chris

  • Like 1
Link to comment
Share on other sites

 

 

 

 

   

One last thing I have assumed you removed the extra thick gasket from the CRITES woofer.  I have removed them completely to use Mortite.  I love MORTITE. Get a little less volume - actually surprisingly less.

 

[

 

Do you get any slapping with the woofers being closer to the mounting pad? I was thinking of actually adding a little space qith my Kappa 15Cs. How much clearance would the Mortite give me?

 

Sent from my SM-T830 using Tapatalk

 

 

 

 

 

Link to comment
Share on other sites

2 hours ago, rickmcinnis said:

Meant to have given my condolences for your hard drive failure.  That is a great loss to you and all of us here.

Well, it's a loss, but probably not a great one.  I lost my measurements from April 2017 to April of this year.  That includes the detailed measurements on my Danley SH-50 and the ESS AMT-1 drivers.  But I still own all the loudspeakers, etc., so once I repair my Xilica USB port connector, I'll be back in business again to capture them again.  That HD drive is dead, however.  I'm thinking the value of the measurements lost aren't worth the cost of forensically retrieving the data via HD recovery service (which is what it would require since the lead-in track is trashed).

 

I need to try to double stack my AMT-1s on the surround Belles (since I'm not using my Khorn bass bins/AMT-1 in the much smaller bedroom/office (13.5' x 12') that I use for storage.  I think that angling the second AMT-1 upwards (i.e., leaning it backwards) to approximate an inverted J-shaped line array will give better coverage plus a little shading of the vertical coverage to the AMT-1s.  I'm thinking that I'll need another miniDSP 2x4 HD in order to control the shading and delay properly.  I'll advise on that project when I get some round tuits.

 

2 hours ago, rickmcinnis said:

It would be great if KLIPSCH would offer the MEH but then I could not afford it and I would probably have plenty of quibbles with how they handled certain aspects.

Can't talk about this subject here--it's bad manners, I believe. I'm working on a solution.

 

2 hours ago, rickmcinnis said:

All of the thinking (?) that goes into these projects is good for us in many ways.

You're right.

 

2 hours ago, rickmcinnis said:

Only reason my curiosity is aroused is whether there could be a good reason to bring the ports a little closer to the mouth.

2 hours ago, rickmcinnis said:

Sine my ports will be much closer to the suspension than in your situation I am thinking longer narrower ports than yours are called for.  I wonder if the longer port aloows for more fudging since they will then be in a wider range of the frequencies of where the horn is working.?  Uh, oh, there I go again!

You can do this.  Just note the effects of this on woofer off-axis loading ("cone tilt").  The off-axis polars will be affected at a higher frequency. Long, thin off-axis ports will also elongate the off-axis polar coverage effects in frequency.  (Those long ports on the Danley Synergies that you've talked about are for those using four woofers per loudspeaker rather than two woofers like the present K-402-MEh configuration.  They did that because each woofer only has one port in the SH-96, etc.)  Also, using four woofers is probably a fair bit of overkill for home hi-fi needs.

 

2 hours ago, rickmcinnis said:

The way I see it  - getting the ports closer to the throat is good for the woofer's high frequency extension but that getting the ports further away from throat allows less interference between woofers and CD.

Yes. The off-axis ports need to be in the right place, and be the right size and shape.  I found that Hornresp provided enough information in terms of woofer on-axis SPL response to place them in the right place.  Experimentally, it worked out.

 

If you move the woofers too close to the throat, you'll have to make their port sizes too small, and you'll seriously affect the efficiency of them if those ports are too undersized.

 

2 hours ago, rickmcinnis said:

Without PEQ we would be out of luck!  

Based on my (now lost) SH-50 measurements using REW, the K-402-MEH woofers have a much flatter response than the SH-50, and the midranges in the SH-50 are almost not necessary--probably they're only there because Danley needed them to hold the phase growth down using passive crossovers.  If he was using DSP crossovers only and was using the loudspeakers at home hi-fi levels (rather than PA levels), I think he would have avoided the midranges altogether.  The midrange drivers only cover just over an octave of bandwidth in the SH-50--hardly worth the effort to put them in.  Danley pays a great deal of attention to phase response.  So do I, now, after hearing the difference.  This is the difference between the Unity horns from Sound Physics Labs, and the Synergy horns from DSL--flat phase response.  And the listening difference is startling.

 

2 hours ago, rickmcinnis said:

One last thing I have assumed you removed the extra thick gasket from the CRITES woofer.

No, I didn't.  Again, I think you're obsessing over "mouse meat" issues.  If you're using 18" woofers, you'll probably need to create a frustum of a cone phase plug to reduce the self-cancellations due to the diameter of the woofer cone itself.

 

Chris

Link to comment
Share on other sites

On 8/28/2020 at 4:22 PM, rickmcinnis said:

I suspect my ports will look as close to yours as i can make them. 

If you want to move them closer to the throat to raise the throat-bounce frequency(ies), that's probably a reasonable thing to do.   I don't believe that they need to be reduced in size if you're only moving them a half inch or so closer (i.e., from 5" from the leading edage of the off-axis port to the horn throat entrance...to ~4.5" for the same data (i.e., the plural of datum), thus raising theoretical the first bounce frequency from ~680 Hz to ~755 Hz, etc. 

 

On 8/28/2020 at 4:22 PM, rickmcinnis said:

Of course, you have real experience with this but I must admit I had assumed you had removed some of that gasket if not all of it.

I just found a picture--it's a single gasket:

 

1378148364_IMG_3003cropforgasketview.thumb.jpg.eceb7c0fa785848f7a836e1d03ded2a0.jpg

 

On 8/28/2020 at 4:22 PM, rickmcinnis said:

Do you think there is any value in giving the drivers a 6 dB per octave slope?

I use an additional out-of-band attenuating PEQ or perhaps two to cut the out-of-band driver levels to the compression driver (high pass), and I use additional attenuating PEQs on the first two woofer response peaks above the effective crossover frequency (the crossover frequency is about 550 Hz currently).  I have attenuating PEQs at ~1300 and 2200-2300 Hz, wherever they show up.  I use enough gain on those attenuating PEQs to keep the bass channel roll-off curve pretty much constant above 550 Hz.  (I don't remember what I used.)

 

On 8/28/2020 at 4:22 PM, rickmcinnis said:

What is interesting and I know you do not have much interest in my measuring at the chair but I find it interesting.  One thing is the slope becomes much more steep  - almost an acoustic fifth order Butterworth slope while maintaining the same crossover frequency but with a hump at the crossover frequency before falling off - so almost getting a Chebychev-looking slope. Knowing as little as I do about this kind of thing I find that very interesting.  Also when both speakers are playing together the slopes become sharper still - again, at the chair, and my chair is about 2.5 meters away from the front of the horns.  

The problem with measuring further back is that the non-minimum phase in-room reflections clutter the measurements.  You can't do anything about those (and shouldn't) unless you're into "head in a vise" listening rooms.  I'm not...

 

If you look at Toole's book, he'll show flat on-axis response at 1-m or closer, but a tilted response (downward as frequency increases) if measured further away from the loudspeaker.  This is the "listening window"...as he puts it.  I see this using Dirac vs. REW measurements (i.e., Dirac measurements taken at the listener's positions, and REW measurements at 1m).

 

You can stick your head into the aperture of the horn and still hear a balanced presentation (as Danley has claimed several times in the past).  The effective minimum listening distance for the MEHs happens to be "zero" because of that phenomenon. I can detect no lobing.

 

Chris

  • Like 2
Link to comment
Share on other sites

Wreote the above at work and had to stop.

 

One thing about my chair response is that my high frequency response increases.  Now this could be since I am not measuring the K402 in its center - a littel bit below - they are sitting up really high in the air and my copper tube is not long enough and it is not that important at this point.  In MEH configuration I will be able to measure at the center.  The SEISMIC cabinets are 35 inches tall and, as you know the K-402 are mounted vertically.  Tall imposing thing!

 

I believe that with the MEH the response at one meter will  not be anyhwere near as diminished as it has been with a speaker with drivers VERY far apart - voice coils easily thrity inches away from each other.

 

I have not posted anything of real usefulness because there is no way to post an REW file here - they are too large and my pictures are useless.  I tried to email them to you, but again, they are too large.  I hope I am missing something and there is a way to do this.

 

One thing I found while measuring at 1 meter which goes along with what you have expanded upon to us from Danley's suggestion (!) to never use a filter with a man's name.  I greatly appreciate his male chauvinism in this respect.  I can just see more craven folk having to insert the obligatory "man or woman's name" in its place. But, what I found was that with no first order filter I got the expected for the greatest part of the range from 100 to 20,000 Hz a straight line at 0 degrees.

 

Then I used a 100 hz first order slope and zero degrees phase became a much larger number.  I went up 100 Hz at a time till 600 hz and by then the line remained just as straight as it was, from memory, but close to -1000 degrees.  When i would use the PEQs to get the response I wanted using no added filter, as you have stated, there is no phase growth to speak of.  So my feeling is that we should not use the first order filters in the XILICA unless one has no other choice.  

 

Will be interesting to see what is required with the woofers.  Maybe that shift will be useful?  Only measuring will tell.

 

I know what filters I need will be different from yours - I am not interested in copying them but It would be interesting to see your approach.  I think I am approaching this as you have since I have tried to use your advice along with the fact I had been trying this many months ago on a whim but with my old system, not the kooky one with the Seismics which actually have some phase linearity the old system did not display the goodness.  It shows me in hindsight that many of the things we have done traditionally were bandaids for hopelessly flawed systems which is how I would describe my system when it contained to horrendous EDGAR 75 Hz horn.

 

What Danley said in the old DIYAudio post was the purloined letter.  It was right there for all to see but from what I have read Chris A was the only person to see the beauty of it and put it into practice and of course bring it to our attention with his measurements.

 

Thanks again, Chris A.

 

Now to work on those holes.  By the 2way if I said I was going to make them smaller I made an error.  Mine will be as close to 1/10 Sd as possible.  I will err on being a little bit large instead of the other way around.

 

Take care,

 

Edited by rick mcinnis
correction
  • Like 1
Link to comment
Share on other sites

You should be able to find some freeware online to resize your pics to a smaller size so that you can upload. There are also online websites that you can upload your images too, and then it resizes to uour desire. Then you just download the new file and upload here.

Sent from my SM-T830 using Tapatalk

Link to comment
Share on other sites

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