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

[Chris A]

Note that the measured the output of this configuration (K-402-MEH) extends slightly beyond the bass extension of the Jub bass bins in the corners of the room.  This is true even though the unit is positioned 30 inches (0.76 m) off the floor with its back touching the front wall.

 

When positioned in the corner, its low frequency output behaves like a typical corner horn, and is basically limited by the room's dimensions and room-speaker coupling. 

 

If positioned on the floor, mid-wall (i.e., not elevated 30 inches off the floor) its bass extension is measurably better than mid-wall.  Setting these units up and getting them dialed-in using REW and a USB microphone is straightforward using email as the transmission means for the REW measurement files.

 

And having the active crossover is the real enabler.  Joe owns a Xilica, and that performed extremely well.  I received a Xilica XP-8080 yesterday to replace the three active crossovers that I've been using.  I paid about half that of a pair of third-party passive crossovers (at ALK's prices).  But this one crossover will suffice for all five surround channels (i.e., biamping the front three, mono-amping the surrounds). It does all the crossover filtering (48 dB/octave), time delays, equalization (for your particular room placement) and fast limiting to protect your drivers...all in one.  It's a 96 kHz sampling, 24-bit-depth unit.  I'd call that pretty spectacular. 

 

 

Chris

[Khornukopia]

1 hour ago, Cantilope said:

Just finishing up my K402 MEH.  Final sanding, primer and paint left, this is hands down the best center channel I have ever heard.  Worth building for sure.  I must say I was surprised by the clean solid bass this thing outputs. Turned out better than I had hoped.. I guess it's time to pretty up the room.

 

 

Nice to read about your center channel speaker. What high frequency compression driver is on your MEH?

[NBPK402]

32 minutes ago, Chris A said:

Note that the measured the output of this configuration (K-402-MEH) extends slightly beyond the bass extension of the Jub bass bins in the corners of the room.  This is true even though the unit is positioned 30 inches (0.76 m) off the floor with its back touching the front wall.

 

When positioned in the corner, its low frequency output behaves like a typical corner horn, and is basically limited by the room's dimensions and room-speaker coupling. 

 

If positioned on the floor (i.e., not elevated 30 inches off the floor) its bass extension is measurably better than mid-wall.  Setting these units up and getting them dialed-in using REW and a USB microphone is straightforward using email as the transmission means for the REW measurement files.

 

And having the active crossover is the real enabler.  Joseph owns a Xilica, and that performed extremely well.  I received a Xilica XP-8080 yesterday to replace the three active crossovers that I've been using.  I paid about half that of a pair of third-party passive crossovers (at ALK's prices).  But this one crossover will suffice for all five surround channels (i.e., biamping the front three, mono-amping the surrounds). It does all the crossover filtering (48 dB/octave), time delays, equalization (for your particular room placement) and fast limiting to protect your drivers...all in one.  It's a 96 kHz sampling, 24-bit-depth unit.  I'd call that pretty spectacular. 

 

 

Chris

I liked my Xilica too, when I had it...the only downside was that it was awkward to navigate IMO. I see yours has the phoenix blocks vs the XLRs I had on mine.

[Cantilope]

8 minutes ago, Khornukopia said:

What high frequency compression driver is on your MEH?

 

Currently using a Klipsch K69, but I built the box deep enough for the TAD 4002.  I am using a Xilica-4080 xover and a Nelson Pass designed Vfet DIY amplifier.

[Chris A]

13 minutes ago, ellisr63 said:

the only downside was that it was awkward to navigate IMO. I see yours has the phoenix blocks vs the XLRs I had on mine.

 

I'm doing the learning curve thing right now.  When finished with the initial install and copying the settings from the old active crossovers to the new one, I'll post any "lessons learned" and short cuts that I find.

 

The Phoenix connectors (also called "Euro") are bare wire connection blocks - you insert the three bare wires (+, -, and shield) into their respective receptacles, and screw down the connections using a screw driver.  No problem. All the inputs are connected into two blocks, the outputs connected in another two.  It's pretty easy to disconnect all four connectors for removal of the unit.

[Marvel]

We used a lot of equipment in the school's a/v department with the phoenix connectors. A lot less expensive than XLRs and 1/4 inch phone plugs. Plus, they are better connections.

[mark1101]

All you MEH guys..............I have a beautiful pair of Klipsch KPT-305s for $2200...........already in the heavy duty and pretty Klipsch pro boxes and with substantial rubber feet.  Perfect condition.  Might as well have them look good too.

[Chris A]

I found a couple of additional things about this MEH design:

 

1) The time delay I'm now using on the HF compression driver channel is about half of what I used last January.  This delay is compensating for phase lag entirely induced by the 48 dB/octave L-R crossover filters. The woofers actually lead the compression driver (physically) by about 0.5 ms in the horn itself. This updated delay setting has smoothed the overall phase of the unit in the 350-550 Hz band. 

 

2) The polars of this loudspeaker are almost unbelievable. The polars are controlled down to ~100-150 Hz...and can be seen at the point at which the off-axis FR curves begin to bend further downward with increase off-axis angles.  At 100-150 Hz, the room's boundaries take over the job of directivity control.  Note that this is about the same frequency that the Khorn and Jubilee bass bins also lose their polar control (they're actually at a slightly higher frequency). 

 

The design of the K-402 horn is to control to 90 degrees--total included angle, so the next-to-last 40 degree off-axis curve (darker purple) is where the horn was intended to start attenuating its SPL output. The green curve  shows the frequency response at 50 degrees off-axis, and shows how the horn does its job at that off-axis angle.  The vertical axis as very similar performance.  These off-axis curves are using 1/48th octave smoothing--basically no smoothing at all, so you're seeing the "dirty curves" as PWK used to call them.  The off-axis performance of the horn is basically ±2 dB from 100-15000 Hz, from 0-45 degrees off-axis.  (The higher frequency deviations are a function of the phase plug of the K-69-A compression driver itself and not the horn).

 

The following normalized horizontal off-axis frequency response plots are in ±2 dB increments--which you will never see from a horn loudspeaker manufacturer. 

 

 

This horn design is spectacular as compared to virtually anything else on the market, and certainly anything else that I've seen.  This is due to the dimensions of the mouth of the horn and the roll-out of the mouth using a tractrix profile.  Its mouth size exceeds both the Khorn bass bin mouths and the Jubilee bass bin mouths.

 

When you listen to this full range loudspeaker, your ears will tell you that something is definitely very different and improved.  I attribute that to the horn's ability to hold consistent polars all the way down to ~100 Hz--unlike other horn-loaded bass bins that I've heard.

 

It's definitely something that you'll want to hear for yourself.

 

Chris

[ClaudeJ1]

3 hours ago, Chris A said:

 

 

This horn design is spectacular as compared to virtually anything else on the market, and certainly anything else that I've seen.  This is due to the dimensions of the mouth of the horn and the roll-out of the mouth using a tractrix profile.  Its mouth size exceeds both the Khorn bass bin mouths and the Jubilee bass bin mouths.

 

When you listen to this full range loudspeaker, your ears will tell you that something is definitely very different and improved.  I attribute that to the horn's ability to hold consistent polars all the way down to ~100 Hz--unlike other horn-loaded bass bins that I've heard.

 

It's definitely something that you'll want to hear for yourself.

 

Chris

The k402 is still the best horn I've ever heard o rmeasured. Proven even more versatile by what you have done with it. Thanks for sharing.

[Chris A]

On 1/11/2016 at 10:21 AM, Don Richard said:

I know you have lots more work to do on this project, but I was wondering if you were going to run any polars on this hybrid.

 

The plot that I posted just above pretty much tells the story that you requested in January.  If you look closely, you'll see that the effects of the off-axis ports are quite minimal and the polar control of this horn is actually better than the other horn-loaded bass bins that you've seen on this site.

 

So the story of "how well does it work" is much better than I would have guessed when I started thinking about this design a couple of years ago. When you get the acoustic sources within 1/4 wavelength at the crossover frequency, they couple and start behaving like a single source (particularly horn-loaded acoustic sources).  I believe that this is the real physics and psychophysics of the secret sauce that Tom D. has been talking about. 

 

There seem to be no true disadvantages to the basic design for home hi-fi...assuming that you use your ears as the final judge of its effectiveness.  Adding one woofer but subtracting an extra horn and cabinet seems to be a strongly net-positive design trade.  I don't see any negatives thus far.  The harmonic distortion plots also show a net-positive effect (very strongly, I might add). 

 

If you wanted to use this horn on top of an existing horn-loaded woofer bass bin, it's clear that you should cross over at 100-150 Hz.  But you don't have to cross that high, so a direct cross to a tapped horn subwoofer or large conventional front-loaded horn bass bin could be done.  However, since the output of this horn is easily EQed down to 30 Hz (or even lower frequency if you have a pair in full corners on the floor) it seems to negate the need for subwoofers altogether. 

 

Also, not having the group delay growth of a ported box is a real discriminator in terms of bass sound quality (much more so than lower modulation distortion, I'm finding).  Since the woofers see the loading of the horn and then the corner of the room as the frequency goes lower, they seem to couple even more effectively than traditional folded corner-horn designs since the horn profiles couple so well to the room's corners. 

 

YMMV.

 

Chris

[ClaudeJ1]

Great post, Chris. I'm doing exactly what you are suggesting, but crossing at 80 Hz to one front and one rear Tapped Horn, each with slightly different frequency bands. It requires minimal EQ to get the bass flat to slightly below 18 Hz. Life is good with Horn Dynamics married to Headphone Definition and great Perspective Geometry, otherwise known as "imaging." We still have to seek better source material. That never goes away.

 

[mark1101]

5 hours ago, Chris A said:

 

 

So the story of "how well does it work" is much better than I would have guessed when I started thinking about this design a couple of years ago. When you get the acoustic sources within 1/4 wavelength at the crossover frequency, they couple and start behaving like a single source (particularly horn-loaded acoustic sources).  I believe that this is the real physics and psychophysics of the secret sauce that Tom D. has been talking about. 

 

Chris

You sure about this?  Danley uses the 1/4 wavelength explanation to determine where to place the woofer within the horn to generate radiation from the front and back of it.  Not sure about how his horns using multiple drivers work.  But his discussion about the tapped horn technology is a very neat idea.

 

What are your crossover frequencies?

 

[Chris A]

37 minutes ago, mark1101 said:

You sure about this?

Yes. 

 

Bill Fitzmaurice also has a thread on this subject for manifolding subs together--with and without a boundary wall gain present. 

 

37 minutes ago, mark1101 said:

What are your crossover frequencies?

I use 475 Hz within the K-402 horn (48 dB/octave). 

 

Why would you question this, Mark?

 

The tapped horn subs just behind each Jubilee are crossed at 40 Hz.  Steep slopes in both cases: 48 dB/octave.  The reason why I put the subs where they are and why I cross at 40 Hz is because the Jub bass bins and the TH subs couple well at that crossover frequency in each corner: a quarter wavelength at 40 Hz is 7 feet. 

 

Chris

[mark1101]

What are the big dips at 400 in the polars even at 20 and 30 degrees?  I would expect adding around the crossover given what you are saying.

 

All I was sayin was Danley's trick to get large output was about cabinet volume, wavelength, and driver placement.  I wasn't aware of any adding between different drivers such as in the Synergy line.  I probably just didn't read all the papers he published.  I was most interested int eh tapped subs and trying to understand how they worked.  That's all.

[Chris A]

42 minutes ago, mark1101 said:

What are the big dips at 400 in the polars even at 20 and 30 degrees?

I was afraid someone would misread the curves that I plotted above.

 

Those aren't "big dips".  The dip at 400 Hz is actually -2dB relative to the main on-axis SPL at the extent of the horn's coverage angle horizontally--about 35-40 degrees off-axis, where the horn wall effects begin to come into play for the K-402.  Remember that I specifically said that loudspeaker manufacturers won't post those curves at that magnification level.  The reason now is obvious: the potential for others to misinterpret the vertical axis scale.

 

If I came into your listen room and plotted the same curve for your MWM bins, you may be surprised to find that the SPL loss with off-axis angle (normalized) in that frequency band is much worse.  The reason why I posted this data is to show Don Richard the effects of putting ports into the horn (because he asked for it):  it's a much smaller effect in this horn (K-402-MEH) than I would have ever imagined.

 

42 minutes ago, mark1101 said:

I was most interested in the tapped subs and trying to understand how they worked.  That's all.

This thread doesn't discuss tapped horn subs, but there is one of those threads in the "Just Jubes" forum that you have access to, along with other threads on the open forums, and at www.diyAudio.com

 

 

Chris

[pbphoto]

According to my wife, I'm the only "big dip" in my listening room.

[mark1101]

OK, so the fault is in the Klipsch design.  Now I get it.  Didn't understand that.  Nothing to do with your experiment.  Sorry I don't know how to interpret your data.  You are just so much smarter than all of us.  Carry on.

[Chris A]

Well hold on there, partner...

 

The cause of that dip is the decrease in wall area at the point where the off-axis ports are at their maximum size--they're actually "rounded slots" oriented radially along the horn's length.  Note that the horn itself (Roy's horn) doesn't do that if those off-axis ports are not there. They're much smaller dips than I could have imagined - basically -3dB based on halving the wall area at that point.

 

If you look at the Danley Synergy and SPL Unity horns horns, those SPL dips will occur at two places along the horn's walls, and they occur at the frequencies where both the midrange and woofer ports are located at the 1/4 wavelength distance from the throat.  That's why Danley never puts a slot all the way across a horn wall.  The Synergies also have bass reflex ports that are even larger, but nearer the mouth of the horn.  The same effect will occur at the 1/4 wavelength point along the horn's walls horizontally.

 

The SH-50 has half the horizontal coverage angle of the K-402, so you will see those dips occur at about half the off-axis angle.  Since they're so mild (about -3dB) at the advertised horizontal angles, it doesn't really show up on the typical polar plots that are published.

 

...and I take umbrage at your "smarter" comment...

 

I merely apply my time more wisely to those tasks that make a difference...

[Chris A]

Here is a plot of the same data, above, but with psychoacoustic smoothing (i.e., what is typically perceived by listeners) applied to the curves and with a more reasonable vertical scale--5 dB per division (although I see a lot of plots that use 10 dB/division, too):

 

 

Note that the bottom two curves are beyond the advertised horizontal coverage angle of the horn, so you don't really need those curves to see the coverage.

[Chris A]

 

I want to also point out that the directivity of the horn still occurs even below 100 Hz (these measurements were taken outside on an elevated platform), as shown by the spread of the curves below that frequency at increasing off-axis angles.  The calculations based on mouth size of the K-402 say that it has directivity to about 40 Hz, at which point the room's boundaries take over.  That's pretty spectacular, and is a testimonial for larger horn mouths than has been typically used in the past.

 

When you go look at smaller "synergy horn" designs--like on the diyAudio forum---you'll see that spread of curves basically stops at ~400-500 Hz, the frequency where those smaller horns lose their polar control due to their much smaller mouth sizes.