Jump to content

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


Chris A

Recommended Posts

Back to the topic of the K-402 based multiple-entry horn, it is the speaker design that keeps appearing at the top of my wish list. Some things I keep wondering is how would I know where to saw holes in the horns, what taper to cut the ports and if there are other engineering details I am not aware of, if I were to attempt this project.

Link to comment
Share on other sites

I was joking about using my Klipschorns as speaker stands, I would actually use a shelf above them. Even if (the not yet acquired) multiple-entry horns sound really good, having Rosewood Klipschorns is to own a masterpiece by a great artist, so I can't imagine letting them move out of my house.

That is one of the most beautiful veneer jobs I have laid eyes on.

Link to comment
Share on other sites

In all fairness to Klipsch Corp., the price difference was $800. each and for the extra money, the customer would also get the K-69 compression driver and the heavy duty, adjustable (I presume) horn mounting bracket/stand. But I do not need the bracket if the K-402 is going inside a cabinet and I would like to use a different driver, so I did not buy the K-402 horns as planned last month, not just about of the price, but because I don't need the extra parts. I am not unhappy with Klipsch for their prices, I just don't want the unused parts.  As I study more about the multiple entry horn designs though, I find a certain appeal about the specs of the 31" wide by 21" tall Danley SM-96, because if I bought some of those, I could still keep my beautiful 31" wide Klipschorns in the room and use them as speaker stands.

As a guy who has all Danley speakers in the main HT room (moved super Heresys downstairs) and a pair of K402's with TAD 4002's waiting for a woofer section and Active Xover, all I can say it's hard to go back once you have experienced Unity Summation Aperture/Synergy Horns with phase coherence and directivity control over a wide band. I will create several solutions based around the 402 in different configurations so that I will be able to do a side by side listening and measuring. Having had 402's before and needing the money and space at the time, I can now return to that now that Chris has made the idea possible.

Link to comment
Share on other sites

Thanks... :emotion-21: 

 

Perhaps with a little more effort I can help make it a bit more possible...and certainly much more affordable... :emotion-55: 

 

...but it will take a little more time and trial-and-error. 

 

Chris

  • Like 1
Link to comment
Share on other sites

Ummm, no...  guess again.

 

Note that what I'm talking about is something that you currently cannot buy, and it will have built-in woofer pads and extra stiffness for the LF duties that it will be performing.

 

Chris

Edited by Chris A
  • Like 2
Link to comment
Share on other sites

Chris,

 

Impressive setup, wish I could hear something like that someday (not sure I have room for a pair of 402s in my small room, though).

 

One technical quibble, of probably little importance, though:  the phase graph you show in the first post of this long thread:

New Center Mid-wall On-Axis.png

 

Is not really showing a linear phase.  The frequency scale of that plot is log, not linear, so a straight line that is other than 0 or 180 degrees doesn't mean the phase is linear.  Linear phase (waveform faithful) would require a straight line for of the phase when the frequency scale is constant in units of Hz (not in octaves, as the slope with a log-frequency plot would show).  Delay (seconds) would be proportional to the negative of that slope -- linear phase implies constant delay.

 

Here is a (rough) plot of what your data might look like in a linear-frequency graph: 

LinFreqPlot.PNG

As shown, the delay (i.e., downward slope) below about 2kHz is much larger than it is above 2kHz.  

 

Here is the same linear-frequency plot with the usual phase wrapping (to get the vertical down to a more visible scale):

Wrapped.PNG

The phase response is quite linear from 2kHz to about 12k and likely minimum phase above that (phase dropping due to magnitude rolloff).  Typical rule of thumb for "reproducing square waves" to the eye is that the phase be linear and magnitude need to be flat from 1/10x to 10x the square wave frequency (and the phase must also flatten near 0 degrees when delay is compensated to achieve a horizontal line).

 

I won't say that this is actually very important, though, just being technically picky. I went to a fair bit of trouble to get my Synergyish speakers to be relatively linear phase in the crossover -- and then added a FIR equalizer to smooth the phase further-- but the difference didn't cause a golden glow with angels appearing in the room or anything.  I'm pretty sure (but, from an uncontrolled test) that there's a positive difference in the midbass sound on transients when phase was linearized, but it's still not something I'd bet much money on.  At the high frequencies, magnitude fixes are definitely audible, but phase differences (comparing to a system with an impulse response that was already pretty compact) didn't seem to make much difference to me.  Possibly if the HF phase had been initially really screwed up then HF phase linearization might have been audible.

 

Bill

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

Hi Bill.  Good to see you here. 

 

I have just tried to plot linear frequency scale phase using REW, and this is currently the best that I can accomplish.

 

post-26262-0-25260000-1457625206_thumb.p

 

For whatever reason, REW doesn't like to go below 1 kHz on phase "overlays view".  I'm crossing at 475 Hz, so you really can't see the phase at that point in this view.  So I had to do two plots, one above 1 kHz, and one below:

 

post-26262-0-47100000-1457625830_thumb.p

 

Note that the driver used is a really inexpensive P.Audio BM-D750 Series I driver (a.k.a., Klipsch K-69-A) that I wouldn't call spectacular by any stretch.  I think that you can go buy a series II model for $169(US) plus shipping.  So this driver has a lot of diaphragm mode breakup at about 11-13 kHz - and it rings like a bell.  However, since the rest of the loudspeaker's performance is so good relative to the two-piece model (the Jubilee) in terms of polars and having all drivers within 1/4 wavelength at their crossover points, that I haven't been terribly motivated to switch it out with a TAD TD-4002 yet.  Perhaps I'll get motivated very soon.  I've got a few things on my plate right now (multiple tax returns included) that I need to get out of the way before returning to the horn problem.  This is a lot more fun...so I'll return to it very soon.

 

Chris

Link to comment
Share on other sites

This should give you an idea of how much the P.Audio 2" compression driver rings at ~13.9 kHz:

 

post-26262-0-41220000-1457629055_thumb.p

 

Chris

Link to comment
Share on other sites

I think that we're discussing the measurement aspects of the configuration. I really appreciate Bill's inputs in that area.

 

Let me introduce you two: Bill is proprietor of Liberty Instruments and a major contributor to the DIY multiple-entry horn realm.  Mark is a long time Klipsch Cinema enthusiast and fellow K-402 owner, as you can see from his avatar.   I might add that The Grateful Dead rank high in his playlist rotation.

 

As far as the discussion of how measurements correlate with what you hear...well, that's always in what I'd lump into the "Audiophile 601" bin.  Right now, just talking about the measurements is a good investment of time, I think.  It's the engineer in us that wants to know the measurements. That's were the real potential gain is had over all the subjectivity of listening only.

 

We can certainly talk about how the measurements correlate with the listening experience.  Presently I actually have few complaints about the K-69-A with the New Center in the center position, but I'm sure that once I swap it out with a TAD 4002, it will sound even better.  I already know how it will measure, since I've got the TADs on the K-402s on each side, and have many measurements of that performance, and I've got an idea of how it will sound, but I'm looking forward to actually hearing it all together nevertheless.

 

Chris

Edited by Chris A
  • Like 1
Link to comment
Share on other sites

I'm sure that once I swap it out with a TAD 4002, it will sound even better.
Are you intending to get some more of the TADs so you don't have to pull them off of your Jubilees?

 

Bruce

Link to comment
Share on other sites

Are you intending to get some more of the TADs so you don't have to pull them off of your Jubilees?
 

 

I'm planning to borrow one off the Jubs until I spring for a new one in a couple of months.  In the mean time, it's nothing to pull one off the Jubs for testing purposes. 

 

Chris

Link to comment
Share on other sites

Edit 19 March 2019:

An impulse spectrogram using a dual-diaphragm BMS 4592ND driver:

 

1387708314_K-402-MEHwithBMS4592tri-amped(1m)spectrogram.thumb.jpg.e2b9c4b27b8bcfae006909cf4934c4fc.jpg

 

Chris

 

  • Like 1
Link to comment
Share on other sites

Group delay of the K-402-MEH in the center mid-wall position (red trace) vs. Jubilee Bass bin in the corner (green trace), both taken at 1 metre on-axis at nominally 95 dBC (rev. 17 Feb. 2019).  The dip in the MEH trace at 172 Hz is due to in-room reflections from the front wall:

 

 

1422788876_K-402-MEHBMS4592ND(redtrace)vs.JubileeTAD4002(greentrace).thumb.jpg.d51fdc0c5f0af5784008e495e8efee70.jpg

 

 

Chris

  • Like 1
Link to comment
Share on other sites

A quick look at the relative phase responses of the K-402-MEH using first order IIR filters (red trace) vs. the Danley SH-50 (blue trace) using passive crossovers, and Jubilee/TAD TD-4002 (green trace) using first order IIR filters (rev. 17 Feb. 2019):

 

 

727251156_K-402-MEHwBMS4592ND(redtrace)vs.JubileeTAD4002(greentrace)vs.DanleySH-50(bluetrace).thumb.jpg.46f97a04ac9904635ee2fd9d3433eb28.jpg

 

Chris

  • Like 1
Link to comment
Share on other sites

Harmonic distortion...first the Jubilee with TAD TD-4002 driver, 95 dBC at 1 metre:

 

1567447320_JubileewithTADTD-4002at95dBHarmDist.thumb.jpg.a95b4a96242e7c06d25c5e2f4a68921e.jpg

 

Next, the K-402-MEH, 95 dBC at 1 metre:

 

528711338_K-402-MEHwithBMS4592NDHarmDist.thumb.jpg.2fe9774068f191a4c4ac733e64db9040.jpg

 

A lot of people pay close attention to harmonic distortion, but the real culprit is modulation distortion.

 

Chris

  • Like 1
Link to comment
Share on other sites

Impulse and step response of the K-402-MEH with BMS 4592ND driver using Zeroth order crossover filters:

 

1249711028_K-402-MEHwithBMS4592NDwithZerothorderfilters.jpg.673d368572499efe67bde22d9d1ece00.jpg

 

Chris

  • Like 1
Link to comment
Share on other sites

Higher order harmonic distortion... The previous two plots repeated below, but this time the second and third harmonic distortion curves are turned off so that you can much more easily see the higher order harmonics -- the ones that are audible because of the phenomenon known as "masking".  First the Jubilee bass bin:

 

Right Jub higher order HD 100 dBC 1 m.png

 

Next the New Center bass distortion (woofers only):

 

New Center higher order HD 100 dBC 1 m.png

 

Note the lower levels of higher order harmonic distortion for the New Center.

 

Chris

From my uneducated eyes...it appears that if you cross them over to a sub @60hz it will be very clean!

Link to comment
Share on other sites

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