A time-aligned top end for the Khorn

[jwc]

JC,

I did a series of plots using the Martinelli Trachorn 420. It's the only other 2-inch horn I have here. The on-axis plots looked exactly the same. I think it's the driver.

The plots are done using a 2-channel analyzer with a reference attenuator in channel A equal to the mikes output at 100 dB SPL. This means ZERO dB on the plot represents 100 dB SPL at the mike. The mike is exactly 1-meter from the mouth of the horn. That means the top line marked 30 dB is actually 130 dB SPL at 1-meter.

It looks like the entire response at 30 degrees drops about 5 dB and smooths out. It's normal to drop like that but the fact that it smooths out surprises me! At 45 degrees the curve side of the horn obscures the driver throat from the view of the mike. The picture below is looking at the horn as the mike sees it. The blur in the foreground is the mike itself. The other side of the horn must be acting like a reflector to disperse the highs. I don't know much about acoustics so I really don't know. Something is bending the sound to make it fan out horizontally. Listening to the sound of white noise through the horn as I slowly move off axis clearly tells me the dispersion is far better that I expected. I'm impressed!

Al K.

That is interesting...I'm surprised. The tractrix I have curved out all lost good coverage above 10K depending on what horn it was (mostly dependent on Fc). The pure exponentials seemed to do the same. When looking at your K2 from the side...it indeed looks pure tractrix that has been slightly shortened from the mouth end. I would have thought adding a conical flare to the start of the horn could give you a good coverage up top....but the horn doesn't appear to have conical in it.

Neat trick Al.

[SSnyder]

Please change the subject! I would rather argue religion or politics than this sales abuse thing!

okay is the pope democrat or republician

[Al Klappenberger]

JC,

No conical sections. Dave used the TRACHORN.EXE program I post on here a while ago to design the K2. It is a clone of the Trachorn 420 that Martinelli made for me which was also designed by the same program. I really have no explanation for the good off axis response other than it's bouncing off the curved side. I suppose some insight might be gained by testing a wide range 2-inch driver and similar horn that can go up to 15 KHz. Most of these Edgar type horns are only used for mid-range use and data on their performance at tweeter frequencies isn't common knowledge to Klipsch fans. I just don't know!

SS..,

I don't know about the pope. Could you accuse him of using his postion to sell bibles? That's the kind of fine line I walk!

Al K

[Al Klappenberger]

I just did the 30 Deg off axis curve again. This time I set the horn on the carpet and positioned the mike very carefully 1 meter away and 30 deg off axis from the front (mouth axis) of the horn and centered vertically. It still looks good!

Al K.

[jwc]

Al. That is REALLY impressive. Seriously.

[Al Klappenberger]

One more time! This is on-axis and 30 degrees done very carefully on the carpeted floor. 1 meter away centered vertically.

Al K.

[jwc]

Al...I have a thought on what is going on here but can't be sure. I hate to keep asking you for info....but I'm really interested in this.

Is there any chance you could label that last image with Hz markers like 400Hz...1000 and 8000 (in addition to the 15K).

Didn't you say the XO point is 8000?

jc

[Al Klappenberger]

JC,

The xo will be at 9000 Hz,

The plot is 0 - 20 KHz with each division being 2 Khz. It's a linear scale. You should be able to find any frequency easily be interpolation.

Al K.

[boom3]

Al,

Good stuff here. Have you considered a "contour" network to flatten out the treble? In other words, an attenuator for the tweeter with a "bypass" (and this is the legitimate use of the term) around the attenuator, with a roll-on of about 12 KHz or so? JBL did this with some of their larger studio monitors that Don Keele worked on some years ago.

[Al Klappenberger]

Boom,

I just don't like passive equalizers. They generate too much loss and screw up the impedance flatness unless they are very involved constant impedance types. Those make even more loss because they require twice the parts count. I count on the drivers being good quality. I believe small variations should be dealt with using active equalizers instead. At the level of quality of these drivers all you are doing is guiding the lily. A good active equalizer can deal with those little corrections and your room too! In this case especially, I am trying to make a Linkwitz-Riley phase-coherent network. Amplitude equalizer also shift phase!

Al k.

[boom3]

Thanks Al, please keep us posted.

[BE36]

Yes, Please.

I am woundering how much my 45 year old ears would notice a tweeter that is 1 - 2 dBs too hot?

I am thinking I might have to have one without the fix.

[Al Klappenberger]

be36,

I figured a dB or two would never be noticed, but there sure does seem to be a lot of people who do. It's why I had to start putting the #4 transformer tap on my universal networks. It gave it an intermediate setting that wound up as the "normal" setting (X - 4). I was setting them at 2 - 5. That's only a 1.7 dB difference! That is on the mid-range though.

AL K.

[BE36]

Al,

I do not worry about your Engineering.

I could hear a dB on Test Tone with mutiple freqs when setting up a Suround system.

But when doing the plots on my system with Stereophile test cd I was hard pressed to hear 1 - 2 dB when stepped up at 1/3 octave intervals, and my experianced Ears seemed to need a little extra near 16k . . .

I could raise my sqauwker 2 dB to get a really flat response in my room but I love the bass and need the extra up high, too many years in the factories or maybe it was that Deep Purple Concert in the 90s. . .

[Al Klappenberger]

All of this sort of thing is subjective. Whatever you hear and like is all that counts for you. I my case though, I have to make things flexible so others can set them to their taste. Setting things the way I like them would only be right for me! That's also part of why I seldom say how something "sounds". I just make instrument tests.

AL K.

[mark1101]

Al, what network are you using to get the FR plots you show? A breadboard of your new design? Does the driver come with any network?

[Al Klappenberger]

Mark,

All the plots are withOUT any network. There are two drivers in the DCX50. I am just plotting them separately but sometimes on the same graph. I have the parts on order for the new network but it's going to be while before they come in.

Al K.

[mark1101]

Excuse my stupidity..........I see what you did now.

[Rudy81]

Al,

I think this is an excellent idea. To date, the best system I have ever heard (by a long shot), is the Oris Horn Orphean system by BD Design

I beleive the 2" co-ax driver mated to the Khorn bassbin - would be equaly good - but even if it isn't, I'm hoping it would be 90% there at less than half the cost of the Orphean system.

Listening to the Orphean system was the number #1 reason why I gave up my Khorns (which I tried with all upgrades including V-trac+2"BMS driver, etc. etc.). But despite all the upgrades - my system was nowhere near as good as the Oris. Granted they are very expensive - but the performance was there.

Last year I got a chance to listen to som Oris horns and had a similar experience. That audition led me to my two way DBB actively crossed setup.

Al, you are on to something here. If you recall, when I had my Khorns I tried really hard to find a reasonably priced 2" driver that could match the bass bin and go high enough to work as a two way setup. The goal, of course was to reduce the number of drivers. Good luck with the project. Looks promising.

[greg928gts]

I've had the BMS coaxial 4592ND compression drivers/V-Trac horns, in my Khorns for over a year now. Hardly a new idea.

My question to the forum is this. Has anyone here heard a pair of 2" outlet coaxial compression drivers in 2" throat midrange horns in a pair of Khorns?

Greg