Al Klappenberger

"How about going all the way to a 1" adaptor and the Great Plains 902?" I did a quick computer run on that possibility. The adapter would be roughly 4 1/2 inches long to take 2 Inch down to 1 inch. Reducing it to 1.4 inch would be shorter. Assuming I started with the right mouth size, both look practical. Al K.

Dave, I have no doubt you could do it. The problem would be finding time! The Eliptrac horn itself needs to have some manufacturing problems ironed out. It is obviously a nightmare to make. A round tractrix extender to reduce 2 inch diameter down to 1.4 inch should be a cake walk by comparison! Hopefully a second generation Eliptrac could be auditioned by someone other them me. I do not trust my ears. I have very little opportunity to listen to live music. All I can attest to is that it sounded very good! With the addition of a 1.4 inch adapter and a full range driver it could be tested for its full potential. BTW: One of the two Eliptrac horns is being shipped to Lee Clinton for more testing. We are hoping to do polar plots. He also has a DCX50 driver on an evaluation / purchase deal from B&C to use with it. The two drivers and LR24-9000 networks are being shipped out tomorrow too. That's all I can do for now. I wanted to have a guy here in town, who's opinion I trust, listen to them, but there just isn't time. AL K.

The initial listening tests shows the merits of the time-aligned high end. The most obvious advantage is the stereo image. There is simply no need for a center channel speaker. I had it turned completely off but you would never know it! The high end is somewhat subdued from what I am used to with the Beyma CP25 crossed over at 5800 Hz. The 9000 Hz crossover is so high that the mid-range driver is doing almost all the work. The tweeter is just adding "crispness" to the sound. You can actually remove the level adjustment plug from the tweeter attenuator disabling it completely without a major degradation of the sound. The midrange quality is very similar to the Bruce Edgar type tractrix horns popular now. The off axis dispersion is as good as with most separate tweeters as long as the throat is visible as it is with the elliptic tractrix horn. Unless you are really sitting close in, that will be the case. Instrument test showed that the curved sides of the Edgar type horn obscured the throat and interfered with the off-axis dispersion. Al K.

The new 9 Khz Linkwitz-Riley network is all done and everything is set to listen to the B&C DCX50 coaxial driver mounted on the Elliptic tractrix horn. I hooked up everything late last night. I expect to do some listening later today. AL K.

After a day or two of listening to all my recordings that I know well and that I consider well recorded, I have to say that the results are not what I expected. I am used to the JBL 2426h drivers on 1-inch Bruce Edgar type wood tractrix horns in my Belles. These are the original Martinelli Trachorn 400 prototypes. I expected the elliptic tractrix 2-inch horns to sound better. The fact is, I can't hear any difference! The Elliptic tractrix sounds fabulous and so do the 1-inch Edgar type horns! I think this has simply not been a fair test. The Elliptic 2-inch horn and B&C DCX50 driver is intended for the 400 Hz crossover of the Khorn, not the 500 Hz crossover possible for the Belle or LaScala. I have always said that 2-inch horns are for movie theater volume levels. Even the levels considered high for home stereo use are not high enough to cause significant distortion in the throat of a 1-inch horn. The JBL 2626 driver simply will not go down to 400 Hz. Distortion rises quickly below 500 Hz. This is the fault of the driver, not the horn. The only 1-inch driver I know of in current production that will go down to 400 Hz without distortion is the stock K55 (Atlas PD-5vh). I made the move from the K55 to a better driver, the JBL, years ago. There was definitely an improvement there. To make the same driver upgrade, a Khorn owner needs to move to a driver like the B&C DCM50 or DCX50 which was designed for a 400 Hz crossover. The 2-inch horn simply comes along for the ride! This test only compared the two horns and drivers in mid-range (squawker) service between 500 and 5800 Hz. Instrument tests showed that the off axis high frequency response of a 2-inch Edgar type horn was degraded when used with the DCX50 coaxial for a full range top end. The elliptic horn was much better off axis. This is the next thing I plan to listen to. I expect to get the last parts I need to finish building up the 9 KHz crossover properly this week. This will let me use the internal coaxial tweeter of the DCX50 driver rather then the ES5800 network and Beyma CP25 combination I have been using. Operating full range with the tweeter time aligned to the mid-range is where the Elliptic Tractrix should show it's worth. I expect to move future comments about that test to the other thread, "A time-aligned top end - Part II". Al K.

Well, the first results are in. My wife walked into the room with her fingers in her ears and asked if I was heard of hearing! I looked at the power meters on the amp, It said 1.5 Watts peak! I don't normally turn the system on when she is here. Now you know why! I really envy guys who's wife understands this stuff! Al K.

The listening test on the Eliptrac 400 horns has begun. I cut the frames down to fit my Belles just in case I decided to actually install them. UPS busted the sides off of both of them in shipping, so I had to do some hacking anyhow! Initially, I am only using the low frequency half of the B&C DCX50 drivers to listen to just the horns as a mid-range. I am still working on the LR24-9000 crossovers to move to the full coaxial configuration. Right now the ES5800 networks are running Beyma CP25 tweeters. Al K.

Guys, The speakers will be used mainly to take measurements. I'll be posting the new design here for anyone who wants to build it as DIY project. Just like I did for the Forte I. Because of the cost I would only offer it as a parts kit when it's done. Al k.

Dennis, That driver looks very pretty! It's another 1.4 Inch. These seem to be pooping into my view more and more. I was looking at a 1.4 Inch Altec (390) design made by Great Planes that might be a good candidate too. I think maybe we need to push Dave into making a 2 to 1.4 Inch round extension to put on the Eliptrac horn. I think it would be a neat accessory. That is if Dave can find the time to do it. I suspect he's got a lot of projects going already! I would love to do some testing on it. Al K.

Guys, I don't think the sidewall obscuring the throat is a problem at all for a simple driver with a single diaphragm used as a mid-range. It just seems to mess things up when the two drivers of a coaxial need to work together. Notice the green plot on the K2 horn goes west right at the 9 KHz crossover. That's where the two drivers are both operating and must sum together cleanly. Earlier tests with the DCM50 driver showed good off axis response all the way to 10 KHz, Dr. Edgar expected his type of horn to be crossed over far lower than that anyhow! I'm not sure what would happen to the off axis response with a wide range driver. I don't have one here to try. For use in the Khorn, it would need to extend all the way down to 400 Hz without distortion too. That's a tall order! That's why PWK couldn't do a 2-way. This is really the reason for doing a coaxial. It's to get as close to a 2-way as possible. Al K.

Dave will be sending me a second prototype of the Eliptrac 400 horn. When I get it and finish building up the networks properly I will actually be listening to his combination! Al K.

The two horns .. I tested the two horns for frequency response using the Linkwitz-Riley network. The plots are shown next to the two horns. The on-axis response of both horns is within +-5 dB. The 30 deg off axis was also good on both horns. The K2 horn, which is the Bruce Edgar type wood tractrix clearly has problems off axis when you get to the point where the curved sides start to obscure the throat. This doesn't seem to matter with a single midrange driver, like the B&C DCM50 but starts to cause trouble with the coaxial DCX50. I'll leave it to the horn experts to explain that! The elliptic horn had a smooth response even farther off axis.

The network ... The Linkwitz-Riley network is simply 4th order singly terminated Butterworth lowpass and highpass filters tied together. The main difference between these two and the usual crossover filters is that the two are 6 dB down their slope rather then the usual 3 dB at the crossover. To make a constant-impedance network both must be 3 dB down. The 6 dB crossover makes the impedance rise at the crossover. In this case, the 6 Ohm design impedance rises to about 12 Ohms at 9 KHz. To compensate, I added a simple R-L-C section. It could be simply left out since it just sucks a little energy out of the amp at the crossover, which makes it draw the same amount of current at all frequencies. This is what constant-impedance means. It does not contribute to the crossover itself at all. The reason for doing it is to provide a constant termination to the low frequency woofer to mid-range crossover network "upstream" that feeds it. One of my dreaded paste-ups is below. It shows the measured complex impedance at the input without the impedance compensator (top left) and with it (bottom left). The upper right is the picture of the jury-rig bread-board. The lower right is the response plot of the two filters. Note how the phase of the two filters virtually sit one on top of the other and the crossover point itself is 6 dB down. Unfortunately, I don't expect to make this design public as a DIY project for two reasons. The first is that a 6 Ohm attenuator is required. That's not an off-the-shelf item. Even worse is that a very small inductor of only 69 MicroHenries is required in the highpass half. It will require an accurate inductance meter to make it. The smallest inductor Solen makes is .10 MilliHenries. You must either pull turns off (a lot of them) or graft two together with a precise spacing between them and the leads connected in parallel properly to use the mutual inductance between them to set the value. That is what I did and it wasn't easy! Al K.

I'm going to start by repeating the opening idea for all this from the first "Time aligned top end" thread: ------------------------------------- I have always maintained that there is no good reason to go to a 2-inch horn to replace the K400 in the Khorn. A 1-inch horn is by far the best bang for the buck. 2-inch horns are for movie theaters but there is some advantages to upgrading the K55 driver. The problem with that is there is simply no 1-inch driver in current production that will operate cleanly down to 400 Hz where the Khorn must be crossed over. Evan the 1-inch JBL 2426h will start making distortion at 450 Hz let alone down at the 400 Hz required by the upper limit of the Khorn woofer. It was designed to be crossed over at 800 Hz but JBL says that it will operate down to 500 at reduced power. That's fine for the Belle or LaScala, but not the Khorn. Only a 2-inch driver will go down to 400 Hz. If you are ready to make the leap to a 2-inch horn and driver you might as well go a step further and make it really worth doing. B&C Speakers makes a coaxial version of their DCm50 driver, which is exactly what the doctor ordered for the Khorn squawker, a design specifically for a 400 Hz crossover. It's the DCx50. It uses the same midrange "guts" as the DCm50 but has a tweeter built right into the throat that extends the frequency range to 16 KHz. That's the same upper limit as the K77 tweeter that most of us have been listening to for years. A response that extends above that is nice but chances are only your dog is going to appreciate it! The big plus with a coaxial is that it's as close as you can get to having the midrange and tweeter occupy the same place at the same time. That means the propagation time from both drivers to you ears is virtually equal! That's TIME ALIGNMENT! Time aligned drivers let you use a far less aggressive crossover than the extreme-slope networks needed to eliminate the driver interference caused by two sources of sound located in two different places. The interference between drivers is significantly reduced. Another advantage here is that the DCx50 requires a crossover at 9 KHz rather than the 6 KHz required by the K55. This adds up to a far less aggressive network with fewer parts of smaller size. It will cost less than an extreme-slope network. Another advantage of the time alignment is the opportunity to us a phase-coherent type of network called a 4Th order Linkwitz-Riley. This type of network is usually only appropriate for sub-woofers as active filters where the wavelength is extremely long. ----------------------------- In the other thread, I had only Dave's (Gothover) Fastrac K2 horn to test with and no parts to build the Linkwitz-Riley network. Since then, Dave has sent me his prototype Eliptrac 400 elliptic tractrix horn. I also have received the parts and have bread-boarded the 9 KHz Likwitz-Riley crossover network. I had a very difficult several days going through all this. I kept getting all sorts of conflicting and illogical acoustic measurement for the two horns I was testing. I first thought the network itself was the problem. I bread-boarded a version of the AP15 series network, which is a 9 KHz 6 Ohm version of the Universal network high section. I also pulled an old ES7500 extreme-slope network out of the attic. After testing all three of these networks on both horns I continued to go bonkers getting all sorts of stupid plots. Finally I moved the horn off the workbench and set it on a chair facing out of the door. WOW, what a difference! It seems all of the hell I went through was due to the room! I don't think it was reflections either. The low frequency driver alone looked fine. I don't know what was going on, but it all went away when I moved the horn! Oh, for an anechoic chamber! The real tip-off to what was happening was the plot below. It's using the ES7500 network that has a "notch" in the lowpass section that rejects everything at 8800 Hz. DEAD! Notice that there is a big dip in the acoustic plot right at the notch. That says that only the tweeter is making a sound there. Obviously the dip could not be interaction when there is only one driver making the sound. It had to be the room! I also discovered earlier that both the two DCX50 drivers I have seems to have their two drivers mismarked. They are out of phase with each other. I am trying to resolve this with B&C Speakers now. That is why I tried both phases on the tweeter. The dip was identical both ways. That is further evidence that it was not an acoustic cancellation. Al K.

BUMP for reference by a new thread - part 2 Al K.

Dave, I think your going to need a second CNC machine and a full-time guy to run it! AL K.

Actually, The big Altec cellular horns are intended for outdoor PA use. They are really not a good choice for a home stereo. It would still be better then the k400. The Altec 511b VOT horn is a good choice. It's actually a bi-radial exponential horn. It just looks like a cellular. Personally, I suggest one of Dave Harris's FasTrac horns. Here's a 511b grafted into my Belle. I later went to the smaller 811b, then to a Bruce Edgar type tractrax like the Fastrac horns. Al K.

Guys, I just discovered something very interesting about the B&C DCx50 driver. The two driver sections are OUT OF PHASE with each other! I don't understand why B&C did this, but both of the drivers I have are the same way! Here is a paste up of the scope displays of the two drivers with the mike in the same place. The two are nicely in time, but one is inverted compared to the other. I inverted one driver and compared their phase shift from the driver to the mike. The phase difference between the two drivers is about 25 Degrees at 9 KHz. That amounts to about 1/10 of an inch! The inversions is very easy to compensate for though. You need only swap the connections to one of the two drivers! The plots are below. Al K. UPDATE: I have to back off of the 1/10 inch statement. Moving the mike back farther from the horn reduced the phase difference. There's something here I don't completely understand! [:S] The slope of the phase with frequency is always nearly the same between mid-range and tweeter drivers no matter the distance to the mike though. Phase slope is "group delay", so that implies time alignment. UPDATE 2: After a few hours of head-scratching: Group dealy = (pi * dPh / 180) / (2 * Pi *dF) From the plot: the mid-range shifts 360 Deg over 940 Hz = 1.06 mSec. Tweeter shift 315 Deg over 940 Hz. = .913 mSec. The difference is 1.73 mSec. speed of sound is 13,397 In/sec * .000129 sec = 1.73 In. That seems to much! But: 1/9000 Hz = .000111 Sec. The two are almost the same! Could the reversed phase be to compensate for group delay difference at their suggested crossover frequency of 9 KHz? DJK, HELP! UPATE 3: After another couple of hours: I think I'm kidding myself! [+o(]

Phenton, Since this is you 2nd post, I'll say welcome to the forum! The bottom line on the K400 upgrades is to DO IT! ANYTHING is better than the K400! Upgrading the K55 is secondary. It's not a bad driver and there is no currently made 1-inch format driver that will handle the Khorn 400 Hz crossover. To upgrade it will require moving to a 2-inch horn. Now you are talking a lot of money! Even the big 1-inch Altec drivers can be blown out by extreme lows that easily get through the 1st order woofer filters in networks like the A, and AA. They will generate a lot of distortion down low too. Al K.

Trey, Yes, I have three good mikes. I have an old Colony Mitty Mike II (an electret), a B&K 4133 and a Larson Davis 2559. I have a B&K 4220 pistonphone for level calibration. I use them with a Hewlett-Packard 3563a analyzer. My problem is no anechoic environment, The best I can do is with a pulse, but I usually use white noise. I have had fairly good results with on-axis measurements, but measuring dispersion on horns has not been easy! I like the noise method because I place a reference attenuator from the amp output into one channel of the analyzer set to equal the Mitty Mike output at 100 dB SPL. This makes the analyzer display 0 dB = 100 dB SPL at the mike no matter how loud I run the test. It's accurate so long as the mike is exactly 1 Meter from the speaker. The gated method is uncalibrated for level. I usually use the Mitty mike for response testing because it's a 1/4 inch and flat to 20 Khz. I mainly use the other two mikes for distortion tests, The Mitty mike starts to overload slightly beyond about 110 dB. The other two take a jet engine to overload them! AL K.

Mark, Don't worry about hijacking this thread. Now that Dave Harris is going to send me the new elliptic tractrix horn he just made I expect to start a new thread. When I get the parts for the Linkwitz-Riley network I'll be testing it and the DCX50 driver on that horn instead. BTW: I just heard from Solen. They will be shipping the parts I need for the crossover prototype today. I should have them next week! Al K.

He's kidding us. It's the fancy anechoic ceiling treatment that's the real secret! [] Al K.

Trey, I did a test once where I substituted the real drivers for the dummy loads when doing a network test. The results confirmed that the dummy load test results were close enough. What the drivers do whith what they get from the network is another story completely! I provide computer run tests results on all the networks I make but state clearly that it is for the network only! I do it partly to make sure I didn't forget something, sort-off like a checklist and to check matching between the two. Al K.

Mark, Thanks. It's just a matter of experience. Dave has been making horns for a while now and I have been writing little programs like this one since I got my first computer (a Heathkit H8) back about 1975. It's why I never learned how to run a spreadheet too. Al K.

Dave, MAN! That looks neater then cat droppings! [] It looks like a bolt-on adapter can be made that reduces it down to fit the K55 driver and maintains the Tractrix expansion properly too! Al K.