[EV Dx38 active xover questions/observations]

Your comment about having the phases 90 degrees apart at the crossover is what I have come to think may be the right thing too.

It depends upon the application. In a home situation, where the listener is in the near field of the loudspeaker, a 90 degree phase difference at crossover probably doesn't hurt much because the resultant main lobe of the response gets sent into the floor or ceiling. In a sound reinforcement application, though, that main lobe gets pointed at a section of the audience, and is very audible to them.

The extreme-slope networks I build are inverted images of each other and are 180 degrees apart at the crossover, which is a very narrow band of frequency because of extremely sharp skirts. This says it would only be necessary to phase it right a one spot.

Essentially true, because the response of the non-dominant driver falls so steeply that it quickly becomes negligible.

It would seem to me that the difference in time dealy between the woofer horn and high frequency horn would throw that 180 degree relationship off.

Correct.

It would have to fall plus or minus from 180. With luck it might hit 90. If not it would be easy to add inserttion phase to either filter in 45 degree steps.The 6 dB crossover loss is automatic becasue the actual zero-loss crossover is the half power point (3.01 dB). The losses in a passive network add to the 3 db putting it pretty close to the nice 6 dB number. What do you think of that theory?

I don't completely understand what you are proposing. Determining the response at the crossover frequency (or anywhere else, for that matter) is pretty simple -- express the magnitude and phase of the drivers as complex numbers, and add them together. The magnitude of the sum is the magnitude of the combined response; the phase of the sum is the phase of the combined response. If each of your drivers has an amplitude of .7071 at crossover, and they are 90 degrees apart in phase, then their sum will have a magnitude of 1.0. If each of your drivers has an amplitude of 0.5 at crossover, and they are in-phase, then their sum will have a magnitude of 1.0. If they are 180 degrees out of phase, then you'll have to invert polarity on one of them. This is done in some Linkwitz-Riley crossovers, for example.

Greg

[EV Dx38 active xover questions/observations]

To the extent that anything I did influenced the parts that you like, thank you.

Greg

[EV Dx38 active xover questions/observations]

so you had something to do with the models for the filters in the 38 if they were grandfathered in?  that is cool!!  i really like this unit and have heard about the merlin as well. 

I was just one of many who contributed to a long line of digital processors in the Altec Lansing, ElectroVoice, Merlin, Klark-Teknik, and Dynacord lines. The DSP firmware in the Merlin was my personal responsibility, but I borrowed some from my predecessors, and others borrowed from me for later products. After I left, I think that all DSP development went to the Dynacord group in Germany, but they had all of my code and, glancing at some of the features of the Dx38 I wouldn't be surprised if they're still using some of it. Many of the built-in configurations of the Dx38 are dead-ringers for what I did in the Merlin.

Greg

[EV Dx38 active xover questions/observations]

This is nothing new and is one reason why a 90 deg releative phase shift is used (usually set by the order of the filters), since that combination will give a relatively flat spectrum given the various assumptions.

The 90 degree relative phase shift is a characteristic of Butterworth crossovers. If a properly-implemented Linkwitz-Riley crossover is used, then the lowpass and highpass sections are in-phase at all frequencies, and each has half-amplitude (-6 dB) at the crossover frequency. In such a situation, perfect time alignment leads to perfect sum-to-allpass frequency response in the crossover, and any remaining anomalies are due to the drivers themselves.

I designed the crossovers in the Dx38's great-granddaddy, the Merlin ISP-100, so I know that EV knows how to implement Linkwitz-Riley crossovers properly. I no longer have any affiliation with EV, so I can only hope that they continued the tradition in the Dx38.

I have been very careful not to mention the issue of the possibility of an audible "group delay".

Group delay is only an issue in a modulated system, which a loudspeaker is not (or should not be).

Greg

[EV Dx38 active xover questions/observations]

Anyone:  It seems as though the MASTER delay on my active is set on a minimum of 1.9 ms.  not the delay between drivers, but the master delay. 

First off, CAN I somehow get the MASTER delay to zero and if I do, then do I need to be concerned with the delays used on the drivers (set at 2271 us)

It doesn't say so in the manual, but I suspect that the 1.927 mSec is the inherent delay through the unit. This is the amount of delay through the A/D converters, processing paths, and D/A converters, and cannot be reduced.

I'm not certain exactly how much delay is necessary before lipsync problems become intolerable, but I understand the general rule of thumb to be that it is on the order of 30 mSec. And it's much less noticeable when the audio is delayed relative to the video than it is the other way around. It seems our brains are accustomed to hearing sounds after seeing the events that cause them, because the speed of sound is so slow compared to the speed of light.

Greg

[Digital madness.....(long no doubt)]

Incoming signals (be they digital or analogue) are converted into Goldmund's own unique 94 KHz / 24 bit digital signal (why - god knows) and then send through the pre-amp processor to the speakers directly.

I'm assuming that this is a typo, and really should be 96 kHz. Otherwise it's truly absurd because it guarantees that every signal will have to pass through at least one format conversion.

Now call me an idiot but there are all digitial amps out there - are there not.

If you're referring to Class-D amplifiers, like those from Bel Canto and Tripath, then no. They actually have an analog stage feeding the modulator.

I realize that in reality none of us have heard a proper digital system (in the way that we might have seen a proper digitial TV setup, for example).

The closest thing to a truly all-digital system that I've experienced was here: http://www.audiomn.org/Pages/feb02/HornShow.html

The only analog stage in the whole signal chain was the modulator for the Class-D amp. The system sounded REALLY good.

Now supposing that being all digital (upto the speaker) might carry an advantage would it not also be a good idea to keep the original digital data in its original form?

It depends. There are some formats that can be converted losslessly (upsampling in integer ratios can be, but usually isn't), others that cannot (DSD to PCM, or vice-versa, for example).

Why convert formats (along with word sizes and bitrates)?

Well, for example, signal processing in the DSD native format is difficult, sometimes impossible. Conversion to a form of PCM is almost necessary if you want digital crossovers.

I agree with you, though, that the format conversions should be kept to an absolute minimum.

Greg

[HEATWAVE]

yesterday evening in the Chicagoland area...100 degrees Farenheit...stifling...today...hot but a little breezy but hasn't hit the 100 mark again...

Bill

Yesterday at 1:00 PM I saw 104° F (40° C) actual temperature, here in Naperville.

Greg

[EV Dx-38 question..]

Do I infer this correctly (for someone who knows), it seems this has a built in equalizer and I can do some room corrections with it at the crossover level?

Looking at the manual (http://www.electrovoice.com/download_document.php?doc=67), it appears that there are 5 bands of EQ per input channel before the crossovers, and several configurable bands per output channel after the crossover.

Also, if I can do all these other wonderful things with it, is it possible that it will also allow me to vary the signal levels (or auto match?) between the hi/low end such that attenuators on my amps will NOT be needed? 

Yes. "Each channel employs a digital level control and a polarity switch." There is also delay compensation available.

Of course, my next question (if it can be done) is how do I figure out what the best setting is...  by ear?

I did all of my ISP-100 settings by ear. It just takes time.

Greg

[Help with mixing balanced & RCA plugs...]

Greg, what is the tradeoff to the resistor approach? Is it just lack of

compatibility with gear of different in/output impedances? What happens

if you use a mismatched resistor?

Other than the obvious (the fact that there is an unbalanced connection, with all of its inherent disadvantages), really the only performance issue that I can think of is the fact that this configuration can pass DC, which would be blocked by a transformer or capacitor. It is sometimes difficult to determine the true input impedance of the destination, and even more difficult to determine the true output impedance of the source, but I have found that getting "close enough" works pretty well. If you're WAY off, then it becomes more susceptible to noise and hum pickup.

And one last comment. When running from unbalanced to balanced you will

need to boost the gain by 12dB.

I think that it's actually 6dB, but it's definitely there. I have not found it to be a problem; there is more than enough gain in my amps to drive them to clipping despite the loss in signal level.

Greg

[Help with mixing balanced & RCA plugs...]

There's a quick and simple way to do it using just resistors, that works fine as long as "chassis ground" and "signal ground" are the same on your equipment. It's commonly called a "impedanced balanced" configuration; I've been searching the Web all morning for a diagram but haven't found anything.

Basically you always balance the lines, and you don't worry about whether the signals are balanced. ("Balancing" actually refers to impedances, not signals.)

So when you connect an unbalanced output to a balanced input, you connect source "hot" and "ground" directly to destination "+hot" and "common", respectively. You connect destination "-hot" to source "ground" THROUGH A RESISTOR EQUAL TO THE OUTPUT IMPEDANCE OF THE SOURCE, with the resistor located on the source side.

Going the other way, from a balanced source to an unbalanced destination, you connect source "+hot" and "common" to destination "hot" and "ground", respectively, but you connect source "-hot" to destination "ground" THROUGH A RESISTOR EQUAL TO THE INPUT IMPEDANCE OF THE DESTINATION, with the resistor located on the destination side.

It's a lot easier to do than it is to write about. I use this with my Merlin ISP-100 (kind of a grandaddy to the DX38).

Greg

[Jubilee bass bin frequency response compared to Khorn]

PWK's enthusiasm for rubber throats notwithstanding, I still think that constraining the first three folded sections to an exponential or tractrix contour will improve the response below 100 Hz. The increase in distortion should be negligible at the SPLs achieved in the home.

Greg

[Captions please for Klipsch Car Audio]

Example of improper installation. Should either be pointed back for thrust or up for traction.

[Preparing for Jubilee build]

Using ASIO is a clever way to go about it too as it will be more compatible with different setups.

ASIO is one of many possibilities; simply the one with which I am most familiar. Unfortunately ASIO is not as device-independent as it probably could be. There is a moderately complex section of code that must be customized for each sound card; the main portion containing the signal processing can then be generic.

Perhaps someone who has more Windows or Mac OS savvy can recommend some other candidates, as well.

The good news is that even complicated crossovers and EQ do not significantly load a modern processor. Been wondering what to do with that old Pentium III machine that's gathering dust? It's more than powerful enough for this.

Greg

[Preparing for Jubilee build]

as the good doctor has said, there is a smaller horn that can be mated to the jub lf and in fact, i demonstrated that last year with active stuff.  if enough people are interested, i can try to whip up a passive for a cinema jub, 510 horn and 69 driver.  maybe the smaller horn can kill some of the horn envy going on!   just as an fyi, Paul and i played with this horn as an alternative to the wood horn.

Everybody; do not overlook the use of a PC and a good sound card as a (fairly) inexpensive digital crossover. Sound cards like the LynxTWO (http://www.lynxstudio.com/compare.html) and others have outstanding audio performance for reasonable prices. (Of course, "outstanding" and "reasonable" are in the eyes of the beholder.) I have a fair amount of experience programming audio processing under ASIO 2.0, so I can help with that if anyone wants.

Greg

[Preparing for Jubilee build]

but i also appreciate creativity and greg was certainly going down a different path, one that we could have explored but because of business constraints, we did not choose (his drawings are going to call for some complicated cuts).  we wanted the cabinet to be buildable and at a profit and Paul really believed in rubber throats.  i was not condoning that plans be made and be made available for sale.  i am sure that the execs would frown on clones of klipsch speakers that are made and then made available for sale.  but in greg's case, it sounds like he wants to be in experimenter land and try something different for his own pleasure.  that is why i asked if he had access to measurement equipment cause i would defintely be interested in data. 

Roy,

I cannot begin to tell you how relieved I am to read your post. As the literal "new kid on the block" who jumped-in feet-first and started splashing around in the water, only to have Coytee ask the very proper and necessary question, "Should we be doing this?", I felt AWFUL. I would NEVER dream of stepping on anyone's toes at Klipsch or anywhere else.

As you pointed out, the design changes that I proposed would be purely experimental. At the moment I have no way to construct an enclosure and no way to test one once it was constructed. I am an engineer myself, so it is in my nature to experiment with numbers and design alternatives. There are no guarantees that what looks good on paper (or on the computer screen) will be good in reality.

At this point I do not know what to do. Having an idea for a design, or even a crude sketch, is one thing. Having a full 3-D CAD drawing is yet another. I certainly don't want to enable anyone to pirate your design.

Greg

[Preparing for Jubilee build]

The Jubilee (as I understand) is avaialbe as a commercial product

Oh, no; this is terrible. I honestly did not know that the Jubilee was still in production. When I searched for "Jubilee" on the Klipsch site, it sent me to the "Discontinued" section.

This is difficult, but under the circumstances I don't think that I can offer the CAD drawings without approval from Klipsch. I am sorry, but I have to respect their right to earn a living. Coytee, thank you for bringing this to my attention.

Greg

[Preparing for Jubilee build]

Can I ask what might come across as a rude question?

I understand, and generally agree. However, according to the Klipsch website the Jubilee is discontinued.

I discovered this forum less than a week ago because I was browsing the Klipsch site to find out what ever became of the bass horn that was described in that old issue of JAES. Thinking that it had probably been made into a product, I was hoping to someday be able to purchase one. One thing led to another, and here I am not only hoping to someday be able to build one, suddenly I'm modifying it as well.

My head is spinning.

Greg

[Preparing for Jubilee build]

  If you can provide the CAD drawings, I can get these cut on CNC. Please let me know how I can get them, thanks!

OK; give me a couple of days to triple-check my work -- I hammered this out REALLY FAST -- and then I'll arrange to get them to you. I work in TurboCAD, but I am supposed to be able to export to AutoCAD .dwg file format. Will that work for you?

Greg

[Preparing for Jubilee build]

i am impressed by your drawings.  for what it's worth, i think that you are on the right track.  will you have a way to measure the lf cabinet when you are done?

Thank you, Roy. Coming from you that's quite an honor.

As for measuring the results ... unfortunately, at the moment I am not in a situation where I can even construct a cabinet, let alone measure its performance. I am willing to make the CAD drawings available to anyone who wants to give it a try.

Thanks,

Greg

[Preparing for Jubilee build]

Also, logically, if exponential/tractix are so close in short lengths, again, I have to ask, what is to be gained by all that "extra" work?

For the horn lengths that I'm talking about in this situation, I compared the tractrix contour with the exponential. I used the horn calculators at "http://melhuish.org/audio/horndesign.html". All values are in cm or cm^2 (I haven't figured out how to get this to format nicely; sorry):

(Dist) (Tractrix Radius) (Tractrix Area) (Exponential Radius) (Exponential Area)

0.0 31.6 3138.1 31.6 3138.1

20.0 36.6 4211.8 36.4 4172.6

40.0 42.5 5670.5 42.0 5548.2

60.0 49.4 7668.2 48.5 7377.3

80.0 57.6 10439.3 55.9 9809.4

100.0 67.6 14358.4 64.4 13043.4

120.0 79.9 20078.5 74.3 17343.5

140.0 96.0 28962.0 85.7 23061.2

160.0 120.1 45348.3 98.8 30663.9

The area of the tractrix is within 10% of that of the exponential up to 100cm, then starts diverging prettly rapidly beyond. In the Jubilee tractrix, the horn is about 150cm long.

I am skeptical that there will be a lot of difference between a tractrix horn and an exponential horn under these circumstances.

Greg

[Preparing for Jubilee build]

Greg, Bruce Edgar has experimented with tractrix bass horns and he reports that they have trouble below 100Hz. But that's up to you to find out for yourself, if you want, of course.

You know, now that you mention it, I seem to recall reading that myself at some point in the past. But I never got around to pursuing the matter to try to find out why.

Also, logically, if exponential/tractix are so close in short lengths, again, I have to ask, what is to be gained by all that "extra" work?

I may have to eat my words on that, too. I just looked at the data -- this "thought experiment" tractrix Jubilee includes 59" of a 63.5" horn! That doesn't really qualify as a "short horn".

However, since nobody's cutting any wood at this point, the only cost is time. I happen to have an excess of that at the moment.

Thanks for your comments,

Greg

[Preparing for Jubilee build]

Edgar, that looks fun - what CAD program is that?

I use TurboCAD. Can't afford AutoCAD.

The problem that I see with what you are doing (an excellent excercise however) is alot of building work for virtually no gain - for me, a major drawback.

The gains, if any, would have to be quantified.

You don't want to use a tractrix expansion on a bass horn. That's why none exist out there already

For such a short section of horn, so near the throat, the tractrix, exponential, and hyperbolic contours are almost identical. As for lack of tractrix bass horns, I think that most bass horns were designed pretty long ago, when using a plane-wave approximation to an exponential characteristic was popular.

Greg

[Preparing for Jubilee build]

Different from previous renditions of the jubilee is the addtional vertical expansion you added close to the end.

I'm assuming that the "beginning" of the horn is at the throat; the "end" is at the mouth. Under those definitions, I did NOTHING to the end, but constrained vertical expansion at the beginning.

My question was simply to get an understanding of the pros and cons of doing the additional vertical expansion immediately rather than waiting til you got to the third section.

Are there any?

Constraining the vertical expansion near the throat changes the horn contour from something resembling parabolic to tractrix. The advantage that a tractrix has over a parabolic is efficiency. The advantage that a parabolic has over a tractrix is lower throat pressure, reducing the chance of horn throat distortion. At the SPL most people use at home the pressure probably doesn't get high enough to be a problem. In a sound reinforcement situation, where the SPL is much higher, that might not be true.

Greg

[Preparing for Jubilee build]

Creative is an understatement.

Any idea if how the specs would change if you ran a senerio using a channel flare begining immediately at the throat and ending at the exit path while wraping around the turns as it expands.

I'm not sure that I understand your question. If the rate of expansion is constant, then the horn is conical. If the horn is tractrix, exponential, or hyperbolic, then in each case the rate of expansion starts out slow and increases as you get farther away from the throat. If the horn is parabolic, then rate of expansion starts out fast and decreases as you get farther away from the throat. There are well-defined mathematical expressions for all of these shapes; anything not exactly following the math is still a horn, just not one of the named types.

Greg

[Preparing for Jubilee build]

Well, I've been playing with the CAD program. I haven't looked at this problem in several years, but finding this forum renewed my interest. Now that I know the shape of the Jubilee "skin" to pretty good accuracy, thanks to members of the forum, I can experiment with the internals.

I have found that creative application of internal "channels" can result in an expansion that is within a few percent of a perfect 39Hz tractrix characteristic over the entire length. See attached drawing.

Greg