[synchronousmosfet]

I really have just one question, which is a front-horn question, but I can't find a thread closer than this, so please forgive me if this is slightly off-topic....(the question comes at the bottom of all this crap you gotta know first!!!!  )

although when I explain, maybe not so off-topic after all....

Having Audio Nirvana's and not being able to afford better....

I have decided to try the following design:

Match up an 8" AN to a front-horn, made of wood, just the way Vincent Brient did it at:

http://vincent.brient.free.fr/round_horns.htm

(very cool DIY project!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!)

for the following REASON (explained to me by M. Brient when I mentioned matching it to a FR speaker): the AN's, being low Qts and low Xmax, but monster magnet, and central whizzer, have a rising frequency response.....the horn will mellow them out, giving more body to the low-midrange and de-emphasizing the high-frequency rise....and match to air impedance better, giving a better efficiency, and better off-axis directivity as well.

So, "shoutiness" of AN's should be tamed, mid-range, especially low-mid-range, smoothed and given more body, the stereo sweet-spot widened a bit....

but the design can still be open-backed, so it has mid-range reflections for ambience (right word?)....

BECAUSE I'll crossover in the amp at either 250 Hz or 100 Hz and send the low-bass to a sub, so the AN doesn't have to deal with the low-bass at all, doesn't even SEE the low-bass...

NOW FINALLY (SHEEZ, I KNOW!) THE QUESTION :

when making the blinking horn using the tractrix computations, how do I set the diameter of the opening that mates to the FR speaker? If the cone is 8" diameter (where the edge meets the surround), then should the opening be precisely 8"? Or is there an advantage, especially since I'm not trying for low bass - that all goes to the sub - would I get better mid-range tonality from the horn by making the opening a bit smaller...say 7 1/2" or 7" or even 6.5"....even though the cone is 8"??

Front horn experts, waddya think??

Best, Charlie

[Magnetar]

Let's see = you are going to build a round tractrix horn because you have driver that is not suited for an open baffle? The AN driver will have mass roll off in a proper horn, no highs. Do you realize this?

I really have just one question, which is a front-horn question, but I can't find a thread closer than this, so please forgive me if this is slightly off-topic....(the question comes at the bottom of all this crap you gotta know first!!!!  )

although when I explain, maybe not so off-topic after all....

Having Audio Nirvana's and not being able to afford better....

I have decided to try the following design:

Match up an 8" AN to a front-horn, made of wood, just the way Vincent Brient did it at:

http://vincent.brient.free.fr/round_horns.htm

(very cool DIY project!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!)

for the following REASON (explained to me by M. Brient when I mentioned matching it to a FR speaker): the AN's, being low Qts and low Xmax, but monster magnet, and central whizzer, have a rising frequency response.....the horn will mellow them out, giving more body to the low-midrange and de-emphasizing the high-frequency rise....and match to air impedance better, giving a better efficiency, and better off-axis directivity as well.

So, "shoutiness" of AN's should be tamed, mid-range, especially low-mid-range, smoothed and given more body, the stereo sweet-spot widened a bit....

but the design can still be open-backed, so it has mid-range reflections for ambience (right word?)....

BECAUSE I'll crossover in the amp at either 250 Hz or 100 Hz and send the low-bass to a sub, so the AN doesn't have to deal with the low-bass at all, doesn't even SEE the low-bass...

NOW FINALLY (SHEEZ, I KNOW!) THE QUESTION :

when making the blinking horn using the tractrix computations, how do I set the diameter of the opening that mates to the FR speaker? If the cone is 8" diameter (where the edge meets the surround), then should the opening be precisely 8"? Or is there an advantage, especially since I'm not trying for low bass - that all goes to the sub - would I get better mid-range tonality from the horn by making the opening a bit smaller...say 7 1/2" or 7" or even 6.5"....even though the cone is 8"??

Front horn experts, waddya think??

Best, Charlie

[synchronousmosfet]

The AN driver will have mass roll off in a proper horn, no highs. Do you realize this?

No, but thank you for telling me. I thought the horn would just tame it, I didn't know it would kill the highs completely.

So, I guess I'll just ebay the AN's and take my loss, because I do want to make these horns, they look like too much fun to pass up. It's really about making the horns, if the AN's won't work, then I'll get new drivers.

A passing thought, though, just for the "intellectual excercise"....couldn't I make a double horn design, a horn for the tweeter part of the FR inside the much larger horn for the main cone? There would be a few struts in the way, but oval struts at 120 degrees apart, bottom and diagonal top left and diagonal top right would probably not interfere with the mid and bass too much. The small openings of both horns can be placed fairly close to the tweeter center and outer cone, since the Imax is only about 1 mm, I think......although the inner horn will have to pretty small or there will certainly be a shadow cast on the upper midrange. There were (and still are) coax's that had an inner horn....

Best, Charlie

[Magnetar]

If you want to build a front horn with treble build it for a compression driver. That's what they are for..

The low QTS AN driver can be used on an open baffle , just use a big baffle and some big mid QTS helper woofers

The AN driver will have mass roll off in a proper horn, no highs. Do you realize this?

No, but thank you for telling me. I thought the horn would just tame it, I didn't know it would kill the highs completely.

So, I guess I'll just ebay the AN's and take my loss, because I do want to make these horns, they look like too much fun to pass up. It's really about making the horns, if the AN's won't work, then I'll get new drivers.

A passing thought, though, just for the "intellectual excercise"....couldn't I make a double horn design, a horn for the tweeter part of the FR inside the much larger horn for the main cone? There would be a few struts in the way, but oval struts at 120 degrees apart, bottom and diagonal top left and diagonal top right would probably not interfere with the mid and bass too much. The small openings of both horns can be placed fairly close to the tweeter center and outer cone, since the Imax is only about 1 mm, I think......although the inner horn will have to pretty small or there will certainly be a shadow cast on the upper midrange. There were (and still are) coax's that had an inner horn....

Best, Charlie

[synchronousmosfet]

Thanks, good advice.

However, since I appear to have the help of someone quite experienced, could I please run something by you?

(assuming....) On a long discussion at diyaudio, there seemed to be somewhat of a consensus that compression drivers have higher distortion due to thermal effects....

is using a tractrix horn with a dome, cone or ribbon tweeter verboten? Intuitively, I'd like to make these horns for non-compression drivers....and I could actually make a round-cell ESL ribbon tweeter myself, that would be a fun project.

I was struck by the plasma tweeter shown recently at a show, an 813 tube oscillating and modulated, the little flame of plasma giving practically distortionless tweeter highs, and a big, gorgeous brass horn situated a good three to five inches from the plasma arc...certainly not a compression driver, but the horn was helping impedance match the little plasma arc to the room air, rather successfully according to the reviews.

While I contemplate how best to make and drive these big wooden horns such as M. Brient made,

I will take your advice and just go ahead and mount the AN's in an OB with a couple of good woofers beneath in an H enclosure. I plan to crossover in the amp, so the woofers will have their own mosfet amp; I don't know whether to crossover at 100 hz or 300 Hz, but that can be made variable and I can find out  by listening....I just need to crossover well before the low bass impedance hump in the AN's.

Best, Charlie

[Magnetar]

For your first horn design it for a 1" throat compression driver but segment it so you can have different sized throats for different sized drivers.

In a home environment compression drivers are rarely driven to the point where heat would be a factor - look at the efficiency! - so like many things over 'there' it's really just a bunch of needless talk

The AN driver's amp can be high passed so you won't have a big cap in series with the driver. All you'll need there is a cap and resitor on the amp inputs.

[D OB G]

Hi synchronousmosfet,

I believe the consensus on DiyAudio is that the compression driver suffers LESS from thermal effects (it is much more efficient).

Certainly Earl Geddes maintains it does.

David

[synchronousmosfet]

Magnetar said:

but segment it so you can have different sized throats for different sized drivers.

Brilliant! I believe I can do that with the tractrix contour...what an outstanding idea. Or do you recommend a different contour than the tractrix, for a compression tweeter?

David said: (paraphrased)...that Gedde maintains the compression driver is LESS subject to thermal distortion...(or to that extent, I think)...

OK, I'll bite. wothehell, I'm flying blind anyway, I might as well accept good advice.

OY, now I have to figure out where to put the crossovers, and which compression drivers to buy....

Let's start here, if you guys wouldn't mind giving me some guidance....if I'm going to go with making horns, then I might as well go three-way. The PCB from Rod Elliott for the Linkwitz-Riley 24 dB/octave crossover has provision to make it three-way, and I can select my own crossover points. I can make the amps for all three drivers without difficulty, that, at least, is not a problem. I really prefer crossovers before the final amp stages, not crossovers at the speakers.

However...

The AN driver's amp can be high passed so you won't have a big cap in series with the driver. All you'll need there is a cap and resitor on the amp inputs.

...I see your point there, allright...I could make the compression tweeter and add it onto the AN in OB, but relieve the AN of the treble...and do so easily by simply RC'ing the input to the AN amp to roll-off the treble. Good, elegant, cheap solution.

My thoughts on this are as follows: for the bass, I should either crossover at about 100 Hz or at about 500 Hz. My reasoning for this is I don't want to have the crossover cutting through the middle of 250 Hz, which from my reading is an area of prime musical content. Crossing over at 500 Hz puts over half the musical power into the woofer...but crossing over at 100 Hz keeps the 250 Hz area within the mid-range driver's spectrum, and still relieves the mid-range driver of at least 20 to 25% of the musical power requirements.

for the midrange, I want to keep the 2.5K to 3.5 K region free of any crossover, because that's the region where operatic voices "ring". (And musicals too; one of the best vinyls I've got is a GERMAN performance of "Oklahoma"...it's not only a hoot, the voices are simply gorgeous...but the Magic Flute in German is not to be surpassed, so either way that 2.5K to 3.5K region is critical.)

On the other hand, what I listen to the most are symphonies, which have most of their frequency content from 125 Hz to 8 KHz...so that would argue for the 100 Hz crossover for the sub, but I don't know if it's reasonable to get a mid-range driver inexpensively to go all the way from 100 Hz to 8 KHz...

but then I don't know if a compression tweeter for a front-horn would cover from, say, 4 KHz to 20 KHz....

So, keeping in mind that I'm POOR and likely to remain so for the foreseeable future, but I CAN make wooden round horns like M. Brient did in that link I listed earlier....

and seeing ahead that the mid-range driver is probably going to be the most expensive component - except for the horns, but I'm going to make those myself (correct me if I'm wrong about the mid-range driver), what compression driver and range for the tweeter would be a good starting place?

I should add that I'd still like to use a cone midrange and cone woofer, so I can still have an OB design for the midrange and bass...but you've talked me into using a compression driver for the tweeter, for my first attempt at making a front-horn. I can always add a rear-firing tweeter if I need to, but I've read posters who say the rear-firing tweeter is not needed for the OB design.

Best, Charlie

[D OB G]

Hi synchronousmosfet,

There are at least two schools of thought regarding crossover points (assuming you accept that a full-ranger really can't handle the freq extremes.  Even an AN).

Avoid areas that the "ear" is sensitive to, or the "physics" approach- crossing over according to the diameter of the drivers, the centre to centre spacing at the crossover freq, and the x-max of the drivers at the crossover.

If you take the "ear" approach, then different people believe different freqs are more important than others.

Earl Geddes believes you should never have a crossover above 1 kHz, but below is OK.

Others believe it should be above the point where harmonics are predominant i.e. above say 3 kHZ, 5 kHz, 7 kHz e.t.c. depending on who you listen to.

Others believe a driver can't properly handle more than three octaves, meaning, say, 20-150, 150-1200, 1200-10000, supertweeter. i.e. a 4-way.

The physics approach maintains that unless you are using digital filters of very high slope, there is sound from both drivers either side of the crossover for at least two octaves (24 dB per octave, minus 40 dB or so being the best case scenario), lower slopes obviously having more simultaneous output. i.e. an attempt to avoid an "ear" sensitive region needs to avoid it by a large margin.

The physics approach maintains that the ear doesn't prefer certain freqs (bearing in mind the varying sensitivity, not preference, at different freqs as shown on the Fletcher-Munson curve.  A flat response is still a flat response).

Using Rod Elliot's active Xover board assumes that the driver has a flat freq response either side of the Xover of at least two octaves, ideally, and that there is no need for notch filtering, or equalisation.

All this is obviously of no help all to someone deciding on a Xover I'm affraid, but it is just a start to addressing the issues involved.

By the way, a 1 inch compression driver, in a waveguide of a big enough diameter, 15 inches or so (oblate spheroid being the best), can go from 800-900 Hz to 16,000 (just ask Earl Geddes)!!

Trying to be helpful ,

David

[synchronousmosfet]

David, thanks, that was extremely helpful.

I find a good overview the best place to start; start global, then find specific examples that help focus the specific strategy.

This matter of front horns has really changed my thinking. I have to say I'm starting to find myself agreeing with Gedde, which is a stretch, he tends to come across as abrasive, so that only added to my intuitive resistance. However, intuition is often wrong, especially when it comes to engineering, as I've found out in doing research.

The specific example that really has me rethinking all of this, and which fits in with one global way of approaching this area, is here:

http://www.onhifi.com/product/amphion_xenon.htm

A tweeter that covers 1200 Hz to 20 KHz. Now, that's a paradigm-shift!!!!!

But I'm psyched. I get it. Now I understand where Gedde is coming from (he really should have somebody do his PR for him...he's his own worst enemy, as so many of us GeeKs are...).

Now I have to decide whether to purchase Gedde's waveguide (and driver?)...which I probably cannot afford...or make my own.

So, here are two specific questions: is there software or an equation I can plug into a spreadsheet that will give me the contour I need for an oblate-spheroid waveguide, or should I just go with the tractrix round horn for the tweeter?

Either way, I will make the waveguide using the wooden disc buildup method, and a homemade lathe large enough for the 15" or so outside diameter. (Although I will go to Gedde's site and check the price on the DIY waveguides).

The second question is this: in the moderate price range, any specific recommendations for a compression driver tweeter that, using a waveguide or tractrix horn, can cover 1 KHz (or 1200 Hz) to 20 KHz? (or 1 KHz to 16 KHz, and I'll just add a super-tweeter, not a big deal, I think....since the critical area is 500 Hz to 6 KHz, and God knows I've listened to my Infinities with the tweeter fuse blown enough times without missing them all that much... )

It also seems to me that a really decent, large woofer can easily deal with 50 Hz to 1200 Hz, so this could be an easy triamped system: tweeter (super-tweeter, if needed, simply added on with a capacitor roll-on), woofer, and sub...if needed. What would be really cool: a biamped system, tweeter with horn from 1 KHz to 20 KHz, and woofer from 25-30 to 1 KHz.

Then, if I want OB-type ambience, use two 12" or 15" woofers in an open-back baffle, add some small rear-firing drivers for the rear-firing midrange, and I've got it all.......

Best, Charlie

[Russell Dawkins]

A tweeter that covers 1200 Hz to 20 KHz. Now, that's a paradigm-shift!!!!!

.. then one that covers 590Hz - 22kHz would really be a paradigm shift!

(SP Technology Timepiece, Continuum and Revelation)

[D OB G]

Here are some oblate spheroid sites:

www.audioheritage.org/vbulletin/showthread.php?t=12751

www.audioheritage.org/vbulletin/showthread.php?t=12126

www.crestviewcable.com/~catapult/Geddes.xls

David

[D OB G]

P.S.

As you can see, there is an efficiency hump around the lower end (happens if you use a dome tweeter too- see Zaph's site), so some equalisation is needed.

[synchronousmosfet]

.. then one that covers 590Hz - 22kHz would really be a paradigm shift!

Umm...Russell, I think you're making a joke, but I don't know what it is....

Is it that 1K to 20K is old news, and I'm just catching up? I always had the mis-impression that front-horn loaded drivers were of limited range...obviously not so.

Best, Charlie

[Russell Dawkins]

I'm not joking - SP Technology actually high-passes their HF driver at 590Hz and it runs right up to 22-24k. The tweeter they use is normally crossed at 2k, but the gain from the waveguide lets Bob attenuate the bottom end of the tweeter's passband by 10dB, saving it from destruction.
Look into the SP Technology circle on this forum.

[synchronousmosfet]

  wow.

well, I'll just mosey on over there and check it out! Thanks.

Best, Charlie

[BrassEar]

DDS or Geddes Waveguide
B&C DE250 compression driver
1 or dual Eminence Alpha 15A drivers used OB
first order passive or active xover

Worked for me.

[synchronousmosfet]

Thanks, BrassEar, that's what I plan to do, glad to have the recommendations, esp for the woofers.

In the meantime, until I have the money to buy those drivers and the Gedde waveguides, I'm going to play with the AN's. Since I want two wood horns, round, tractrix, for 8" drivers ANYWAY, I'm going to go ahead and try them on the AN's.

I have this conjecture...we'll see if it's true...that the phase plugs in the center of the AN's will prevent the horn from killing the treble. Now, I wonder what happens if I make a much longer phase plug to use with the horn..... ...keep making it longer and longer and see how the treble balances with the mid-range...

Best, Charlie