quote:

---

to kind of go over what I've found, since there seems to be no risk in revealing it...and no reason to email me I realized

1. The hard thing is to put the cone drivers near of the high frequency driver. The best spot is 1/4 distant from apex (on main axis horn). This is the reason for small cone drivers use. If you can lower the x-over, you could use bigger cones.

2. Try small holes because you want omnidirectional radiation and maximum coupling between the cone drivers (same frequency band) at horn plane entrance*. Locate the holes at corners for same reasons.
If the holes are too big, the horn could not reveal the constant coverage pattern.

3. From what I've noticed all unitys have two holes on both corners, on each plane of the horn (so 8 holes)

4. "How do you calculate the point where the midrange exit holes are?
>Their distance from the tweeter? Is this the distance from the tweeter
>faceplate or its diaphragm? How exact does this have to be?

Ideally you would "time align" them. To do that you need to know the
acoustic center of the drivers at the crossover frequency. There is no easy
way to guess this that I know of. All the reactances in the driver create
phase shift which will move the effective acoustic center away from the
driver diaphragm. If you can measure the acoustic center, then the path the
acoustic waves travel from the acoustic centers of the drivers through the
horn need to be offset by the amount of phase shift introduced by the
crossover at the crossover frequency. For example, 90 degrees of phase
shift at 5kHz is approximately 343m/s / 5000Hz * 90degrees / 360 degrees =
1.7cm. The easiest way to do this is perhaps to mount the drivers to the
horn in such a way that you can move the midrange exit holes around, feed
the horn an impulse signal and then move the midranges around until the
impulse response looks good. Tom Danley probably modeled everything using
his fancy programs and only needed one try to get it right, but for the
rest of us, there is no substitute for trial and error.

5. the use of 6 midranges proves to be advantageous in allowing the horn to go lower

6.Depends on the time-alignment thing and the flare of the horn. The home
Unity is a 60x60 degree pyramid. You don't have to do it this way, though.
You can adjust the amount of horn loading you get by changing the angle. I
suspect that with a 5kHz crossover point and a large distance between the
acoustic center of your midranges and the exit of your midranges that the
midrange exits will have to be located pretty close to the tweeters, in
which case, you'd have to have a fairly narrow horn angle (less than 60x60)


7. "The
approach was also "trial and error" sort of.
I built and measured the "first guess" and THEN tried to refine the
relationships with acoustical or electrical theory.
With further prototypes the model got refined and the understanding of its
operation improved.
You must have the drivers aligned both in time and phase (it is possible to be
in phase but N cycles off in time) to make it work properly. The ranges have to
be adding together where the dimensions are less than 1/4 wl or spatial
problems result. You will need some test gear which can see this stuff because
it is unlikely you will find the dimensions by trial and error and when its
"right" it works very well."

8. If you look in the Lambda design... from the holes back to the throat of the compression driver it is slowing smoothed out until it is radial

9. "To make such a horn (where all 4 parts are the same) requires compound angle
cuts.
For a 60 deg horn angle, the blade angle is 37.7 deg and crosscut angle 26.5
deg.
If you are going to cut these out, do yourself a favor and make a cutting board
jig.
This is a board about 1/2 inch thick with strips attached that ride in the saw
table slots.
This sliding table then has a stop on it to hold the stock at the right angle
and a hold down to keep the wood from moving on the jig.
Remember, the exact angle of the horn wall angle is much less important than
having all the parts the same."

10. "What is the area at the throat end where the compr. driver attaches?
The actual throat for a 1" exit driver is usually equal to between a 5/16 and
3/8 inch diameter circle way back at the surface of the diaphragm. The Unity
horn has a 1" entrance if that is what you meant."

11. "A "secret" of the Unity horn is in utilizing the fact that the flare rate in
effect changes depending on where one is or where the drivers attach. Driven at
the apex, such a horn only goes down to about 1 kHz but if the drivers are
connected in a little down the expansion (where it is slower) the same horn
goes down to 300 Hz.
The other concern is to make the crossover network cancel out the phase/time
difference between the mid and high driver so that it really looks
like/measures like/sounds like a single horn and driver even though it is 5
drivers."


This is some "secret" stuff I've turned up from way way way back (2000)

If anyone can help me figure out how exactly one mounts the drivers to the horn and also what holes one should cut and where (a fromula to figure this out???)

the horn is simple.... there is plenty of info on how to make a horn

basically to make a 30 hz low cutoff horn you would double the area every 24" and for a 300 hz low cutoff double every 2.4"

I believe there is a formula for that but I Don't have it

From what I can tell the best way to try this is build a horn and try to make your best guesses with what you know about the 1/4 wavelength to the apex of the horn (there are formulas to calculate the 1/4 wavelength for that but I don't know what they are ... help?)... and then like John suggests, move the holes and use a impluse tone and equipment to check cancellation or whatever (again any help with what equipment is needed and what to do would be a godsend )

the importance of TAing the midranges I still dont' understand why you do this, anybody?

also who would one eliminate the phase issues (like tom suggests you need to solve both for perfection)

any help with equipment and procedures to fix these things... would be unimaginably helpful...

THanks and I hope I helped ya'll as well

---

quote:

---

Hi

Some answers to a few questions.
If your going to DIY a Unity horn, I would urge you to make one (not two) and play with it until you are happy, then make two with what you learned.

The holes in the corners that admit the mid and or low sound into the horn form a low pass acoustic filter, the purpose is to attenuate the unavoidable distortion components (all above the operating band). To copy something like Nick’s version of the Unity horn, you want the holes to be about 3/4 to 5/8 inch dia for a 3/4 inch or 1/2 inch thick horn wall.
The low pass “filter” is formed from the compliance of the air under the cone and the mass of the air in the hole (duct).

These intersect the horn forward of the compression driver entrance at the traditional spot at the apex. I would start with the mid holes entering the horn where the interior dimension is about 3 to 3 œ inches across.
Look for some sealed back 4- 5in mid drivers as a good place to start.
Typical “off the self” sealed drivers will give a high sensitivity which means they are loafing in operation.

The down side (which maybe isn’t really) of a straight sided horn like these tend to be is that they have less effect on the drivers response than a curved wall horn.
That is to say, if one measured the acoustic power from the driver on a plane wave tube, then a “perfect” CD horn has the identical response as it distributes the energy into a constant angle.
A curved horn does not, it can very significantly confine the energy at high frequencies into a small angle.
In the perfect case, the two cancel and at least at some small place on axis the response can be flat.
The CD horn must be compensated to account for the fact that the compression driver’s actual acoustic power begins to roll off at about 2 – 4KHz depending how long you had your wallet out when you bought the drivers. The $Tad 2001$ for example rolls off at about 4 KHz.

The up side of this is that what you eq to make flat is also fixing the acoustic phase associated with that response shape.
Second, the CD horn then delivers essentially the same frequency response everywhere in its coverage angle and this makes the spectrum of the far field very close to the near field spectrum. The effect o this is that even in another room, the CD system sounds normal where the one with changing directivity will sound murky and dark way off axis or anywhere other than on axis.
That is because the second system has a very much rolled off acoustic power at the top even if “on axis” it measures flat.

Crossovers.
I have Berringer 2496 and BSS366 that I use some times but most of the products I design have passive crossovers. Its kind of nice smoking a processed, 3 way active system with a 3 way passive one and one power amp.
I have not used the DXEQ but it looks interesting and possibly very powerful.

Conceptually the idea is simple but model a greater than 1st order crossover and sum the outputs and you see that while the response is flat, the phase rotates 90 degree per order going from well below to well above.
Using simple filters, no combination of physical off set, phase reversal or what ever will allow one to cancel the phase shift AND get flat response.
Worse, if one intends to have the lowest distortion or greatest output, one needs higher than a 2nd order high pass to limit each ranges excursion below band.

Once one can turn an intelligent computer program loose, one finds a solution that gives the right summed output BUT is not an integer order slope or may even be a changing slope at roll off.
It sounds like dxeq can do this adaptively BUT as I recall, it used LMS to measure the drivers. That is fine so far as making the driver to driver interaction work but I would still have a concern that it is not actually using the real acoustic phase for each range.
In a perfect world, one would be shooting for a condition of flat response AND zero degree acoustic phase, the only condition which allows a speaker to replicate a complex input waveshape in air.

The issue is made a bit more complicated when you remember the fact that all loudspeakers drives act as an all pass filter, they have a delay (the time between the input signal going in and sound coming out at zero distance) and as the drivers upper cutoff goes down, the longer that delay is. A large woofer may indeed have the equivalent of a number of feet of intrinsic delay with respect to its physical position, all drivers have a radiation point in time somewhat behind the physical radiator.
Add to the driver delay is the fact that ALL filters add delay to, that phase shift one see when summing the outputs is delay which increases as the frequency falls.
The lower the corner frequency, the greater the delay between the high and low pass sections.

Since being “in phase + - 90 degrees” but behind several signal cycles in time is to my mind obviously also not desirable being a gross time error (but the basis of some DIY and commercial horn systems, where the higher driver is in front of the lower) , it is imperative that the driver delay AND filter delay be compensated by physical placement and / or electronic delay.

Honestly, one has extraordinarily little chance of arriving at a suitable filter by trial and error or cookbook filter formula’s. Just making a simple passive filter that is happy with a compression driver for the load is somewhat involved.
Something adaptive like DXEQ or LSPcad would be the way to go.
On the other hand, there have been people who were delighted with the results on the first try with little test equipment. One thing people don’t realize is how much directivity improves imaging by reducing the room effects at the listening position.

One last thing, while I am a DIY’r too, I am also a single parent trying to figure out how to send his kids to college. I would ask that you NOT produce / sell these horns, that is what I am trying to do. On the other hand, making them as a DIY project for your enjoyment is something I am fine with.
Have fun.

Tom Danley

---

quote:

---

Här ser jag ett sammband, kanske tydligare än vad jag själv förstår...

Lite gammla nyheter från pro-världen :

http://www.servodrive.com/SPL-td1.html

---

quote:

---

Var hittade du allt detta ?

/Anders

---

quote:

---

Rekommenderar DE25 alt. DE250 som den numera heter från B&C, de andra elementen har jag tyvärr sämmre koll på, men som 12" skulle ju exvis B&C kunna passa också, enligt mitt knappa minne skall de ha någon PE modell med relativt hög verkningsgrad (ca 101dB/W/m) har samma fast 15" med en känslighet på 101,5dB/W/m, vill minnas att EV har någon modell med 103, likaså JBL, kan vara värt att ta en kik på, men vad det 5-6,5" gäller erkänner jag ett visst bortfall i förslag, alternativt om B&C har någon bra 6,5".

---

quote:

---

Får man inte problem med utsläckningar, funkar verkligen den dära horndesignen i ett hifisystem?

---

quote:

---

Jag tror att du skall försöka få till den där andra vinklen på tratten för det hjälper till att bryta upp punktligheten och på köpet får du en ljudbild med större bredd och mjukhet

---

quote:

---

Hrmm, Herr ordförande... hur går det med Unity ? Innebär det att man ska köra sina kära Altec till återvinningen ? Edgar goodbye ?
Har inte läst artikeln pga att en Klant lyckades tappa sina glasögon i sjön

---

quote:

---

men hur lågt kan ett sånt horn gå ? 80 Hz utan att bli för stort ?

---

quote:

---

Innebär det att man ska köra sina kära Altec till återvinningen ? Edgar goodbye ?

---

quote:

---

Benefits and Specifications of the Unity Horn From the Lambda Acoustics web site

This patent-pending technology is based on utilizing a pyramid-shaped, multi-band wave guide, to properly combine the amplitude and phase from a multiple number of drivers into a single coherent acoustic output. It is in effect a 2-way horn, 5 driver system including a simple passive crossover network that effectively acts as a single driver on and off axis. Unlike similar sized enclosures, the 16-inch square horn exit provides an essentially constant 60 x 60 degree pattern down to several hundred Hz. Using this technology, more of the sound is aimed directly to where the cabinet is pointed minimizing side and floor reflections. This benefit assists in controlling sound in highly reverberant rooms or where the maximum preservation of a recording's stereo image is desired.
Clarity Unmatched voice intelligibility and precise music reproduction due to the time-aligned phase coherent design. This is the only multi-way design to accomplish such in both the on and off axis response.
Controlled Directivity 60° x 60° dispersion pattern provides superb energy time curve response over an extended bandwidth while lowering undesirable side wall and floor reflections from reaching the listener and clouding the sound.
Frequency Response 300hz ~ 22,000Hz +/ - 3dB
Efficiency 100dB @ 1W1m with an 8ohm nominal load
Power Handling 100W RMS (120dB) typical music program assuming 6dB high pass crossover at 300hz
Distortion @ 112dB/1M 300 Hz = .477%
600 Hz = .194%
1200 Hz = .386%
2400 Hz = .967%
4800 Hz = 1.5%
Distortion @ 90dB/1M 300 Hz = .078%
600 Hz = .029%
1200 Hz = .062%
2400 Hz = .110%
4800 Hz = .049%
Unparalleled Performance Very low distortion, very high output. Consider at even 116dB each 5" mid range driver is receiving only 1W, the power compression and harmonic distortion levels of the speaker rivals those that only amplifiers could previously attain.
Dimensions 16.5" x 16.5" at the mouth outer lip, 14" deep overall, very small size for such high performance.
No "Horn" Sound! Due to the large radiating area of the 4 midbass drivers, the "horn" sound has been eliminated in this advanced patented design.

---

quote:

---

Men aldrig av horn-männen eller
TL-folket (märkte ni den lilla verbala skillnaden?).

---