Full range in a front loaded horn

Sella

Member
2004-08-12 2:19 pm
Espoo
In my new project I am going to use full range driver in a front loaded horn. With hornresp program I can calculate theoretical frequency response with knowing horn geometry and driver's T/S-parameters. The problem is that I would like to add drivers hf response to the analysis. Then I could optimize the horn to match driver's hf-response.

I have measured the driver's frequency to half space (ie fitted to infinite baffle). It isn't clear to me how to add these two responses together. I was thinking following:

1. It is an 8" driver so, frequency response start's to roll off when wavelength is over 4" and about 3400 Hz in frequency. It is my intention to use driver's response from here on upwards and add a roll of 6 dB/oct downwards to the measured response.

2. This response is add to simulated response.

My idea is to divide the response to two subsystems: the horn system and approximating that from 3400 Hz upwards the driver behaving same as radiating to half space. The goal is to make sure that horn's upper cut off frequency matches driver's rising response (Fostex 206) and not causing a remarkable dip or peak.

I am making any sense with this? :confused:
 
Last edited:

Sella

Member
2004-08-12 2:19 pm
Espoo
Yeah, I tackled down that problem as I noticed that hornresp is also able to add the piston response to the horn's response (trough: tools - directivity - response) :)

But how the **** you guys get any sense to this front horn + full range thing? I tried to model Azurahorn 204 and Oris 200 and I always get the frequency band which the horn horn amplifies to play about 5 dB louder than the hf-range (according to the manufacturer's measurements and my own).

I think there is some piece missing from the puzzle..

I posted as an attachment my hornresp simulation of the Oris 200 horn and Fostex 206 driver. According to Fostex driver's spl is somewhat 100 dB up from 2k Hz. So without horn it is 5 dB supressed and with 5 dB attenuated.
 

Attachments

  • Fostex206e.pdf
    14.7 KB · Views: 301

Ivo

Member
2005-01-25 11:16 am
Amsterdam
I built a number of tractrix horns for Lowther drivers a few years back. I had a lot of contact with Bert Doppenberg of BD-Design/Oris horns. Hornresp only simulates the predictable behaviour of the driver and horn combination. It won't be able to simulate the cone break up for you, so Hornresp cannot predict what happens in the higher frequencies. The horn also won't let through the higher frequencies unaffected. In an Oris-like configuration, it forms a a mild low-pass filter, rolling off the highs at about 3 dB/octave (if I remember correctly).

So what you have been doing is roughly right, the horn-response is dominant in the horn pass-band, above that you eye-ball the response from the graph and the expected roll-off. And then measure to see what you've got. Bruce Edgar had some tips about acoustically modifying response in his midrange horn article (see Volvotreter's website, the download section). Playing with the backchamber can also be helpful to adjust response.

The Fostex FE206 is used with success in front horns, so you have a good chance of succeeding.
 

Sella

Member
2004-08-12 2:19 pm
Espoo
Thank you for a good answer. It always gives confidence knowing that is heading to right direction.

That SPL pump still worries me. Has anyone ever posted any measurement of Oris or Azurahorn with any full range driver? I could not find any.. if there is they must be buried deep in some thread.
 

CLS

Member
2005-06-17 6:58 am
Taiwan
I got rough RTA measurements by the built-in function of Behringer DEQ2496.

My Oris 150 is loaded by a Focal midrange Audiom 7K rated 98 db/w, and xover'ed to a waveguide loaded tweeter Eminence APT50 rated 105 dB/w.

The overall pink noise response of my system shows downtilt - bass/mid are higher than the treble, i.e., sensitivity of Oris 150 + Focal is higher than 105 dB, I'd say by 2~3 dB in an overall shallow slope. If comparing both ends of spectrum, the total drop on the top might approach 10dB. I tuned it that way on purpose, I like warmer balance, and often listen quietly.

Interestingly, I once briefly tried Lowther PM4 in Oris. It played obviously louder than the Focal (even though their rated efficiency is the same at 98db/w). I didn't have APT at that time, so no direct comparison.
 

CLS

Member
2005-06-17 6:58 am
Taiwan
Oops, it's not 10dB drop, but 20.

IMG_2920.jpg
 

Sella

Member
2004-08-12 2:19 pm
Espoo
Thank you for your kind answers. I have tought this matter quite a lot lately and I think I will go to coaxial driver instead of full range. I think that full range drivers poor hf would need a super tweeter in any case. Currently I am running PHY-HP KM30 in IB and I think that I would take a step down that way.

I have tought of these drivers:

BMS 12CN680 (12")

and

Beyma 8CX300Nd (8")

Both drivers employ a compression driver and a waveguide for the hf-range.

Here are some initial simulations with Beyma driver with closed back chamber. I guess cross xo-points would be somewhat 300 and 1000. With vented rear chamber It is able to go down near 200 Hz.

I also started to wonder where I could find 12" coaxial driver which fs would be near 30 Hz? I was thinking to go Avantgarde Solo and do a full active system with larger front horn, back chamber tuned near driver's fs and the slope taken care with active xo.
 

Attachments

  • Beyma.pdf
    14.7 KB · Views: 77
...and I think I will go to coaxial driver instead of full range. I think that full range drivers poor hf would need a super tweeter in any case.
I'm listening to a coaxial in a 160Hz Azura.
It can sound quite nice IMHO.
Using 10" Tannoy Reds, w/ 15" 60Hz front horns below.
no crossover used on top or bottom of the Tannoys, or on the 60Hz horn.
A pair of Tapped horns for the very bottom, and super tweeters crossed in @ 10k.

Definitely not a "full-range" system (so doesn't go on this forum), but I don't know how else to actually get full range.
 

CLS

Member
2005-06-17 6:58 am
Taiwan
I'm listening to a coaxial in a 160Hz Azura.
....

Hi,

Would you please share more?

I've been thinking of horn loaded coaxial (maybe 8~12" with 1" compression driver), but it seems there'd be directivity problems which are hard to resolved.

At the typical xover point of such assembly - around 1kHz, the sound from the treble unit is largely guided by the cone of mid-bass unit, which is mostly very wide conical. So I'd imagine the waveform of the outer edge would hit the horn wall (near the throat). The treble sound leaves the smaller horn and then meet another bigger one. The transition between the 2 are not smooth :( So I think a horn or WG for such driver should be very wide and shallow, thus it can minimize or eliminate this issue. (while the loading and directivity control of the midbass unit are another 2 stories.... )

That is in my thinking, though. I don't know how it behaves in reality, and whether it's a real problem or something very small which can be ignored.

Hmmm......
 
hmm, where to start...
this was just thrown together days ago, just as a general check to see what it would sound like, and if it would be worth the work to do right.
also, it's just meant to be an "interm," then second system, until i get wooden Eleven Horns - lower midrange horn.... 140 Hertz Le Cleac'h... and
Eleven Horns - upper midrange horn.... 600 Hertz Le Cleac'h... ,
(hopefully will be in this weekend),
to use with compression drivers.

the azura has a more rapid expansion ( 0.8 T factor), so is shorter than an oris (~38" x 17" / 950mm x 450mm); doesn't sound as "dark."
however, the azura throat is for a 8" Lowther, i just made an adaptor and bolted the 10" Tannoys on...
i've made a fiberglass mold though, and intend to experiment w/ different materials (glass, carbon fiber, kevlar), make some for Phil's field coil Lowthers, and another pair cut back shorter, so the throat will be larger, and match up properly to the Tannoys.

actual directivity-- easy enough to ck/match in hornresp, but i haven't even simmed anything for these yet (just the compression driver system).
mid-treble x-over is ~1800 on the Tannoys;
one of the cool features is the shape of the compression driver to cone interface, and that the shape of the cone curves/flares, (isn't just conical).

the mid-bass horn sims +/- 1 dB from 70 - 350Hz, -3 is ~60 & 500Hz (a little to high, i just adjusted the angle for now, will try a choke to roll it off a little sooner).
that was the only thing that has appeared as a problem (by ear, not worth running measurements yet, i'm tearing it down tomorrow).
actually, i would say the mid-bass is the best part of this system, (actually, IMHO it's probably the second best mid-bass i've ever heard).

i really was suprised how well this sounds, with so little work, (just adjusting angles).
don't know if that helps, pm me if you want more details--
again, this doesn't even belong on this forum...
 
Hi,

Would you please share more?

I've been thinking of horn loaded coaxial (maybe 8~12" with 1" compression driver), but it seems there'd be directivity problems which are hard to resolved.

At the typical xover point of such assembly - around 1kHz, the sound from the treble unit is largely guided by the cone of mid-bass unit, which is mostly very wide conical. So I'd imagine the waveform of the outer edge would hit the horn wall (near the throat). The treble sound leaves the smaller horn and then meet another bigger one. The transition between the 2 are not smooth :( So I think a horn or WG for such driver should be very wide and shallow, thus it can minimize or eliminate this issue. (while the loading and directivity control of the midbass unit are another 2 stories.... )

That is in my thinking, though. I don't know how it behaves in reality, and whether it's a real problem or something very small which can be ignored.

Hmmm......

You raise interesting questions that I think about as well.

Different units have different directivity, so choosing a unit with a narrower dispersion would be step one.

Secondly, the larger the woofer, the shorter the horn. The effect is that the horn walls are further away from the main "lobe" of the tweeter's response, so less reflections.

Thirdly, the part that I am not certain of. In a simplified assumption, the woofer and tweeter should have the same response at the crossover frequency. In balance, in phase and also dispersion pattern. If that is the case, then the two should act together as a single driver. If the woofer's interaction with the horn at (or just below) the crossover frequency was not a problem, it also won't be for the tweeter at (or just above) the crossover frequency. Directivity will only increase above that frequency. In other words, the problem might not be that big anyway.
 

CLS

Member
2005-06-17 6:58 am
Taiwan
Hi,

The 3rd point would be very difficult, I guess. For example, in Oris 150 (or Azura 160 might be similar in this regard), horn loading is effective up to 2kHz or so. If the xover point is set right at here or slightly higher, than the dispersion of the treble unit is mostly much wider than the midbass unit due to the source size.

According to the long Wave Guide thread by Geddes, the design guideline is controlling the directivity of the 1" CD to be exactly the same as the 12~15" midbass unit at and around the xover point which is near 800~900Hz range (that is about 90 degree).

So if the horn load for midbass is to be controlled under such frequency range, and also fulfill other needs (get out of the way of treble, and proper loading, directivity control of itself... ), that would be very difficult to design. Eventually, it might be an extremely wide and shallow one which is just like a slightly recessed baffle:(

Or, in the end, maybe this idea will inevitably evolve to the Unity horn concept. That'd be a special design coax with its treble unit extending to the entire horn, and the sound of midbass unit coming out through several passages around somewhere near the throat of treble unit.
 

Ivo

Member
2005-01-25 11:16 am
Amsterdam
The Unity horn is fascinating and when properly done I bet it's very good. I won't be able to do it properly...

Good points about the mismatch in directivity. An option then would be a smaller woofer, to push the point of beaming higher. Again, there are coaxials with a horn tweeter with a smaller directivity. I haven't looked extensively now, but I saw coaxials with a 70 degree disperion and I know I have seen 60 degrees as well. I don't remember at which frequency dispersion is controlled like that, which may not be as low as the proposed crossover frequency.

I have read accounts of people using Altec coaxials in their VOTT cabinets, which are frontloaded and rear reflexloaded. I did wonder how they were getting away with that and assumed they had narrower dispersion patterns and, possibly, the benefits of a fully hornloaded pointsource outweighed the negative effects of reflections inside the horn.

It is always hard to draw lines of sanity or efficiency in our hypothetical DIY discussions and I think this is where my experience falls short. I would love to hear from someone who tried his Altec coaxial in his VOTT or related horn and tell us what his impressions were.

I hope to try something like this out some day, it would be nice to know at least what gives a better chance of succes, a laaaarge driver or a small driver.
 

Ivo

Member
2005-01-25 11:16 am
Amsterdam
Fair enough, but my curiosity is how a large Oris-style horn with a coaxial could be made to work. I have heard a number of conventional systems and a a good number of far out designs. When tastefully done, they can all get very close to each other. My personal experience with a multiway compression driver system was very, very positive (probably my favorite so far).
 
Or, in the end, maybe this idea will inevitably evolve to the Unity horn concept. That'd be a special design coax with its treble unit extending to the entire horn, and the sound of midbass unit coming out through several passages around somewhere near the throat of treble unit.

This one has been used that way in Tom Danley's latest:
Overview

Sheldon