I admit, I really like the look and the form factor of the Altec 816A cabinet, I am just trying to tailor something more specific to my needs.
I haven't played around much with hornresp other than very simple examples. I am really at a loss to see how the BR portion of the response can be up to the level of the horn loaded portion. That has always seemed puzzling.
Like Art mentioned earlier, the space between the two pathways at the mouth causes frequency response ripples. A single exit mouth does not suffer from these issues. See POST #134
It usually won't be, the response will usually be "stepped" down.
The room response may make up for some of the step.
You can generally go with a 3/1 compression ratio or even more with a relatively low Bl speaker like the Alpha 8.
Remember, horn wall angle on a conical horn determines the upper beamwidth, that ultra long horn would be -6dB at about 10 degrees.
Ultra beam.
That horn is trying to get too much bandwidth from the driver, with a FLH don't bother trying to get the Fc (horn cutoff) below Fs.
I agree with JLH that the folding in post # 159 (looks like the EV MTL4 and MTL 2 type designs) is good, probably one of the best compromises if one is planning a separate HF horn. The compression chambers usually should be smaller than shown.
I am using Alpha 8 in an offset horn design with the acoustic crossover around 1000 Hz, and Alpha 6" in a straight horn/BR with a 1" slot with the acoustic crossover at around 3500 Hz.
Art
My apologies for beeing so stuborn 🙂 I grasp the problem with the delay / combfiltering caused by the pathlenght difference of the Jubilee horns double exits.
I ment the "red" reflector like thing in the picture i quoted in post #155. In Hornresp loading two 12 inchers sideways as in the drawing above, leaves an offset to the hornbeginning of about 5 inches. This offset causes a notch above 6 a 800Hz, when compared to the solution where the drivers fire "bifurcated" into the horn and then combine to one horn.
The "red" reflector could be used to decrease the amount of offset for the drivers. But would this still count as 1 fold? Since the driver is 90 degrees of horn-axis? Hmm, i am probably not making myself much clearier. In the above picture, do both horns have the same number of folds?
An externally hosted image should be here but it was not working when we last tested it.
I ment the "red" reflector like thing in the picture i quoted in post #155. In Hornresp loading two 12 inchers sideways as in the drawing above, leaves an offset to the hornbeginning of about 5 inches. This offset causes a notch above 6 a 800Hz, when compared to the solution where the drivers fire "bifurcated" into the horn and then combine to one horn.
The "red" reflector could be used to decrease the amount of offset for the drivers. But would this still count as 1 fold? Since the driver is 90 degrees of horn-axis? Hmm, i am probably not making myself much clearier. In the above picture, do both horns have the same number of folds?
Last edited:
The long horn idea came from one of Patrick Bateman's threads, where you and he had discussed that a horn with a narrower dispersion would have a smoother frequency response, so I made that my 'first port of call'.
I've been playing around a bit again, after your advice in relation to the compression ratio, and came up with some more promising results...
I was hoping to get to 1kHz, if I can. The HF horn goes down to 1kHz, so maybe I don't need to - possibly 800, and then 1.6kHz using a 3dB per octave crossover?
By straight, do you mean a horn that isn't folded or bent, or the horn itself has straight walls (conical)?
Straight = Not folded or bent. Straight path to the throat -
Conical doesn't load so well. For good loading exponential works. You can piece together conical sections along the expo profile and get many of the benefits with less complex build
Conical doesn't load so well. For good loading exponential works. You can piece together conical sections along the expo profile and get many of the benefits with less complex build
Thanks for clarifying...my head's been spinning on this subject a bit!
So, I've done it, I've ordered two of the Eminence Alpha 8a speakers, and I'll start with those as single driver horns! Hornresp seems to show a promising response within the band I need...at least until I can get the second pair and make those dual driver enclosures! 🙂
So, I've done it, I've ordered two of the Eminence Alpha 8a speakers, and I'll start with those as single driver horns! Hornresp seems to show a promising response within the band I need...at least until I can get the second pair and make those dual driver enclosures! 🙂
Hi unaHm,
The ripples in the response are due to the impedance mismatch at the horn mouth, which for the given design has a relatively small cross-sectional area. You will find that masked and unmasked results are quite similar, for the example shown.
Kind regards,
David
Remind us again of what the goals are for this horn and I can provide you with a Horn Response simulation. For lower frequencies I've always found Horn Response to be very close to real measured performance.
It’s not the path length differences (both bifurcated paths are the same length), it’s the flat panel separation between the two exits which result in reflections due to mouth diffraction. This will result in frequency ripples equal to one wave length of the panel dimensions (both height and width).
Loading two woofers in a horn with your “red” reflector does not result in any kind of offset. It behaves as a single pathway as far as the horn is concerned. A bifurcated horn is one that splits at the throat and does not recombine until the mouth. It is the distance between the two mouths (exits) that causes problems. What you have with that “red” reflector horn is simply a split throat horn that immediately combines after the throat entrance.
The “red” reflector horn I would count as a single fold horn. This is because you are bending the sound at least 90 degrees and redirecting it down the horn. If you can’t physically fit a straight (unfolded) horn in your room, then the “red” reflector horn is the next most benign thing you can do with sound.
I would recommend you go to the volvotreter website and read both the “Show Horn” and “Monolith” horn articles by Bruce Edgar. They are located in the down load section: Volvotreter Download Section
Also see: Huygens Fresnel principle - Wikipedia, the free encyclopedia
Pay attention in particular to Huygen’s wavefront reconstruction principle mentioned in the second page of the Monolith article. Design your reflectors in accordance with this idea in mind and the high frequency performance should not suffer much from the bend. One thing I would add is try to maintain the area of expansion through the bend while following Huygen’s principle at the same time. My post #159 kind of shows how to do that.
Dear JLH,
Thanks for the very thorough answer to my question. I have read both Showhorn and Monolith article's and will read them again until i fully grasp these finer details about Huygens and the wavefront. Thanks for beeing so patient with me and my "red" reflector 🙂
unaHm, good luck with your Eminence adventure!
Thanks for the very thorough answer to my question. I have read both Showhorn and Monolith article's and will read them again until i fully grasp these finer details about Huygens and the wavefront. Thanks for beeing so patient with me and my "red" reflector 🙂
unaHm, good luck with your Eminence adventure!
Last edited:
Thanks JLH, that's very kind of you! I'd like to get a usable range of 150Hz to 1kHz.
I think Bill Fitzmaurice noted in his T18 subwoofer design that the ideal crossover for it was 80Hz, but it's quite flexible, and has quite a flat frequency response up to 250Hz. I know the Alpha 8a only has a resonant frequency of 73Hz, so 150 still seemed to make sense, with it being an octave up.
Do you plan to run this horn on a tube amp, or solid state amp? How do you plan to wire the woofers - in series, parallel, or one woofer per amp channel?
- Status
- Not open for further replies.