Most brands are found here, or can be sourced. I mentioned JBL CMCD drivers because they seem to be a natural choice, but unfortunately those in particular are not available.
It would be good to know what is an ideal driver for what I'm trying to achieve.
I had so much grief with Norton that I switched to the AVAST! freeware ages ago with no issues except for it initially putting HR in its 'sandbox' and asked if I really wanted to open it, but after that, no issues since then through many years of renewals and moving to several 'new' computers since then.
GM
Right, but it 'wants' a smaller one than just sealing up the basket, which I've done before, so need to add plugs to reduce the volume inside the rear, so way too much hassle for me or even suggest.
GM
Hi LewinskiH01,
It really depends upon the dimensions and geometry of the throat chamber. The chamber I specified should hopefully work okay over the frequency range of interest without the need for a phase plug, as path length differences from any point on the driver diaphragm to the horn throat do not appear to exceed about 1/16th of a wavelength at 2500 hertz.
If by "underestimate sensitivity drop at the higher end" he means that the pressure response predicted by Hornresp will be higher than that actually measured, then I am not sure that I would agree.
Kind regards,
David
It really depends upon the dimensions and geometry of the throat chamber. The chamber I specified should hopefully work okay over the frequency range of interest without the need for a phase plug, as path length differences from any point on the driver diaphragm to the horn throat do not appear to exceed about 1/16th of a wavelength at 2500 hertz.
If by "underestimate sensitivity drop at the higher end" he means that the pressure response predicted by Hornresp will be higher than that actually measured, then I am not sure that I would agree.
Kind regards,
David
Hi David McBean
Hornresp is a fantastic program, cudos to you!
My listening room is rather narrow (315 cm) and I am thinking of making a midbasshorn which has two straight sides: to the floor and to the sidewall, thus making these becoming a "Mirror" for the horn mouth. The top side and outher side I plan to have in a hypex/jmlc-curve, with a 40/60-split dimension wise (=e.g. if the hight is 80 cm, then the Width will be something like 120 cm)
Is it possible to calculate such a horn in Hornresp?
The horn will be used in the 4-500 to 100 hz range, and I will be using a JBL 2220H driver. I will possibly be using a phase plug, as this will make it easier to have two parallell, straight sides along the floor and side wall. (And it will also enhance upper range frequency response if I am correct).
I have looked into an Excel-spread sheet made by Le Cleach (quasi cylindrical horn) for Stand bass Horn, but that spread sheet assumes parallell top and bottom sides, while I would like to have a curved top side for better (?) SQ in the top range of the horn, and that will also make it possibly for me to sit closer to the horn mouth (abaout 2.4 meters).
Best regards
Gisle
Hornresp is a fantastic program, cudos to you!
My listening room is rather narrow (315 cm) and I am thinking of making a midbasshorn which has two straight sides: to the floor and to the sidewall, thus making these becoming a "Mirror" for the horn mouth. The top side and outher side I plan to have in a hypex/jmlc-curve, with a 40/60-split dimension wise (=e.g. if the hight is 80 cm, then the Width will be something like 120 cm)
Is it possible to calculate such a horn in Hornresp?
The horn will be used in the 4-500 to 100 hz range, and I will be using a JBL 2220H driver. I will possibly be using a phase plug, as this will make it easier to have two parallell, straight sides along the floor and side wall. (And it will also enhance upper range frequency response if I am correct).
I have looked into an Excel-spread sheet made by Le Cleach (quasi cylindrical horn) for Stand bass Horn, but that spread sheet assumes parallell top and bottom sides, while I would like to have a curved top side for better (?) SQ in the top range of the horn, and that will also make it possibly for me to sit closer to the horn mouth (abaout 2.4 meters).
Best regards
Gisle
Hi Gisle,
Hornresp can export details of cross-sectional areas at progressive distances along the axial length of a Le Cléac'h horn. Once you have the exported data for your horn, make a scale plot of the curve shape you would like for the top side of the horn to determine the heights at the same points that Hornresp uses for the cross-sectional areas, then divide the cross-sectional area by the height to get the width at that point. This should give you what you want.
Kind regards,
David
What are best practices as far as geometry go? Should the edge of the throat be radiused, chamfer, or 90°?
BTW, in your example you used 100cm3 as Vtc. My estimation of Vtc using the Hornresp utility and online info about the M5N12-80 came up to be around 200cm3. Was your figure based on better information than mine? Obviously reducing Vtc improves the response.
I believe he meant the other way around: in my example, if Hornresp is showing 100dB at 2500Hz, that in reality I would likely measure higher than 100dB at 2500Hz.
The original Bell Labs/W.E. 1.4" exit driver, which the Altec/GPA is its modern day replacement, was designed for the 300-3 kHz decade, though admittedly it takes a very large horn to use it this low in a PA app and as the need for even greater acoustic power became required they initially used 2-4 on a manifold, then the 'bean counters' had their say, so the engineers shifted the XO point to 500 Hz, but technically it's the most correct design for the BW AFAIK without hand crafting one starting with a larger driver that terminates at a 1.4" throat.
GM
GM
Last edited:
Thank you very much, David
Hi Gisle,
Not a problem
.If you are happy to use the data as generated by Hornresp, then it would not even be necessary to make a scale plot of the curve shape you would like for the top side of the horn, thus simplifying the overall task.
If you care to provide the following information, I can quickly produce the dimensional data you would need to construct the horn.
* Horn throat width in cm
* Horn throat height in cm
* Either horn mouth area in cm2, or horn axial length in cm
* Mouth aspect ratio height : width (for example 1 : 1.5 if H = 80 cm and W = 120 cm)
* Horn cutoff frequency
* Horn T
Kind regards,
David
Hi LewinskiH01,
Smoothed edges are always good practice, but I doubt that you would hear the difference in the frequency range of interest.
I kept the throat chamber as small as I thought reasonable because the chamber acts as a low pass filter - the larger the chamber the lower the rolloff frequency. Also I wanted to minimise destructive interference effects due to different path lengths from points on the diaphragm to the horn throat, so that hopefully a phase plug would prove not to be required. Setting Atc = Sd and using a volume of 100cm3 gave a chamber length of just over 1 cm, which seemed about right to me.
In that case, he got his logic wrong.
Perhaps he is confusing power response with pressure response? Pressure response results are dependent upon the directivity model used in the simulations. The pressure response will be higher than the power response at upper frequencies due to "beam width narrowing".
Kind regards,
David
Smoothed edges are always good practice, but I doubt that you would hear the difference in the frequency range of interest
.I kept the throat chamber as small as I thought reasonable because the chamber acts as a low pass filter - the larger the chamber the lower the rolloff frequency. Also I wanted to minimise destructive interference effects due to different path lengths from points on the diaphragm to the horn throat, so that hopefully a phase plug would prove not to be required. Setting Atc = Sd and using a volume of 100cm3 gave a chamber length of just over 1 cm, which seemed about right to me.
In that case, he got his logic wrong
.Perhaps he is confusing power response with pressure response? Pressure response results are dependent upon the directivity model used in the simulations. The pressure response will be higher than the power response at upper frequencies due to "beam width narrowing".
Kind regards,
David
Hi Gisle,
Not a problem
If you are happy to use the data as generated by Hornresp, then it would not even be necessary to make a scale plot of the curve shape you would like for the top side of the horn, thus simplifying the overall task.
If you care to provide the following information, I can quickly produce the dimensional data you would need to construct the horn.
* Horn throat width in cm
* Horn throat height in cm
* Either horn mouth area in cm2, or horn axial length in cm
* Mouth aspect ratio height : width (for example 1 : 1.5 if H = 80 cm and W = 120 cm)
* Horn cutoff frequency
* Horn T
Kind regards,
David
You are very very generous, David!
I will send you an email/pm as soon as I have gathered my thoughts.
Best regards
Gisle
I'm not sure what driver you are referring to here. The 555W was used in the 300Hz-3kHz range in the W.E. Widerange system, but has an exit diameter of approx. 18mm. It was originally used down to ~60hz (on 11-14' long horns). Typically 1-4 drivers were used on these horns, but several horns were used, both for power handling and to get the required coverage.
The W.E. 594A has an exit diameter just short of 2", and was based on the HF driver of the Fletcher system, which had a slightly smaller exit. Intended frequency range was 300Hz-10kHz. 1-2 drivers where used on multicellular horns.
The first 1.4" driver is, I believe, the Lansing 284, designed for the Shearer system. The Altec/GPA 288/299 drivers are the modern replacements. The Shearer system had a crossover frequency of 250Hz, and was intended to cover up to 8kHz.
That's today's history lesson
-Bjørn
That's today's history lesson
Hi Bjørn,
It seems that your research visits to the AT&T Archives were days well spent
.Kind regards,
David
It seems that your research visits to the AT&T Archives were days well spent
Definitely, although most of the information in my post above comes from published papers
-Bjørn
Hi Bjørn,
It seems that your research visits to the AT&T Archives were days well spent
Kind regards,
David
That would be an interesting place to get into and spend a couple of weeks.
I know a couple of fellows who worked in the Canadian version of that enterprise.
Definitely, although most of the information in my post above comes from published papers
-Bjørn
Hello Bjørn.
I really enjoyed reading your AudioXpress articles. I also looked into your Thesis as I'm particularly interested in rectangular horns to reduce the center-to-center distance between midrange and tweeter. Maybe I missed it, but the thesis seemed to focus more on differences between MPM and BERIM calculations for rectangular horns rather than the analysis of those horns themselves. Yet you of course have analyzed rectangular horns and experimented with models (as shown on your website).
Is there anything I can read to learn towards what I'm hoping to design? My goal is 350-2500Hz from a rectangular DIY horn driven by a cone. Pretty much like the Edgar Tractrix, but benefiting from 30+ years since it. Or maybe I simply need to drop the idea and stick with a round horn, be it Tractrix or LeCleac'h.
I found Volvotreter put together a spreadsheet to use for rectangular horns and in conjunction with Hornresp. It's a good starting point but I'm wondering how the Hornresp simulation will deviate since it assumes a circular cross-section.
Congrats on your work!