Hornresp

Everything n the internet is true!

Or NOt???

That is why I ask you guys as experienced people.

I did see that read a real book about horns is the best way, special the old ones, constant directivity horns are also old, for me I see what happens.

now I have a tractrix from X design,, however it beams to much, I need some more spreading, so need to reed the right stuff.

Directivity and Imaging


thanks.

kees
 

Attachments

  • tractrix1.jpg
    tractrix1.jpg
    261.5 KB · Views: 264
Last edited:
Hi Kees,

I think that the author of the Blog is trying to convince us that conical horns are the best because that's what his company makes and sells :).

Apparently he is not too concerned about the lack of low frequency performance compared to say an equivalent-sized exponential horn.

Incidentally, he refers to Edgar "Villachur" when the name is actually "Villchur"...

Remember - if something sounds too good to be true, then it probably is. If conical horns were as good as claimed then every manufacturer would be using them.

Kind regards,

David

The tractrix from X I have experience with do not bad, however it has a peak because of the ports and the tractrix curve, however a 24 dB filter do wel correct that.

I have to say, it is sim with a exp flare in hornresp, it can be different with real world, see picture two of real measurement..

The book do explane that it never the way se want, so the conic is busted, I do not now the guy and that he do sell them however making such blogs is not very well for bussines, I like to be honest, and so I can always do promisses who are real. But nice book. thanks.

DIY is just experiment, the new things we need for perfection.
 

Attachments

  • ScreenHunter_ Aug. 06 19.12.jpg
    ScreenHunter_ Aug. 06 19.12.jpg
    54 KB · Views: 246
  • tractrixwoofer.jpg
    tractrixwoofer.jpg
    71.4 KB · Views: 250
As all books abou the fysics of audio, it are all compromisses we deal with, conical is for the home maybe a option to get wider image but when I use a K-tube in the tractrix it is better because these do radiate wide and is a way to prevent beaming in that area and I keep the cone loading intact for low distortions, what concerns speaker, it is not for nothing that in al these years there are still not yet invented the best system, conical horns do load the speaker very slowly, white a tractrix or a exponentional horn do load it mucho lower, and is better for the bass, a conical subwoofer do bad, so a combination of horns on the places where she are oke is the best way then, conical constant directional horns even have a exponential part in throat for speaker loading.

The synergy is a special kind, because it is a injected one, and the compression driver do not beam much because of the form.

Nice stuff, but I like amps more, looks less complicated.

David

When I export a txt file for a synergy and want a 60 x 90 horn, is that possible in hornresp or do I need the synergycalc.xls file for making one to test? I do use sketchup for cutting plan.

Thanks.

regards
 
When I export a txt file for a synergy and want a 60 x 90 horn, is that possible in hornresp or do I need the synergycalc.xls file for making one to test?

Hi Kees,

If we assume a rectangular cross-section conical horn where:

H1 = Throat height
W1 = Throat width

H2 = Mouth height
W2 = Mouth width

L12 = Axial length

HA = Height included flare angle
WA = Width included flare angle

Then:

H2 = 2 * L12 * Tan(HA / 2) + H1
W2 = 2 * L12 * Tan(WA / 2) + W1

Numerical example:

H1 = 3 cm
W1 = 3 cm
L12 = 30 cm

HA = 60 degrees
WA = 90 degrees

Then:

H2 = 2 * 30 * Tan(30) + 3 = 37.64
W2 = 2 * 30 * Tan(45) + 3 = 63.00

S2 = 37.64 * 63.00 = 2371.32

The horn having the required 60 x 90 flare would therefore be specified in Hornresp as:

S1 = 9.00
S2 = 2371.32
L12 (Con) = 30.00

The data should be exported as shown in Attachment 1 to give the required 60 degree height profile as shown in Attachment 2, and the required 90 degree width profile as shown in Attachment 3.

Kind regards,

David
 

Attachments

  • Attach_1.png
    Attach_1.png
    31 KB · Views: 218
  • Attach_2.png
    Attach_2.png
    41.3 KB · Views: 78
  • Attach_3.png
    Attach_3.png
    42 KB · Views: 76
I'm trying to learn how to use hornresp to model a transmission line. I think I understand the basic principles but I'm not 100% sure I'm translating a physical design into a model correctly. I imagine the information I need has been posted before but I haven't found it so far so it would be appreciated if someone can cast an eye over this.

I'm going to use a published design to keep this simple, the Woden Labyrinth Lance which is on p2 of http://wodendesign.com/downloads/Woden-BabyLabs-081015.pdf

This has the following dimensions

schematic.png

- rear chamber is 101.6mm wide, 76.3mm deep
- driver centre line is 82.6mm down
- rear chamber extends 165.5mm from the driver centre line down to the base of the cabinet
- opening into the TL is 50.8mm high and 101.6mm wide (i.e. full width of the box)
- TL extends back to the height of the box as that same depth (i.e. 50.8mm) so another 247.7mm
- TL narrows to 25.4mm and extends back down to the bottom of the box

So, in terms of hornresp segments this seems like there are 4 segments

1) top of the rear chamber down to the centre of the driver
2) centre of the driver down to the bottom of the chamber
3) 1st section of the TL
4) 2nd section of the TL

In hornresp schematic terms, I visualise it like this

real.png

I think my basic question is that I don't know how to transition from 1 segment to another given that the physical model has abrupt steps in the cross sectional area where hornresp doesn't seem to do this. What's the right way to model this?

My other question is that this seems like it models the driver as firing into the TL but in reality it is the rear of the speaker firing into it, does this matter?

For reference, my attempt at doing this in hornresp

hornresp.png

zero out all the chambers; Vtc = Atc = Ap = Lpt = Vrc = Lrc = 0

1st segment = top of box down to centre of driver
S1 = width of rear chamber * depth = 10.16*7.63 = 77.52
S2 = S1
Con = 8.26

2nd segment = centre of driver to bottom of box
S3 = S2
Con = 16.51

3rd segment = 1st section of TL
S3 = ??? (I don't know how to "step" down in size )
S4 = 10.16 * 5.08 = 51.6128
Con = 24.77 + 1.27 = 26.04 (height of box + width of wood dividing the line from the rear chamber)

4th segment = 2nd section of TL leading to the output
S4 = ??? (I don't know how to "step" down in size )
S5 = 10.16 * 2.54 = 25.8064
Con = 24.77 + 1.27 + 1.27 = 27.31 (height of box + width of wood dividing the line from the 1st section + width of wood to the exit)

Cheers
Matt
 
Last edited:
I think my basic question is that I don't know how to transition from 1 segment to another given that the physical model has abrupt steps in the cross sectional area where hornresp doesn't seem to do this. What's the right way to model this?

You can model the discontinuities with a very short segment e.g. .01 cm. but then you'll run out of segments to do the model. You'd probably be better off building it with full ramps instead of steps.


My other question is that this seems like it models the driver as firing into the TL but in reality it is the rear of the speaker firing into it, does this matter?

I can't imagine that it matters; especially not with only low frequencies coming out the end of the TL. When you get done with the model, I'd love to see how it compares to a bass reflex. Will it be worth the trouble?
 
With regard to abrupt cross sectional area transitions in Hornresp, nc535 is correct, the only way it can be done is with a very short (.01 cm) segment transitioning from one cross sectional area to the next cross sectional area, and he is also correct that this is going to use up too many segments to simulate this design in Hornresp.

But you can simulate it quickly and easily in tl.app at Leonard Audio | Audio Engineering Resources
Arguably this program is even easier to learn to use than Hornresp. Or you can use Akabak which is lot harder to learn than Hornresp.

One more thing - when simulating the design as you did it, it should be PAR segments, not CON. If it were simulated as stepped (as it's supposed to be) it wouldn't matter if the segments were CON PAR or EXP, the results would be the same. But with the tapered segments that you simulated it would make a difference and PAR is the correct way to do it, assuming you are using two parallel side walls as the original design plans show.
 
Hi Matt,

I think my basic question is that I don't know how to transition from 1 segment to another given that the physical model has abrupt steps in the cross sectional area where hornresp doesn't seem to do this. What's the right way to model this?

The easiest way is to simply average the areas on either side of the steps. The predicted results will be close enough for all practical purposes.

Using your figures, the Hornresp horn data inputs then become:

S1 = 77.52
S2 = 77.52
S3 = (77.52 + 51.61) / 2 = 64.57
S4 = (51.61 + 25.81) / 2 = 38.71
S5 = 25.81

L12 = 8.26
L23 = 16.51
L34 = 26.04
L45 = 27.31

As 'just a guy' has mentioned, it would be theoretically more correct to use Par rather Con segments, however the dimensions of your design are such that I doubt it would make any noticeable difference to the results.

My other question is that this seems like it models the driver as firing into the TL but in reality it is the rear of the speaker firing into it, does this matter?

No.

Kind regards,

David
 
Thanks for the replies. I tried the Leonard Audio software and compared to hornresp, using David's suggested approach, and the modelled FR (for the lance) that I could find (http://www.diyaudio.com/forums/full...3-box-designs-baby-labs-more.html#post3753912) and they seem to be in pretty close agreement (hornresp has a somewhat softer knee but I guess this is the differences between how damping is applied/calculated as you enter the values in different ways in each one).

leonard.png

hornresp.png
 
Hi Kees,

If we assume a rectangular cross-section conical horn where:

H1 = Throat height
W1 = Throat width

H2 = Mouth height
W2 = Mouth width

L12 = Axial length

HA = Height included flare angle
WA = Width included flare angle

Then:

H2 = 2 * L12 * Tan(HA / 2) + H1
W2 = 2 * L12 * Tan(WA / 2) + W1

Numerical example:

H1 = 3 cm
W1 = 3 cm
L12 = 30 cm

HA = 60 degrees
WA = 90 degrees

Then:

H2 = 2 * 30 * Tan(30) + 3 = 37.64
W2 = 2 * 30 * Tan(45) + 3 = 63.00

S2 = 37.64 * 63.00 = 2371.32

The horn having the required 60 x 90 flare would therefore be specified in Hornresp as:

S1 = 9.00
S2 = 2371.32
L12 (Con) = 30.00

The data should be exported as shown in Attachment 1 to give the required 60 degree height profile as shown in Attachment 2, and the required 90 degree width profile as shown in Attachment 3.

Kind regards,

David

:hypno1:

Thanks David

I go on with it to try.

Or is this program a option? it calculates for me and can input that on hornresp.

Here I can also see the dimensions needed. I also need the extra flare as seen on attachment.


Kees
 

Attachments

  • ScreenHunter_ Aug. 11 11.44.jpg
    ScreenHunter_ Aug. 11 11.44.jpg
    269.8 KB · Views: 205
Last edited:
It is not a Hornresp option - you will have to do it for yourself :).

Thanks David.

I did hear that keeping the horn square it is better for directivity, like for in a home and as high end design.

This program do only put out the hornresp data for simulation I did read, after that go back and design with the program itselfs, it give dimensions of the horn, I think I go this way it is more easy because I am not a math boy (stargate universe):D
 
Hornresp Update 3970 -160814

Hi Everyone,

Three significant changes have been made to the Multiple Entry Horn Loudspeaker Wizard tool, along with a number of other minor alterations. The major changes are:

CHANGE 1

Individual power response chart results were not being shown correctly, even though the total power result was correct. This has now been fixed.

CHANGE 2

The number of entry point options available has been doubled from 5 to 10, the maximum theoretically possible for a system having up to four segments with either one or two offset driver entry points.

CHANGE 3

The S4 slider in a four segment system now has the same Manual / Auto feature as provided on the S2 and S3 sliders.

Kind regards,

David
 

Attachments

  • Attach_1.png
    Attach_1.png
    54.2 KB · Views: 191
I will enjoy trying this out!

Hi Mark,

Have fun :).

Five minutes playing with the Multiple Entry Horn Loudspeaker Wizard tool will show just how difficult it can be to design a good Synergy-type horn system. It becomes very obvious indeed that the multiple inputs interact in quite complex ways, and that not unexpectedly, the overall performance is very sensitive to the positioning of the entry ports.

Kind regards,

David
 
need help with the Multiple Entry tool

I'm had trouble too, just trying to get to the Multiple Entry Horn Wizard, despite reading the help carefully and trying to do exactly what it says. I have a more or less working synergy horn that I'm fine tuning and a simulation of the mids on it. I took that model and followed the instructions for a 2-way horn, creating this Nd record, which simulated as expected:

oml0fp-UMv4FZkXw4kTYeBxyjGtyT95zPN9vwnUpxDY_pKpUI71LRN0vSxDF4MwtgnN8sX8Q-nHT54E9T7c58aPpCHYmJyz1pinI9LF8E_0IhEgFSyoB5scUun6p7oXBGST0CCS5xs48xsct-E1ZdCT5ASI01w9wTeyrbjpKXAIQ4yTptl8nhkuWtv0_LUJgvuR24_TpMNrejMmojIE6XRfo15NgSK2z2cbRJPElZ0wIY1XaDLrMwjX2Lt_YlYej-z7f6hqfh2suo-ujLaPXnt4Gyoux5VAoMDzrIH4tRC5VwgU9uZI79pjY6vhX8guYODEQhZ_T9k7CnpxEYfgzz56AbeZtqz6fyZy_FXlIwphC1vAxpXkIc67lPcj4DaMAIIW9QqCIu0zn5EUQhDM546TfXC4wGccmg29mYbAs6LUnJReDeZPs79_hZ6Y2IEaSJDNrblOEThwzIgMW6HiSTY3sPC5yhkbtBfZ1Pfw3PPF6-ZGyG4aOYYStkPZ7DJgGguIOL1cJy8RiZqYdEG2SeuvSI3SsVTFyLPFNggMo5U8PwVyCJVIa5a0ElIg-2vZ1IkKf4YzC88fyUpVCaIrntAZFOy8Ag6w=w919-h691-no


The next thing was to create an ME1 record for the mids. That is the part I thought I had trouble with.

What I did was to double click on Nd on Input Parameters screen until OD appeared and then went into the driver arrangement tool, specified 4 mids in series parallel, and selected ME1. I then got back to the input parameters screen, entered my mid driver and champer parameters, and pressed "activate" and got this:

xOt6UnlOdSaK8VeYFNlJHMfWzPKMS0ycwrYk9EwCOgPZNr3w55DPehs8UyASjjCKlkSA2Vx_-cxZCcU9bM7sSIB5fsdveOYKzsHp-49DY6z5_0ww4F7PbQhhbMFoNcM3p5530eQMoTpy8iKmJVxy7eTv86CVU4mSrxU3_KFEs32nhLQFxKkYUY7U714tJ0iG4ZrDGr6HDdYnITUw_CdcbEiBaE5KadzvPOT0k8J0tXoA79TlM8o1fqf6nJfYFdzE4f8_-bT7FK-iEc-HNNzSjVivAe8DmYQYjppRTH17-jfAHZC9DKfnrQ5ynbbqjFbFm2VQhAR3lfSJfTYk8SPOMzT_z9Xic6KgOKwKaJ9HVSPruhdEwhSOjLQIOSnNI1UGgB_qIFKx8uva-2zknwlcfXtRhTNwBgaCHsZgevMYpLKXveBzPCOaI0yPJciPAdc87s0lyqStENrgzkjbj149H-3VdjMigxf-TDrc2Wd2bWJz0faNxIr8kS9NMdo0RgAMnrJc8ZFdte5wcqEDPumgX0FRR6WYAizBzLqsgWvrziwZ8CdRHIw41GwImSuH1oe5xugvrHkm4CtXirXfxO2lciNxRm_HCX8=w1407-h1058-no


At this point, I was going to ask for help because I still didn't see the Multiple Entry Wizard in the tools menu, but I persisted and finally got it. Fortunately, after I entered the data for the Nd record, I had pressed "Add" on the bottom of the Input Parameters screen because I didn't want to enter it yet again. This created a separate file record for the Nd section, which is needed.

Then I entered the data for the ME1 record and activated it as described above. To see the Multiple Entry Wizard in Tools, I had to press "Prev" on the bottom of the Input Params screen to get back to the Nd record. Pressing Prev gives you the option to save, which I believe is required.

With the wizard up you can look at a number of things. The mid simulation is pretty much as before and as measured, after adjusting params for a match, but I have no filters defined.

In the physical world, my simulation has a number of PEQs to flatten the mid and CD response in addition to LR4 high pass and low pass. I use just 2 PEQs on the mids, a few more above 2 KHz... The jury is still out on whether I'll use any delay. (eventually I will use FIR filters) If you're finding it difficult to get a good response, it might be you are not allowing enough filtering options.

Thanks, David.

Jack
 
Hi Jack,

I'm had trouble too, just trying to get to the Multiple Entry Horn Wizard, despite reading the help carefully and trying to do exactly what it says.

Note that there are two sections in the Help file to read:

1. Multiple Entry Horn Wizard (under the TOOLS MENU heading).
2. Two-Way or Three-Way Multiple Entry Horn (under the LOUDSPEAKER MODELS heading).

To simulate a two-way system it is necessary to create two separate records. A Nd record is required to specify the complete horn system and the HF driver at the throat. A ME1 record is required to specify the offset MF drivers and associated chambers and ports. To create a ME1 record, simply add a new record and then either Ctrl+double-click on the Nd (or other) label in edit mode until ME1 is shown, or directly select the ME1 option using the Driver Arrangement tool.

Once the Nd and ME1 records have both been created, click the Activate button on the ME1 record, and then switch to the Nd record and select the Multiple Entry Horn Wizard tool.

At this point, I was going to ask for help because I still didn't see the Multiple Entry Wizard in the tools menu,

The Multiple Entry Horn Wizard tool can only be accessed from a Nd record, and only after a ME1 record has been activated.

Kind regards,

David
 
nice feature tnx david
ive not yet been able to get it working,but i,m sure il find out in time :)

Hi epa,

It's all pretty straight-forward, once you get the hang of it :).

The key points to remember are:

1. Two records are required for a two-way system (Nd and ME1).
2. Three records are required for a three-way system (Nd, ME1 and ME2).
3. The Multiple Entry Horn Wizard tool can be accessed from the Nd record after the ME1 record has been activated (or after the ME1 and ME2 records have been activated in a three-way system).

Kind regards,

David