|
15NW100 BC Horn |
Post Reply 
|
Page <12 |
| Author | |
KaphaSound 
Registered User 
Joined: 22 July 2020 Status: Offline Points: 102 |
 Post Options
 Thanks(0)
 Quote Reply
 Posted: 29 December 2020 at 6:43pm |
Very curious about this!
|
|
 |
|
|
kipman725
Registered User
Joined: 02 September 2020 Location: Warrington Status: Offline Points: 231 |
Post Options
Thanks(0)
Quote Reply
Posted: 29 December 2020 at 10:47pm |
|
Looks great, certainly easier movement splitting the horn in 4. |
|
|
|
|
citizensc
Young Croc 
Joined: 16 October 2015 Location: Perth,Australia Status: Offline Points: 532 |
Post Options
Thanks(0)
Quote Reply
Posted: 30 December 2020 at 1:13am |
Just to clear things up a little, Danley is talking about legacy AKABAK, I have been using AKABAK 3 which is completely different software. Legacy AKABAK is a very old program, so old that it doesn't run on modern computers but lots of people still use it. AKABAK 3 is for BEM modelling which is far more computationally demanding, simulations can take hours or even days to run on modern computers if they are very complex. The basics to AKABAK 3 are 1. Getting your geometry in to the software, if you are feeling masochistic you could draw the whole horn in AKABAK 3 by setting coordinates and creating planes from those coordinates. I think the more sane method is to draw it in external software and import the drawing in to akabak. Make sure the drawing is as minimal as possible, I only include geometry that matters acoustically, eg. I don't draw internal cavities in detail. 2. You then have to define a lump element network which basically defines the electromechanical properties of the driver. 3. You then have to set up all the measurement stuff, which is basically like setting up a mic array in the software. The learning curve to all this is a bit steep but once you get over it you start to realise you can sim pretty much anything you want in the software, including the room the speaker is in. It took me hours to get my first sim working but now I can build one from scratch in about 15 min. I have been thinking about making a little tutorial series that covers the basics needed to get started, it wouldn't go in depth as I don't feel qualified to do that, I may mislead people. If I do, it wont be soon, I have been super busy since getting out of lockdown and moving back to my home city, so much so that I haven't had time to prototype my unity horn despite having the components :( Ps. The HF1440 is HUGE!! I have big oafish hands and it makes them look tiny.
 |
|
|
|
|
KaphaSound
Registered User
Joined: 22 July 2020 Status: Offline Points: 102 |
Post Options
Thanks(0)
Quote Reply
Posted: 30 December 2020 at 2:06am |
|
A tutorial sounds awesome already these points were very helpful! A synergy horn would be next on my list of builds but I think that could be some time down the line. I’ll take a look and see if I can’t figure out a bit more and give it a go, but if the local mill gets back and can CNC/mitre everything for me I may just pull the trigger we’ll see!
Edited by KaphaSound - 30 December 2020 at 10:49pm |
|
|
|
|
KaphaSound
Registered User
Joined: 22 July 2020 Status: Offline Points: 102 |
Post Options
Thanks(0)
Quote Reply
Posted: 30 April 2021 at 9:47pm |
|
Still working on this, looking to get the first cab built this coming month (subject to wood costs and CNC work) and would appreciate any last input! The nature of this project is a bit experimental but again after seeing Citizensc's BG horn modifications and AKABAK 3/hornresp models I suspect the end result shouldn't be too far off. One real question is still how well the horn path will couple with the face and if we can expect to see much improvement in sensitivity, as well as what directivity might look like. Especially curious if we had 8 of these lined up what that would mean for directivity.
I modeled moving the volume of the cone around a bit and not much changed, but I just want to confirm S1 should start half-way across the driver, then the volume shaded in yellow here including the cone volume should be added to vtc?  Here's some possible stacking options, although I think the whole point of this design is to really use the perpendicular mouth for the low-frequency extension. I suppose you could use this as a seriously overpowered kick bin in the conventional orientation?  Here’s the inputs incase anyone sees anything that looks off:  Can’t say enough how much help this community has been. Started learning everything probably just before the start of the pandemic and I'm really hoping to build the first box and see/hear/feel the results. I’m also hoping this design will be able to keep up with the newer paraflex designs...
Edited by KaphaSound - 01 May 2021 at 9:39am |
|
|
|
|
DMorison
Old Croc 
Joined: 14 March 2007 Location: Aberdeen Status: Offline Points: 1649 |
Post Options
Thanks(0)
Quote Reply
Posted: 01 May 2021 at 1:38pm |
|
I haven't kept up with the BC and stacking parts of this, but for modelling the throat area Hornresp specifically has the Offset Driver option for this kind of layout - S1 becomes the section at the very end of the horn and S2 that at the entry point of the driver.
This does use up a horn section though, so you may need to condense your current s2-s4 into one section. This doesn't always make much difference, but I'd try it both ways just to be sure.
|
|
|
|
|
snowflake
Old Croc
Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3121 |
Post Options
Thanks(0)
Quote Reply
Posted: 01 May 2021 at 5:17pm |
|
it depends what sort of throat you use - letterbox, v-slot, full Sd cutout etc
|
|
|
|
|
KaphaSound
Registered User
Joined: 22 July 2020 Status: Offline Points: 102 |
Post Options
Thanks(0)
Quote Reply
Posted: 01 May 2021 at 8:03pm |
|
Using full sd cutout, and I originally did model with the offset driver without the perpendicular exit then moved the volume of s1-s2 over to vtc freeing up S5 and that modeled the same.
Condensing S2-S4 to keep offset driver mode with the perpendicular exit of S5 did seem to make a decent difference in the sim but I would imagine that’s to be expected since that would be quite a drastic change in the expansion rate throughout the mid section? This particular horn seems to benefit a lot from a very slow expansion rate to start really until S3 to S4 (the traditional end of the horn).
Edited by KaphaSound - 01 May 2021 at 8:14pm |
|
|
|
|
KaphaSound
Registered User
Joined: 22 July 2020 Status: Offline Points: 102 |
Post Options
Thanks(0)
Quote Reply
Posted: 30 November 2021 at 11:03pm |
|
Still keen on creating this but would really appreciate one last look through the inputs and design before I go get wood cut. Here's the inputs where VTC encompasses the volume of the driver cone + volume of the cutout + volume of what s1+s2+s3 would be in the OD arrangement:
 And here's another look at the design where VTC is highlighted. After reading through previous posts including citizensc's 15BG horn I believe this should be the correct approach in order to free up the additional segments:  The last "gamble" on this design that I'm thinking is really S4 - S5. Currently I have S4 - S5 as half the width of the cab (24.1cm), but I've experimented with slightly more and slightly less and each scenario is pretty close. This is also assuming that the path touches S4 at the bottom left then moves out directly towards you, but I've experimented with a shorter path and smaller S4 area and it would also take quite a difference to see drastic changes. If there's any thoughts here I'm all ears!
Edited by KaphaSound - 30 November 2021 at 11:18pm |
|
|
|
|
toastyghost
The 10,000 Points Club 
Joined: 09 January 2007 Location: Manchester Status: Offline Points: 10919 |
Post Options
Thanks(0)
Quote Reply
Posted: 01 December 2021 at 12:56am |
|
Unfortunately Hornresp won't model anything beyond the designated mouth S5, and the use of an axisymmetric condition means that it doesn't really represent the three exits you have. Results for a single horn in of this configuration might be quite far off.
You also don't get to choose the 'stacking arrangement' for the multiple horns option in Hornresp. It doesn't really make a stack, the segment areas are simply multiplied by a factor to produce a single horn of equivalent volume as the individual horn paths of a stack. The issue is the lack of any simulation of directivity, diffraction and the like. Once you get to 'big horns' the model starts to become much less valid. I'd treat the design as a single box and see how it compares to other horns of similar size and volume if you're limited to Hornresp models. The prediction of throat impedance and therefore magnitude response on-axis is reliable and good, if the source mouth radius is small compared to the wavelength of the intended operation range. If you did want to play with the new Akabak, you can build an equivalent model to Hornresp (with unlimited elements in series or parallel) in the LEM section, and couple the 'mouth' element to a mesh of the cutout in a model of just the external cabinet for the BEM part. That way, you can ignore all of the internal stuff for now, and get some results that will inform any further development. |
|
|
|
|
KaphaSound
Registered User
Joined: 22 July 2020 Status: Offline Points: 102 |
Post Options
Thanks(0)
Quote Reply
Posted: 01 December 2021 at 3:03am |
|
Has Akabak been updated recently or are there any tutorials you’d recommend? I know it’s much more powerful but I haven’t been able to piece together all the steps quite yet. Having said that though you can see citizensc’s previous Akabak work on this same idea here: https://forum.speakerplans.com/bc-horns-from-standard-flhs_topic105665_page1.html
As far as orientation it is still a bit of a guess with respect to directivity but the idea would be to always use the smaller exit for the lower extension. You could consider it a corner horn as well since the corner would cover all exits but S5.
|
|
|
|
|
kipman725
Registered User
Joined: 02 September 2020 Location: Warrington Status: Offline Points: 231 |
Post Options
Thanks(0)
Quote Reply
Posted: 01 December 2021 at 11:50am |
|
Akabak 3 is almost totally different software to the classic Akabak which causes great confusion. Akabak 3 includes a BEM simulator and an LEM simulator which can be coupled. Whereas Akabak was only an LEM simulator. Hornresp is also an LEM simulator but with a more limited interface than Akabak which makes it easier to use but also means some very complex boxes can't be simulated. The software you want to simulate a BC horn is Akabak 3 as you need BEM simulation to take into account the mouth shape and coupling of box mouths. The linked thread uses Akabak 3. There are some video tutorials around for Akabak 3: (I haven't used this one just came up in search) I would recommend following some to get an idea of what it does and the interface. Also the program includes some examples. I wouldn't say its the easiest thing to learn but its well worth learning. The only limitation of the free version is that you can't save results so have to re-solve the model every time you open the file and want to look at the results. |
|
|
|
|
Post Reply
|
Page <12 |
| Tweet |
| Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |
Topic Options
snowflake wrote: