Hello everyone. I am fairly new to both ABEC3 and AKABAK. Would anyone be able to give me some guidance in modelling a horn?. I have a compression driver with me. I need to model a Horn/Waveguide for it. However, I do not have some essential driver parameters. My plan is to measure them.
*Should I model the driver in Akabak first?
*Is there a way to import the frequency response of my driver to ABEC?
*Is there a better way to model the horn for my driver?
Please excuse me if my questions are not clear. I can explain in detail. Any help would be greatly appreciated.
*Should I model the driver in Akabak first?
*Is there a way to import the frequency response of my driver to ABEC?
*Is there a better way to model the horn for my driver?
Please excuse me if my questions are not clear. I can explain in detail. Any help would be greatly appreciated.
If you are new to Akabak, I would suggest you not try to model a compression driver as it is very complex and not just putting in TS parameters like a cone driver or even dome tweeter. There are phase plug flow passages and the system has to be modeled in great detail. There is a nice model of a typical 1in D220 compression driver that most people use (modified) from the basic one given in the manual. It does a pretty good job of approximating most 1in drivers in a horn. Your specific driver will be somewhat different depending on native response curve but it shows you the typical sensitivity you can expect.
Do a search for a Synergy Akabak script - you will find the code for modeling the D220. Use that as a baseline.
Akabak does not import FRD responses to manipulate like some XO simulators. Akabak calculates the response from theory and equations rather than take a measured response to manipulate. You can plot externally imported data in your Akabak graph though, but for plotting I like to import FRD or TXT file curves into REW.
There maybe some one who knows how to model a compression driver ab initio in Akabak but I have not seen any examples. They would all require having either a dissected compression driver or having engineering CAD drawings to know channels and passages exactly. Then you need to know motor parameters to put into model of the gyrator. If you just look at the example in the manual you can see how complex this is.
Do a search for a Synergy Akabak script - you will find the code for modeling the D220. Use that as a baseline.
Akabak does not import FRD responses to manipulate like some XO simulators. Akabak calculates the response from theory and equations rather than take a measured response to manipulate. You can plot externally imported data in your Akabak graph though, but for plotting I like to import FRD or TXT file curves into REW.
There maybe some one who knows how to model a compression driver ab initio in Akabak but I have not seen any examples. They would all require having either a dissected compression driver or having engineering CAD drawings to know channels and passages exactly. Then you need to know motor parameters to put into model of the gyrator. If you just look at the example in the manual you can see how complex this is.
Last edited:
Suggestions
Dear xrk971,
Thank you for this information. Yes, I have been using such a script as a baseline. However, I do not know if Akabak can be used to design a well suited horn for the driver I have. What is your opinion on using ABEC to do that? If ABEC is a good option, do you know if the Akabak script for the driver can be used in ABEC without any hassles? My efforts to do so resulted in several errors popping up. Most of them were 'un-recognized parameters' like annular ring depth 'Ha',etc. Thank you for spending time to clear my doubts.
Dear xrk971,
Thank you for this information. Yes, I have been using such a script as a baseline. However, I do not know if Akabak can be used to design a well suited horn for the driver I have. What is your opinion on using ABEC to do that? If ABEC is a good option, do you know if the Akabak script for the driver can be used in ABEC without any hassles? My efforts to do so resulted in several errors popping up. Most of them were 'un-recognized parameters' like annular ring depth 'Ha',etc. Thank you for spending time to clear my doubts.
Akabak works well for 98% of horn designs. Only if you are looking at 3d effects on horn profile shape and how that impacts polar patterns will you need to use ABEC3. If you have a profile you want to simulate break it up into segments to approximate the curve. Since tractrix curve was not built in, I simulated it with many segments. There is a copy of my script for the tractrix synergy or Trynergy in the 4th post of this thread.
http://www.diyaudio.com/forums/full...ing-trynergy-full-range-tractrix-synergy.html
ABEC3 uses the Akabak engine for its solver so they should be compatible. I do not use ABEC3 yet so cannot help you on specifics. All I can say is try their example and modify it from there for your application. But if you are new, start with Akabak first - it is 1d and easier to use.
http://www.diyaudio.com/forums/full...ing-trynergy-full-range-tractrix-synergy.html
ABEC3 uses the Akabak engine for its solver so they should be compatible. I do not use ABEC3 yet so cannot help you on specifics. All I can say is try their example and modify it from there for your application. But if you are new, start with Akabak first - it is 1d and easier to use.
Because you are using a compression driver, I assume that you want to simulate a midrange horn. This really requires the use of an isophase wavefront model. AkAbak uses a plane wavefront model. Approximating the profile of a midrange horn using a series of plane wavefront elements will not give a particularly accurate prediction of performance.
What flare profile did you have in mind?
Assuming that the compression driver is up to the task, a Le Cléac'h horn with cutoff frequency fc = 320Hz and T = 0.8 would be a good place to start. If off-axis performance is particularly important then an oblate spheroidal waveguide could be considered as an alternative.
If implemented correctly, the overall performance of a Le Cléac'h horn is likely to be superior to that of either a tractrix horn or a spherical wave horn with the same fc.
Akabak noob here - has anyone successfully/accurately simulated a downward-firing subwoofer? I'm thinking of a box with the woofer (and vent(s) if any) on the bottom surface, supported ~100mm from the floor by a leg near each corner.
My assumption is that that means I need to couple the front of the driver (and ports) to a Horn or Waveguide to represent the space under the box but my question is, what values do I use for the throat areas?
My brief reading of the Radiation section of the manual seems to indicate that the Baffle model implies a half-space whereas what I have is basically some slots (under each edge of the box), each radiating into quarter space. How do I model that, Reflectors?
My assumption is that that means I need to couple the front of the driver (and ports) to a Horn or Waveguide to represent the space under the box but my question is, what values do I use for the throat areas?
My brief reading of the Radiation section of the manual seems to indicate that the Baffle model implies a half-space whereas what I have is basically some slots (under each edge of the box), each radiating into quarter space. How do I model that, Reflectors?
HornResp would be an easier simulation and easier to learn if you don't already know it.
Then, if not satisfied with HR, you can export an Akabak model from it and then add more detail to the simulation in Akabak. But I think the HR sim would be sufficient if you are only interested in subwoofer frequencies...
Then, if not satisfied with HR, you can export an Akabak model from it and then add more detail to the simulation in Akabak. But I think the HR sim would be sufficient if you are only interested in subwoofer frequencies...
Not worried about learning more akabak, just looking for some guidance on the physical structure of the simulation I should use in order to best represent this kind of cabinet.
My main areas of concern are:
- presumably I should make the circumference*height of the gap around the driver at least as large as the piston-area of the driver
- modeling the radiation from the slots into quarter-space
I could do the former as 4 parallel ducts perhaps, each representing 1/4 of the cone and 1/4 of its circumference? But they have a 90-degree bend.
The latter, I could model as 90-degree horns? Radiation summing is going to be weird though as they are not coplanar and are not even facing the same direction. At the 15-50Hz I care about, the directivity will be meaningless/zero; I'm just trying to make sure the radiation resistance is about right, because that will (I assume) affect the box tuning.
My main areas of concern are:
- presumably I should make the circumference*height of the gap around the driver at least as large as the piston-area of the driver
- modeling the radiation from the slots into quarter-space
I could do the former as 4 parallel ducts perhaps, each representing 1/4 of the cone and 1/4 of its circumference? But they have a 90-degree bend.
The latter, I could model as 90-degree horns? Radiation summing is going to be weird though as they are not coplanar and are not even facing the same direction. At the 15-50Hz I care about, the directivity will be meaningless/zero; I'm just trying to make sure the radiation resistance is about right, because that will (I assume) affect the box tuning.
Let me then try to answer the first question you asked.
I think its more of a bandpass chamber than a horn.
I would model the volume of air between the cabinet and the floor as a throat chamber and the horn S1 and S2 as equal to the circumference of the cabinet times its height off the floor. L12 would be very small. A throat chamber without a horn. Easy to predict the effect will be an HF roll off but not until well beyond 50 Hz.
I think its more of a bandpass chamber than a horn.
I would model the volume of air between the cabinet and the floor as a throat chamber and the horn S1 and S2 as equal to the circumference of the cabinet times its height off the floor. L12 would be very small. A throat chamber without a horn. Easy to predict the effect will be an HF roll off but not until well beyond 50 Hz.
In fact, more than reasonable. This is the first time that I have a simulation that gets close to the published, measured response of the Shiva downfiring vented boxes. The volume under the box does actually have quite an appreciable effect on the tuning of the box.
This is my script:
Yes I am somewhat living in the past. I have 3 Shiva Mk.II sitting on the shelf that are looking for cabinets, and recently got access to a CNC router.
This is my script:
Code:
Def_Driver 'Shiva Parallel'
SD=481cm2
fs=21Hz
Cms=0.47e-3m/N
Rms=2.54Ns/m
Bl=10.8Tm
Re=2.9ohm
Le=2.1mH
ExpoLe=0.618
Def_Const 'Con'
{
VOL=142.5e-3; | box volume
LEN=55.0e-2; | box height
SS=481.0e-4; | Radiator area
PORTD=10.0e-2; | Port diameter
PORTL=43e-2; | Port length
| Shiva whitepaper says physical port len is
| 12" plus 2.5" flares or 17" for straight pipe
| not sure if I have the k-correction set correctly (one free end, one flush)
SDBOTTOM=0.203; | area of slot around bottom (3.5x22.5" x4)
VOLBOTTOM=9e-3; | volume under the driver
THROAT=0.1; | horn throat area = circumference*height of VOLBOTTOM
HORNWIDTH=2.286; | 4x22.5"
HORNHEIGHT=8.89e-2; | 3.5"
HORNLEN=10e-2;
}
Driver Def='Shiva Parallel'
Node=1=0=2=3
| main box volume
Enclosure 'E1'
Node=2
Vb={VOL}
LB={LEN}
| the port
Duct 'Port1'
Node=2=3
dD={PORTD}
Len={PORTL}
| air volume under the driver+port
| treat it as a 14x3.5" cylinder: 9L volume, 9.9e-2 output area
Enclosure 'E2'
Node=3
Vb={VOLBOTTOM} | result seems insensitive to this (good!)
| model rest of space under box as a horn, less the 9L under the drivers
| input area is 0.0993m^2 -> diameter 35cm
Horn 'Bottom'
Node=3
STh={THROAT}
WMo={HORNWIDTH}
HMo={HORNHEIGHT}
Len={HORNLEN} | result seems insensitive to this (good!)
Conical
x=0 y=0 z=0 HAngle=0 VAngle=0
| insensitivity to the E2 volume / horn length indicates that the arbitrary choice of boundary be$
| under-driver volume and horn throat is not important. Net result is very similar to using a
| horn with throat chamber of the full volume and no depthYes I am somewhat living in the past. I have 3 Shiva Mk.II sitting on the shelf that are looking for cabinets, and recently got access to a CNC router.
Last edited:
Another question - does anyone know if Akabak has a batch or command-line mode? I want to generate a bunch of input scripts, feed them to akabak one by one, and read output files containing simulation results.
The reason is that I would like to be able to wrap akabak in an optimisation algorithm, e.g. sweep through a range of a couple of parameters (say, driver position in a tapped horn) and plot the variation in response due to that parameter. Ultimately, use akabak as the fitness function for a genetic optimiser or something like that.
Anyone done similar?
The reason is that I would like to be able to wrap akabak in an optimisation algorithm, e.g. sweep through a range of a couple of parameters (say, driver position in a tapped horn) and plot the variation in response due to that parameter. Ultimately, use akabak as the fitness function for a genetic optimiser or something like that.
Anyone done similar?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.