Please could someone check this LEM script if this is enough to define a direct radiating driver (like the 5" Faital above)? I'm still a bit helpless.
Code:
Def_Driving
Value=2.83 IsRms
Def_Driver 'Drv1'
dD=103mm
Mms=10.5g
Cms=0.57e-3m/N
Rms=0.52Ns/m
Bl=6.9Tm
Re=5.4ohm
fre=35kHz ExpoRe=1
Le=0.1mH ExpoLe=0.618
System 'S1'
Driver 'D1' Def='Drv1' Node=1=0=10=20
// Rear volume
Enclosure 'Eb' Node=20
Vb=5e-3m3 Qb/fo=0.1
RadImp 'Throat' Node=10 DrvGroup=1001
From ATH - Advanced Transition Horns : "...expression for a very general horn profile that smoothly terminates into a flat baffle...."
Is this really true with a roll-back?
//
Is this really true with a roll-back?
//
The whole sentence is "The above formula provides a simple closed-form expression for a very general horn profile that smoothly terminates into a flat baffle" and that is true. Rollback is not integrated into the formula (can't be). It would be nice to have some simple parametric expression for the curve including the rollback but I doubt it's possible.
Last edited:
Why not take a driver without a conical exit section?
Attachments
Last edited:
With a suitable 1.4" driver with an open enough throat extension one could get the HF directivity of a 1" driver but with the low frequency extension of a 1.4" driver. At least that's the idea. The same principle holds for any driver size. Also the loading would be a little better with the decreased throat angle.
Driver without a conical exit section doesn't offer this possibility.
Driver without a conical exit section doesn't offer this possibility.
Last edited:
Dunno, but here are some extra plots.
Attachments
This is the Le Script that I used for a model of the 15PR400
Looks pretty much the same on a quick view. Watch the length of the rear chamber, I found leaving it out was the better option.
Code:
//*****************************************************************************
//
// ABEC3 Lumped Element File
// Project:sealed woofer
//
//*****************************************************************************
// Thiele-Small parameter of electro-dynamic driver
Def_Driver '15PR400' // 15inch paper woofer
Re=5.1ohm // datasheet
Le=0.75mH // datasheet
Mms=56.8g // datasheet
Cms=0.24e-3m/N // datasheet
Bl=16.7Tm // datasheet
Qes=0.34 // datasheet
Qms=6.0 // datasheet
Rms=3.11Ns/m // use HornResp to derived
Fs=35Hz // datasheet
//SD=857cm2 // datasheet
//fre=3.1kHz //
//ExpoRe=0.824 //
ExpoLe=0.3 //
Def_Driving "Driving"
DrvValue=2.83V // assumption is gnd=node "0" and output is node="1"
IsRms
// Lumped element network
System 'Woofer'
//Filter 'WooferLR4LP' f0=700Hz V0=1 Order=4 LR Lowpass; // implicit input
Driver 'Drv1'
Def='15PR400'
Node=2=0=111=112 // connected woofer to LP filter, lowest node# is input
// DrvGroup=1011,1012 // front,rear of woofer 2 port
DrvGroup=1011 // front only,
RadImp 'WooferFront' // rad imped, a different object
Node=111 // connect drive source node
DrvGroup=1011 // the DrvGroup links the LE to BE
// RadImp 'WooferRear' // rad imped, a different object
// Node=112
// DrvGroup=1012
Enclosure "WooferRearChamber" // using a lumped compliance model with damping
Node=112 // connect to drive source node
Vb=60L // rear chamber volume
//Lb=10cm // rear chamber length
etaB=2.7e-2 // medium dampingLooks pretty much the same on a quick view. Watch the length of the rear chamber, I found leaving it out was the better option.
The benefit of having both woofer and waveguide being axisymmetric and of similar sizes is that their directivities will be similar and give a good polar match both horizontally and vertically.
Slot loading the woofer will create a bandpass chamber that will cause a peak in the response before it rolls off much earlier than normal. The long thin exit will not be a good polar match at 500Hz or above which are the likely crossover points, the effect can be simulated in Hornresp or with greater precision in ABEC/Akabak.
Edit: Seems like the post I responded to vanished.
maybe you could even add a bandpass design to create an aggressive roll-off
that would mean that the drivers harmonic distortion would be suppressed. could be a good trick to make it sound even cleaner...
This probably also depends on the type of diaphragm and the specific design of the phase plug.
The outer slit of a 3 slit annular phaseplug may, or may not interfere with a throat insert.
The same may be true for some ring radiators in which the wave front is directed past a centrally located plug.
Attachments
Last edited:
Pure OS throat (k=1) most probably wouldn't fit to any extension, for any reasonable coverage angle. The whole point is to start a curvature-continuous WG profile at a smaller throat. Then even for k>1 the performace would be better than a pure OS throat starting at the exit of the extension. At least this stems from the simulations. Starting at a smaller throat is always an improvement. Making the profile smoother (without the curvature change) is another improvement.