Sd is speaker displacement area, generally the diameter of the cone + about 1/2 of the surround diameter, the outer portion of the surround hardly moves so Sd is always a bit smaller than the cut-out area.
Atc should be the area equivalent to the diameter of the entire surround, a bit larger than Sd.
The HF rolloff will probably be less than the simulation shows (no cone breakup is modeled), but by all means compare different compression ratios and compression chamber size.
Attachments
Thanks again. The drawing was clarifying indeed. I had a wrong understanding of Sd.
New run, with adjusted Atc and Vtc. To reduce the latter from 200cm3 I envision a recessed fit to the horn. I need to draw it.
If the HF is likely to really roll off slower than simulated then I should be pretty good for 300-2500Hz. Right?
What is Hornresp assuming in terms of shape of horn mouth? Round? Does it matter if I want to model a rectangular Tractrix like Edgar's? I would imagine it does matter. But how can I reflect it in Hornresp?
Attachments
If the expansion is the same, some aspects of the sim will remain the same.. ie the throat impedance may, the polar response may not. Hornresp has a neat feature if you look at the schematic then go: File>Export>HornData, Rectangular horn, set the width flare, Preview - width/height
That guys loony tunes - avoid at all costs
Definitely crazy - he kept building stuff instead of talking about it...
This strange need diyers have to make phrases...
Btw : @Rewind ; did you have a look at old Fertin and the one they make now at EMS speaker, good reviews here and there, but pricy ! The old full range 8" from Fertin company was famous for its good sound, not sure it is the case yet...
Never see in my life a distorsion measurement (from independant source) of these french drivers, like the Supravox e.g.
Btw : @Rewind ; did you have a look at old Fertin and the one they make now at EMS speaker, good reviews here and there, but pricy ! The old full range 8" from Fertin company was famous for its good sound, not sure it is the case yet...
Never see in my life a distorsion measurement (from independant source) of these french drivers, like the Supravox e.g.
This strange need diyers have to make phrases...
Btw : @Rewind ; did you have a look at old Fertin and the one they make now at EMS speaker, good reviews here and there, but pricy ! The old full range 8" from Fertin company was famous for its good sound, not sure it is the case yet...
Never see in my life a distorsion measurement (from independant source) of these french drivers, like the Supravox e.g.
No, I have not, but I will look into it, thanks.
Thank you.
Any tricks to make the top end drop at a slower rate than what was depicted on my last post?
The lower end looks good to about 350Hz. Smoothness from 350 to about 1800Hz looks good too, around 105dB. I wish I were -3dB at 2000Hz.
Simulating the M5N12 vs 605Nd shows the latter achieving a higher sensitivity, but dropping faster at the higher end. I'm skewing towards M5N12.
Excellent. Indeed, dropping Atc from 113cm2 to 47cm2 (Sd/2) improved the higher end significantly. Beyond Sd/2 resonances seem to show up - I'm guessing they resonances...wiggles in the SPL response.
Off axis at the higher end is not looking too good. But it hasn't worsened when going from Atc 113 to 47cm2. I find this a little odd. A 5" direct radiator shouldn't have such a narrow dispersion at 2500Hz as this horn seems to have. 53° at 2500Hz...Puzzling.
Attachments
The horn would be constricting the beamwidth. Along its length where the horn becomes large enough the primary mode of the highs will continue with the angle as thus defined, and influenced to some degree further along including diffraction. Compare a beamwidth plot for tractrix, and a conical horn.
EQing a conical waveguide is acceptable/necessary, and the flip side of the narrow beamwidth for a tractrix horn may be that EQing for flat power makes it too hot on axis (not that I'm implying anything).
EQing a conical waveguide is acceptable/necessary, and the flip side of the narrow beamwidth for a tractrix horn may be that EQing for flat power makes it too hot on axis (not that I'm implying anything).
Stupid question of the day
I continue to read papers and play with Hornresp. I am not sure if I should be modelling using 2xPi or 4xPi for the radiation angle. I can rationalize 4xPi because the midrange horn will be far enough from any walls and floor...and can also rationalize 2xPi because at these frquencies the horn itself prevents any waves to go around the plane of the mouth of the horn...
Bottom line, I'm confused
And it has a significant impact on the simulation, per the attached.
I continue to read papers and play with Hornresp. I am not sure if I should be modelling using 2xPi or 4xPi for the radiation angle. I can rationalize 4xPi because the midrange horn will be far enough from any walls and floor...and can also rationalize 2xPi because at these frquencies the horn itself prevents any waves to go around the plane of the mouth of the horn...
Bottom line, I'm confused
And it has a significant impact on the simulation, per the attached.
Attachments
But it does not simulate that shape. After choosing your rectangular shape, the rest is guessing until you listen to it.
What you say is reasonable and it would probably lie between the two but as far as I choose to assume, the radiation space parameter applies a generic loading condition. I'd want to ask David to be sure.I continue to read papers and play with Hornresp. I am not sure if I should be modelling using 2xPi or 4xPi for the radiation angle. I can rationalize 4xPi because the midrange horn will be far enough from any walls and floor...and can also rationalize 2xPi because at these frquencies the horn itself prevents any waves to go around the plane of the mouth of the horn...
Bottom line, I'm confused
And it has a significant impact on the simulation, per the attached.
The way I see it I prefer to go with well defined terminations and boundary conditions although they aren't always essential. This may for eg. mean taking conical and hyp/ex horns to a baffle when appropriate, and running LeCleach horns far enough around, and using walls properly when corner/wall loading.
What works for the highs may be different to what works lower down. The natural termination space of some horns isn't clear, ie: a tractrix curve appears geometrically to end in a baffle but perhaps it should be as the LeCleach.
I prefer to examine each case at each frequency to decide how the mouth should be loaded. For example: by controlling the directivity index.
FWIW, I decided to buy Faital Pro M5N12-80. I should get them today.
Coming up next: break them in a bit, measure T/S parameters with REW (will likely need help there as well...), adjust horn design on Hornresp and build a first trial.
I might start a new thread to get input specifically on the horn design. This has been an outstanding learning process. Thanks all!!
Coming up next: break them in a bit, measure T/S parameters with REW (will likely need help there as well...), adjust horn design on Hornresp and build a first trial.
I might start a new thread to get input specifically on the horn design. This has been an outstanding learning process. Thanks all!!
An externally hosted image should be here but it was not working when we last tested it.
Just sent inquiry from 165H´s. not same horn but some similarities are shown. kind of feel that fst does not have strenght to keep up over 1k on my horn. Not sure, if it´s not 4months to wait and price isn´t too bad maybe i cultivate myself.
yesterday i measured el cheapos 20€ sica in that quido, that goes low