Here's the influence of adding various lengths of a straight duct segment to a 1" throat, zero exit angle.
Obviously, as a result, it 1) increases efficiency at LF, 2) decreases efficiency in the middle of the passband, 3) adds some FR ripple, not nearly as pronounced with the real devices, however - this can be vital. Note the directivity remains virtually unchanged.
Effects 1 and 2 together cause the driver's response to become more extended and flat - that's the only explanation I have. Perhaps different drivers require different lenghts of the extension. Anyway, starting the duct right at the exit of a phase plug, I think this can be a promising way how to proceed, after all.
0 mm
25 mm
50 mm
75 mm
Obviously, as a result, it 1) increases efficiency at LF, 2) decreases efficiency in the middle of the passband, 3) adds some FR ripple, not nearly as pronounced with the real devices, however - this can be vital. Note the directivity remains virtually unchanged.
Effects 1 and 2 together cause the driver's response to become more extended and flat - that's the only explanation I have. Perhaps different drivers require different lenghts of the extension. Anyway, starting the duct right at the exit of a phase plug, I think this can be a promising way how to proceed, after all.
0 mm
25 mm
50 mm
75 mm
Last edited:
Do you have some impulse response or ETC or something with time domain visible to wonder is that resonance with the extension? I wonder if sound gets better or worse, or just different, assuming frequency responses with or without duct were equalized with DSP?🙂 what if the input driving modes are manipulated, does effects of the extension change, like ill effects on top octave(s)? Basically, where is the trade-off?
https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7431530
https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7431507
It's the ATHEX 460-36 with a 6° extension 101 mm long (to get a 1" throat at the end).
https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7431507
It's the ATHEX 460-36 with a 6° extension 101 mm long (to get a 1" throat at the end).
The above was for k=1. This is for k=3 (the 75 mm extension).
Basically it makes the transition less abrupt, the impedance less wavy (lower-Q peaks) with all the features shifted lower in frequency. The downside is the HF beaming (which would be the same without the extension, that's just the result of a higher k).
- With a tubular duct it's even not necessary to have any advanced model for that - one can easily try different lengths.
And I assume a piece of open-cell foam inside the duct wouldn't hurt either.
Basically it makes the transition less abrupt, the impedance less wavy (lower-Q peaks) with all the features shifted lower in frequency. The downside is the HF beaming (which would be the same without the extension, that's just the result of a higher k).
- With a tubular duct it's even not necessary to have any advanced model for that - one can easily try different lengths.
And I assume a piece of open-cell foam inside the duct wouldn't hurt either.
Last edited:
As for the time domain, it seems that all of this is still minimum-phase phenomena, so it should be possible to EQ easily, as without the extension. I tried to find any excess group delay effects but found none. Maybe someone else will have a better eye, or an approach.
Probably worth a try.
Is it possible to simulate with accuracy?
The 0mm above seems best as the impedance trace is flat!? Do I miss something?
//
The 0mm above seems best as the impedance trace is flat!? Do I miss something?
//
The swing from negative to positive in the imaginary part is very much the same as what happens in a horn with a defined cutoff.
Yeah, that's what a sharp-enough peak/change will always do, as the real and imaginary parts are actually tied together by some (mathematical) transform. You could calculate one from the other.
Last edited:
You get some help to go lower with the driver but once that swing happens the rolloff is steeper with the associated phase change. As you say all very predictable.
Interestingly, the effect is not at all dramatic with the real drivers, as was shown. It's still a nice and smooth roll-off.
Probably the Q is actually much lower (the peaks broader and the imaginary swing also lower) than what the simple model predicts.
Probably the Q is actually much lower (the peaks broader and the imaginary swing also lower) than what the simple model predicts.
The 75 mm duct with its inner boundary damped (Damping = 0.05 and 0.1):
(I don't claim this is anywhere near the real thing, but I believe this impedance curve could be closer to truth...)
(I don't claim this is anywhere near the real thing, but I believe this impedance curve could be closer to truth...)
Hi Marcel,
could you repeat the simulation with conical extension, e.g., 1.0 in to 1.4/1.5 in?
Kindest regards,
M
could you repeat the simulation with conical extension, e.g., 1.0 in to 1.4/1.5 in?
Kindest regards,
M
Straight vs conical (0 vs 8 degrees), 1" throat, 75 mm extension:
With the conical it's basically a 1.4" throat. To retain the directivity of a 1" throat, it needs to be tubular, maybe requiring further dampening (?).
This would need to be actually tested, I guess, with different lengths of the tube.
With the conical it's basically a 1.4" throat. To retain the directivity of a 1" throat, it needs to be tubular, maybe requiring further dampening (?).
This would need to be actually tested, I guess, with different lengths of the tube.
Last edited:
How about the opposite?
Conical adapter like 2”-1”. We know from TL speakers that this can be made shorter for the same tuning, and the higher harmonics gets damped somewhat.
And as long as we talking home audio, there shouldn’t be any significant downside to trying to squeeze the sound thru a small throat.
Conical adapter like 2”-1”. We know from TL speakers that this can be made shorter for the same tuning, and the higher harmonics gets damped somewhat.
And as long as we talking home audio, there shouldn’t be any significant downside to trying to squeeze the sound thru a small throat.
After the latest experience, I wouldn't dismiss anything beforehand...
Last edited:
I had trouble telling if you tried this with the K=3 example, but would it help to change the throat angle to 0 deg to better match to the straight duct?