Google Translate
This has some interesting designs. You can view the finished concepts @the bottom of the page.
This has some interesting designs. You can view the finished concepts @the bottom of the page.
i agree an aperiodic TL or labyrinth enclosure would be the best at losing rear radiation. However, unlike many designs though, NO rear chamber and a large pipe area are probably good design goals, and maybe better than existing designs. Those that i am aware of anyway. reduce pressure in the pipe, and damp heavily. Seems like the best method to me.
And how is the driver aware of this unbalanced impedance curve? Both sides are drive by the same voice coil. The impedances can not be separated. They are combined in the same point in the equivalent circuit.
With regard to standing waves within a box (sealed or vented) vs. an OB design, the OB designs are far from resonance free. Front diffraction issues are effectively doubled by having the open back. Rear wall reflections are also increased. An H frame enclosure adds very significant depth related resonances and multiple cancelation nulls. A well stuffed box, on the other hand can easily have internal resonances lowered to the point where they are undetectable.
David S.
One of the advantages of a TL -- it sucks the back wave down the pipe where there is a much thicker blanket of damping to pass thru.
And one could always do it like the nautilus, and entire w/l of damping at thelowest frequencies.
dave
One of the things you learn when designing horns, the best load for the other side of the driver is an identical horn.
One would expect that varying the load on either side would affect 2nd order harmonics, whether positively or not depends on the design of the driver.
dave
TLs are not really the best at losing rear radiation.
Most TLs are designed to give some bass contribution. As such they can't be very well damped or the bass output will disappear. If you damp them well enough to lose the back wave then they are effectively a sealed box without any TL properties. (And might has well have been designed as sealed from the beginning.)
Don't confuse the ideal of a transmission line, a constant impedance conduit down which energy travels without end, and the typical damped resonant pipe we call a "transmission line".
David S.
An H-frame should be operated only below the first dipole peak, then the nulls are avoided. The first null of my own very large H-baffles happens at approx 250 Hz, which is not a problem for a woofer system.
The dipole peak is not a resonance that rings in the time domain like a standing wave, its just a response peak because the front and rear radiation sum in phase with each other.
Rear wall reflection cant be avoided in the bass range with dipoles. The same goes for regular closed or vented boxes. But - if the dipole is placed near the wall - in the pressure zone of the standing waves - it can not energize the standing waves in the room. That's one of the major advantages of dipole bass.
Member
Joined 2009
Paid Member
Perhaps a 'sealed TL' is in fact, a good approach for minimizing box colourations. It becomes an Infinite Baffle. It will still be more efficient than an OB for bass because there isn't any l.f. cancellation.
If I understand correctly, this helps to reduce uneven bass response in a room, but it isn't a 'box colouration' issue ?
StigErik said:
In the case of an H-frame, you have a pipe resonance due to the impedance mis-match where the H opens out into the room.
Creating a TL (in the audio sense of the word) is going to involve tweakng the levels of more than just the lowest peak. You need to let a certain amount through the line exit in order to achieve a smooth response. You can try to damp it out completely but you may as well just put more stuff in a closed box.
Last edited:
If using the speaker exclusively below the modes of resonance is allowed then all damping issues are moot. I can make a subwoofer with no cabinet stuffing at all if the first modes are above 100 (assuming my crossover is really steep). The question is what cabinet or damping approaches are best if we need to use the woofer box over a fairly wide range.
A standing wave is a poor counter example as it, like the OB example, is the response of two sources: the sound generator at one end and its far end reflection. In any case, we will see ringing in the time domain whether the cause is comb filter cancelation or mechanical system resonance.
For long wavelengths the dipole may have an advantage due to its lateral nulls. For higher frequencies (the discussion was about problems from internal cabinet resonances and how to cure them) a dipole will have considerably stronger back wall reflections. That seems like a poor trade off versus the choice of damping rear energy within the cabinet.
David S.
Once upon a time, the 80s, I got the impression that a TL gave me closed box benefits with LF extension to Fs. And flattened impedance to boot. Stupid? Lots of folks want to make TLs into vented boxes with long ports but they don't sound like that. Wrong? If I want a 300 hz passive x-o, what other LF enclosure can actually work? Yeah I'm a woodworker. Baffles don't bother me.
A heavily damped TL where the mouth is open would ensure that the energy that is not absorbed is released. Since this energy is a fraction of the back wave it should not not interfere too much with the front wave.
but eventually the remaining back wave energy needs to either absorbed or released right?
What if the other end is open but the output is damped to the point that the box behaves more like a lossy (Aperiodic) box instead of a sealed or TL?
Comb filtering from a single reflection or diffraction event will cause ringing in the time domain ? How ?
Smearing in the time domain yes, since some of the impulse energy arrives delayed in time, ongoing ringing at a spot frequency like a high Q resonance due to standing waves I can't see how though.
Last edited:
Comb filtering from a single reflection or diffraction event will cause ringing in the time domain ? How ?
Smearing in the time domain yes, since some of the impulse energy arrives delayed in time, ongoing ringing at a spot frequency like a high Q resonance due to standing waves I can't see how though.
True, a single reflection doesn't "ring" but does have a messy time response. We also know that cancellations due to reflections will be perceived as having a distinct pitch.
I tried some convolution in Cool Edit Pro with a single reflection nearly cancelling a 1k interupted tone and also a 5 reflection combo cancelling the same tone. In the second case the primary convolution impulse was 50% and then 10% impulses at 0.5, 1.5, 2.5, 3.5, and 4.5 mSec. (The half delays assured cancellation.)
Multiple reflections looked more like the tone ringing you would expect. Certainly back when magazines published tone burst measurements we saw a lot of long term ringing.
I'll have to think about the reason for that.
David
Attachments
You can't really have it both ways.
The original statement was that a TL would be best at absorbing the potentially damaging rear wave from the driver. This is not the case.
If you want significant bass contribution from a TL then you will have some delayed midrange output as well. Damp it heavily to lose the upper range output and the box might as well be sealed. (It effectively is.)
As to closed end TLs, they are really just sealed box enclosures with particular dimensions. These can be problematic also, if the cabinet is especially long in one dimension the low frequency mode will need special effort in damping. For example, tall narrow tower speakers, especially if the woofer is at one end, will have a strong length resonance that can be difficult to damp. In a cabinet with a constant lining thickness the bottom mode for the long dimension will give problems.
In that case a squat box, even with all dimensions equal, can give better results.
David
It is the case. THe purpose of the damping in the line is to dissipate the unwanted higher harmonics of the line. If you don't you get some huge ripple in the FR as the terminus output moves in & out of phase. So how much of the back wave is absorbed (converted to heat) by the damping can be clearly seen as the (not) ripple in the FR.
dave
Except this is not the case. There is a threshold where upper range output or harmonic ripple as planet10 refered to, is well attenuated, while still providing around 3db of response boost over the intended passband. can you get more output with a vent or less stuffing? sure.
But to say it's "effectively sealed" is unfair. Once the lower impedance peak has been broadened you've effectively found a good middle point between a vent and sealed - that is the TQWT - more output for its passband than sealed, and less output and consequently midrange contribution than ported.
Member
Joined 2009
Paid Member
Seems to me, that if you replace "TQWT" with "Aperiodic" in your paragraph above it still looks perfectly correct.
The arguement is getting a bit circular but remember the starting assertion was that the best cabinet design in terms of absorbing all secondary radiation was the transmission line because "energy goes down the transmission line never to return". Except a labyrinth cabinet is not a transmission line, it is a damped tuned pipe.
Every simulation and measurement I have seen shows that for 2 to 3 dB of bass output you will have to accept 1 to 2 dB of upper range ripple. I'm not knocking TL speakers, the ripple is a modest cost to pay for some free bass output and isn't likely to be visible in a real room curve.
Still, if people are loosing sleep over what happens to the back wave of a loudspeaker, a lightly stuffed sealed box is clearly better than a typical TL design.
As to "effectively sealed" look at the Augspurger paper. All his curves show the woofer only output as well as the combined output. With low damping you can approach 6dB greater bass output from the combination of line and woofer but the ripple will be unacceptable. To get to moderate ripple, say 2dB, the line output must drop to 10 to 15 dB below the woofer output. For any better than that you will need high line damping and the output will be primarily from the woofer, i.e. the cabinet becomes effectivley sealed.
David
Every simulation and measurement I have seen shows that for 2 to 3 dB of bass output you will have to accept 1 to 2 dB of upper range ripple. I'm not knocking TL speakers, the ripple is a modest cost to pay for some free bass output and isn't likely to be visible in a real room curve.
Still, if people are loosing sleep over what happens to the back wave of a loudspeaker, a lightly stuffed sealed box is clearly better than a typical TL design.
As to "effectively sealed" look at the Augspurger paper. All his curves show the woofer only output as well as the combined output. With low damping you can approach 6dB greater bass output from the combination of line and woofer but the ripple will be unacceptable. To get to moderate ripple, say 2dB, the line output must drop to 10 to 15 dB below the woofer output. For any better than that you will need high line damping and the output will be primarily from the woofer, i.e. the cabinet becomes effectivley sealed.
David