When large bandwidth is desired, a conventional front loaded horn may be a better alternative.
A similar "extended low-bass shelf" response may be obtained from an undersized rear-loaded horn with the right choice of sealed chamber volume and throat area for the driver employed. The main difference is much smoother response above 80Hz with no comb filtering.
A similar "extended low-bass shelf" response may be obtained from an undersized rear-loaded horn with the right choice of sealed chamber volume and throat area for the driver employed. The main difference is much smoother response above 80Hz with no comb filtering.
3 oktaves are a max.
http://www.lenardaudio.com/education/07_horns.html
FLH bass horn corner horn, no crossover, 3 m distance
http://www.lenardaudio.com/education/07_horns.html
FLH bass horn corner horn, no crossover, 3 m distance
MikeHunt79 said:
There's nothing to stop you from mixing alignments. IIRC, Geddes at one point had sealed, vented and bandpass speakers all in the same room, covering different ranges. You could use a sealed box for the upper bass, where efficiency is less important, and a TH for the sub-bass, where efficiency is hard to come by.
3-way with horns is already a headache when it comes to get proper summing of the acoustical output (group delay matching is required). More horns covering narrower frequency ranges equal more headaches and poor summing...
Wider bandwidth even at the expense of some efficiency is always desirable (at least for music). Then again, most of the efficiency in a multi-way system is *lost* if good summing is not achieved, and the narrower the frequency range covered by each speaker, the worse the summing problem becomes.
Wider bandwidth even at the expense of some efficiency is always desirable (at least for music). Then again, most of the efficiency in a multi-way system is *lost* if good summing is not achieved, and the narrower the frequency range covered by each speaker, the worse the summing problem becomes.
Summing is difficult in the higher frequencies. But in the bass area, the wavelength is so long that we are allways in the < 1/4 WL area, concerning driver distance. A digital delay dialed in by pink noise rta measurements will work fine.
Apart from this, i also favour a wide bandwidth for each driver, just out of unreflected preference.
Apart from this, i also favour a wide bandwidth for each driver, just out of unreflected preference.
Summing is equally hard at all frequencies. LF horns are much longer resulting in proportionally longer group delays. Furthermore, the small mouths used in LF horns result in proportionally longer group delays (higher Q resonant modes) in comparison with HF horns, which usually have gentle mouths due to the lack of size constraints.
Note that a 1/4 wavelenght error (90 degrees out of phase) already results in a 3dB/W sensitivity loss across the crossover region, which is 1.5 octaves wide for 24dB/oct LR filters. This is wasting half the amplifier power. Phase ripple in horns, and particularly in tapped horns, makes it even worse.
There is nothing particularly wrong with group delays, but good summing requires matching with the help of parametric EQs and all-pass filters (DSP).
Different alignments may be freely mixed in narrow frequency bands, but the group delay matching task becomes more complex.
Note that a 1/4 wavelenght error (90 degrees out of phase) already results in a 3dB/W sensitivity loss across the crossover region, which is 1.5 octaves wide for 24dB/oct LR filters. This is wasting half the amplifier power. Phase ripple in horns, and particularly in tapped horns, makes it even worse.
There is nothing particularly wrong with group delays, but good summing requires matching with the help of parametric EQs and all-pass filters (DSP).
Different alignments may be freely mixed in narrow frequency bands, but the group delay matching task becomes more complex.
This is the first time i hear of group delay matching of different sources. The concept sounds logical though. Is this usually done in the professional sound area?
When i said that summing is easier, i referred only to the summing of two signals with constant group delay. Since all you need to do here is put the sources closer than 1/4 WL apart and adjust the delay, summing gets easier as WL gets longer. If one takes non constant group delay into account, it gets harder, just as you say.
When i said that summing is easier, i referred only to the summing of two signals with constant group delay. Since all you need to do here is put the sources closer than 1/4 WL apart and adjust the delay, summing gets easier as WL gets longer. If one takes non constant group delay into account, it gets harder, just as you say.
Hi,
isn't group delay simply the amount of time a signal of certain bandwith needs to make it through a system (block)?
In that sense, aligning two signals in the time domain comes out as being the good old analogue all pass filtering (or digital delay) of speaker ways, which then prepares those ways for phase alignment at the desired frequency (point).
Correct me if I'm wrong.
isn't group delay simply the amount of time a signal of certain bandwith needs to make it through a system (block)?
In that sense, aligning two signals in the time domain comes out as being the good old analogue all pass filtering (or digital delay) of speaker ways, which then prepares those ways for phase alignment at the desired frequency (point).
Correct me if I'm wrong.
Yes, near the crossover region group delay ripple has to be corrected in order to get a smooth polar pattern (except for co-axial systems, of course). This is easy to achieve with a DSP but it's only seen on the best PA systems. A simpler scheme involves correcting only the (theoretically smooth) group delay due to the lower and upper natual roll-off of each driver (and propagation delays), but this is hardly enough for bass horns.
As I think that I recall that Mr. Cowan had one of his Perless XLS almost "smooking" when under use, I wonder what you can do to protect yourself against driver failure in your tapped horn during the planning stage. That is, there "must" be some safety guidelines to go by when selecting drivers? I guess the two most obvious would be Xmax and VC size, and perhaps power handling?
I will go for tapped horns, using 12" car subwoofers, or maybe a 12" PA driver as a starter project due to size restrictions.
I will go for tapped horns, using 12" car subwoofers, or maybe a 12" PA driver as a starter project due to size restrictions.
G'day buggsson
To put this into perspective, it must be realised that this was part of a maximum output domestic system in a large room and not at all representative of what most of us want or need. The two 32Hz tapped horns could easily peg an RS SPL meter on it's highest range at 1M from the adjacent mouths. (>126dB) The Tapped Horns were being driven at >500W RMS each with a heavily limited Slayer track. The music I listen to would never cause undue voicecoil stress with 1KW per driver. I did find it interesting that at this power level, and with content containing most of it's energy between 40 and 80Hz, the excursion was very low. This probably led to the hot voicecoils too. The voicecoils weren't damaged, not even discoloured, but they did smell hot. Next time I will try a 9mm x-max driver with a 4" voicecoil. Excursion is not a concern in a ~30Hz tapped horn.
Cheers
William Cowan
To put this into perspective, it must be realised that this was part of a maximum output domestic system in a large room and not at all representative of what most of us want or need. The two 32Hz tapped horns could easily peg an RS SPL meter on it's highest range at 1M from the adjacent mouths. (>126dB) The Tapped Horns were being driven at >500W RMS each with a heavily limited Slayer track. The music I listen to would never cause undue voicecoil stress with 1KW per driver. I did find it interesting that at this power level, and with content containing most of it's energy between 40 and 80Hz, the excursion was very low. This probably led to the hot voicecoils too. The voicecoils weren't damaged, not even discoloured, but they did smell hot. Next time I will try a 9mm x-max driver with a 4" voicecoil. Excursion is not a concern in a ~30Hz tapped horn.
Cheers
William Cowan
Hi William,
Were you using the 830564's? I don't remember seeing your Hornresp inputs for the '564 ... could you post it?
What is it with you Aussies, first T/S, then Lenard, and McBean, and of course you! Is it something in the water. How come more than half of the good sound guys come from "Down Under"?
Well, at least we've got Tom Danley!
~Don
P.S. My tapped horn has been banished from the house, everything kept moving around.
Were you using the 830564's? I don't remember seeing your Hornresp inputs for the '564 ... could you post it?
What is it with you Aussies, first T/S, then Lenard, and McBean, and of course you! Is it something in the water. How come more than half of the good sound guys come from "Down Under"?
Well, at least we've got Tom Danley!
~Don
P.S. My tapped horn has been banished from the house, everything kept moving around.
G'day Don
Here is my model for my latest tapped horn. It's a bit shorter than the one on my web page, and the throat has been tweaked a bit. Measurements were remarkably close to this sim. Except for the amplitude of the high Q peaks, it measured essentially the same. The driver won't take enough power to reach X-max in the passband or even a little below it.
You're too nice Don, putting me up on a pedestal with the others!
I do think there is something about Aussies that drive them to experiment. We have less off the shelf choice here, so DIY is sometimes a necessity. You can't just buy Danley or Geddes speakers here! Freight would cost 1/2 as much as the speakers.
The workshop out the back has always been held in high esteem by most Aussie blokes, almost a home away from home I needed some new beds for my daughters and was unable to find one off the shelf that my wife and I liked, so I built them. They match the rooms I also built for them. I could never live with a commercial speaker system either, there's just nothing decent here. (See http://www.cowanaudio.com and click on Unity, the finale).
Chris8sirhC.......That's a tough one. They are good at different things, but both work as intended. Both are very clean, have low group delay and high output. One is portable and one is not. The tapped horn costs less, but doesn't go as low. Choose the mix that suits you, either option is capable of excellent performance. Don't procrastinate, just do it!
Cheers
William Cowan
Here is my model for my latest tapped horn. It's a bit shorter than the one on my web page, and the throat has been tweaked a bit. Measurements were remarkably close to this sim. Except for the amplitude of the high Q peaks, it measured essentially the same. The driver won't take enough power to reach X-max in the passband or even a little below it.
You're too nice Don, putting me up on a pedestal with the others!
I do think there is something about Aussies that drive them to experiment. We have less off the shelf choice here, so DIY is sometimes a necessity. You can't just buy Danley or Geddes speakers here! Freight would cost 1/2 as much as the speakers.
The workshop out the back has always been held in high esteem by most Aussie blokes, almost a home away from home
I needed some new beds for my daughters and was unable to find one off the shelf that my wife and I liked, so I built them. They match the rooms I also built for them. I could never live with a commercial speaker system either, there's just nothing decent here. (See http://www.cowanaudio.com and click on Unity, the finale).Chris8sirhC.......That's a tough one. They are good at different things, but both work as intended. Both are very clean, have low group delay and high output. One is portable and one is not. The tapped horn costs less, but doesn't go as low. Choose the mix that suits you, either option is capable of excellent performance. Don't procrastinate, just do it!
Cheers
William Cowan
Attachments
when i try to get my tapped horn to reach full volume it starts 2 make funny noises and go out of control, i am running it of the right channel of a old sansui AU-317 that my mate found at his job at the dump
it is supposed to be 50w per channel,
im guessing its the amp causing it to severely distort, or could it be the driver struggling with the pressure in the compression chamber? cos it moves alot of air, i can feel it can move more than it does without such distortion, as it does if i play lower frequency's
it is supposed to be 50w per channel,
im guessing its the amp causing it to severely distort, or could it be the driver struggling with the pressure in the compression chamber? cos it moves alot of air, i can feel it can move more than it does without such distortion, as it does if i play lower frequency's