Hi sdad,
Maybe take a look in this thread:
http://www.diyaudio.com/forums/subwoofers/176089-easy-horn-excel-spreadsheet.html
Regards,
Maybe take a look in this thread:
http://www.diyaudio.com/forums/subwoofers/176089-easy-horn-excel-spreadsheet.html
Regards,
Hello,
my easy way is:
Calculation of expo horns
If you have AH, AM, l,
k still lacks the flare rate:
ln (AM / AH) / l = k
so you can see the surface
each point out (Al):
INV ln (l * k) * AH = Al
l m, AH + AM + AL in square meters
By Al e.g. m in width and divide
you have the distance in meters
AH = Horn throat
AM = Horn mouth
k = flare rate
l = length
Lengths and flare rate VALUES
K VALUES:
0.37 ~ 10 Hz
0.55 ~ 15 Hz
0.7 ~ 19 Hz
0.9 ~ 25 Hz
1.1 ~ 30 Hz
1.3 ~ 35 Hz
1.48 ~ 40 Hz
About an octave above k plays the horn,
if the length matches. At BL horn lengths
over 3.4 m, you get a suck out
below 100 Hz
4/lamda lengths VALUES
3.4 m ~ 25 Hz
2.8 m ~ 30 Hz
2.4 m ~ 35 Hz
2.1 m ~ 40 Hz
1.7 m ~ 50 Hz
1.4 m ~ 60 Hz
Most of my horns open
the last~ 75 cm (look ALPHORN)
with a double or triple k,
to get a better mouth+room
crossover.
my easy way is:
Calculation of expo horns
If you have AH, AM, l,
k still lacks the flare rate:
ln (AM / AH) / l = k
so you can see the surface
each point out (Al):
INV ln (l * k) * AH = Al
l m, AH + AM + AL in square meters
By Al e.g. m in width and divide
you have the distance in meters
AH = Horn throat
AM = Horn mouth
k = flare rate
l = length
Lengths and flare rate VALUES
K VALUES:
0.37 ~ 10 Hz
0.55 ~ 15 Hz
0.7 ~ 19 Hz
0.9 ~ 25 Hz
1.1 ~ 30 Hz
1.3 ~ 35 Hz
1.48 ~ 40 Hz
About an octave above k plays the horn,
if the length matches. At BL horn lengths
over 3.4 m, you get a suck out
below 100 Hz
4/lamda lengths VALUES
3.4 m ~ 25 Hz
2.8 m ~ 30 Hz
2.4 m ~ 35 Hz
2.1 m ~ 40 Hz
1.7 m ~ 50 Hz
1.4 m ~ 60 Hz
Most of my horns open
the last~ 75 cm (look ALPHORN)
with a double or triple k,
to get a better mouth+room
crossover.
First, you need a good horn model. I recommend, "On the Specification of Moving-Coil Drivers for Low-Frequency Horn-Loaded Loudspeakers," and it can be found here: users.ece.gatech.edu/mleach/papers.html
Now, you need the go to paper on THs. That is the TH patent, under the name: "Sound Reproduction With Improved Low Frequency Characteristics" There are several sites where you can find it, and make sure it has the images as well. Several omit them, and it makes no sense without them.
You could also do as tb46 suggested, and check out my horn spreadsheet thread, and the excel document itself (the one one the last page.
) The sheet has all the needed equations in it, and you can see them to the far right. There are some good explanations of why I did things the way I did, see post 24. There is also a similar thread at AVS & Hometheatershack with slightly different info, and explanations of my madness.I have found this to help some people. Read the TH ripoff paper from Wisdom Audio, they make Tapped Pipes.
http://www.wisdomaudio.com/pdfs/RTL.pdf
Skip the first two PR sections, and read the rest. Pretty simple right? Now read the Danley white paper. It will make more sense now. http://www.danleysoundlabs.com/pdf/danley_tapped.pdf Then go back, and tackle the patent doc.
Good luck.
I am experimenting with a few ideas, and in doing so, I am working with AkAbak, as hornresp doesn't seem to offer the flexibility I need. I have created a script, but have a question regarding how to converge 2 waveguide segments into a common horn. This requires 2 input nodes and I see that a horn has a single input node. To merge the 2 input waveguides together I used ducts with a very small thickness. Is this the correct approach? Here's the script, feel free to comment. Wouldn't post if I was afraid of learning something here.
Here's the easiest way to make a decent tapped horn that I am aware of:
Audio Psychosis • View topic - Insanely Easy Tapped Horns
It takes a lot of the guesswork out of tapped horn design.
Danley gets between 7 and 9db of gain.
I have one in my shed I did for a winter audio show that has 7 db of gain.
What Patrick shows is about 5 db. Better than nothing to be sure. But not fully optimized. To get real mechanical efficiency out of a box you need the correct driver.
And to call a box a horn at least in my books it must have mechanical gain. There must be a rise in efficiency.
A self imposed "rule" is to have a minimum of 6db of gain before I'm willing to call a tappered resonant enclosure a horn.
I have one in my shed I did for a winter audio show that has 7 db of gain.
What Patrick shows is about 5 db. Better than nothing to be sure. But not fully optimized. To get real mechanical efficiency out of a box you need the correct driver.
And to call a box a horn at least in my books it must have mechanical gain. There must be a rise in efficiency.
A self imposed "rule" is to have a minimum of 6db of gain before I'm willing to call a tappered resonant enclosure a horn.
A small killer driver that works well is this one:
HiVi SP10 10" Subwoofer
Smallish box and stupendous output. Two in tandem are a real treat.
HiVi SP10 10" Subwoofer
Smallish box and stupendous output. Two in tandem are a real treat.
No AkAbak experience, but Prof. Leach's compression horn math will calculate the driver's parameters.
GM
GM, Allow me to read Leach's work before commenting. Having said that, allow this comment - question - will his work include the merging of 2 paths into a single manifold? ( Please, this is tongue in cheek, let me read that over ;>) ) An obvious work around, and this is NOT what I'm after here, is to build 2 completely separate boxes and put them side by side. Thing is when I do this in AkAbak, my results are in fact comparable to my duct approach, making me believe the duct is appropriate. I'm looking for an authoritative approval of this approach, is all. This is my first crank with AkAbak, and I hold water like a sieve.
I am after a common ouput horn as I am in need of addressing the next frequency range up from this TH. I am looking at using the existing horn as output for this next range. This makes me believe that I will need additional modelling interfaces into this chamber.
I am after a common ouput horn as I am in need of addressing the next frequency range up from this TH. I am looking at using the existing horn as output for this next range. This makes me believe that I will need additional modelling interfaces into this chamber.
Sorry, not really following along, my mind is still pretty much mush from days of high fever, so all I'll say is that ML's math designs a 4th order band-pass (BP) compression horn. As such, you must provide the proper design goals to get any useful info, especially for a TH. Once you have an alignment, then you can divide the horn up into multiple smaller ones that can be stacked/whatever or just multiple drivers feeding one large one or any other combination required such as truncating it to shorten its axial length to meet the needs of the app.
WRT THs in general, there's different alignments with end loading to create a simple 6th order band-pass (BP) being the most common since they have the widest BW/bulk once EQ'd. A true TH though is one with high gain, so as a general rule its BW is limited to its tap point, so to get a wide BW (for a TH) requires a very powerful motor and suspension that can handle the loading, though for typical HIFI/HT apps the latter can be relaxed somewhat.
HTH,
GM
WRT THs in general, there's different alignments with end loading to create a simple 6th order band-pass (BP) being the most common since they have the widest BW/bulk once EQ'd. A true TH though is one with high gain, so as a general rule its BW is limited to its tap point, so to get a wide BW (for a TH) requires a very powerful motor and suspension that can handle the loading, though for typical HIFI/HT apps the latter can be relaxed somewhat.
HTH,
GM
Nasty business, this bug. In this household, as well. Just read over Leech's paper. Certainly enough info to derive my solution. Will start on that derivation after this post.
At this early on milepost in my study, I have come this conclusion: A recursive optimization algorithm appears to be worth considering. Current programs all seem to be a point snapshot and trying to find the "optimal" fit for a specific set of conditions isn't present in anything I've run across. This not at all surprising, defining "optimal" is an OMG! However, with the dozen(s) of variables available , some being user taste controlled, it would interesting to see what effects are created when messing with these items. I'm sure some are of no consequence, others so finicky that not worth considering.
Sorry, I ramble on.
At this early on milepost in my study, I have come this conclusion: A recursive optimization algorithm appears to be worth considering. Current programs all seem to be a point snapshot and trying to find the "optimal" fit for a specific set of conditions isn't present in anything I've run across. This not at all surprising, defining "optimal" is an OMG! However, with the dozen(s) of variables available , some being user taste controlled, it would interesting to see what effects are created when messing with these items. I'm sure some are of no consequence, others so finicky that not worth considering.
Sorry, I ramble on.
This software will do Leach's calcs for you http://fullrangedriver.com/singledriver/download/MLutil.zip
You do not need the ducts. The horn segment only has one input node, but you can attach as many other segments to it as you want.
That said, you really don't want to use a horn segment there though, as it messes the radiation/FR up. Long story short, if you are simulating a single section horn with a standard single flare rate you can use the "Horn" component. For anything else use a waveguide and Radiator combo endcap instead.
Here's a quick sketch of what I am trying to model . the 2 spkrs front load into a common cavity (what I called the horn) The rear of each spkr is kept separate from each other, as I want to investigate the effects of different resonant lengths being summed inside that common cavity.
I hope this makes sense.
Thanks for your time, by the way. This may all fall apart on me, but at least I will have learned something in the process.
I hope this makes sense.
Thanks for your time, by the way. This may all fall apart on me, but at least I will have learned something in the process.
The basic concept is no different than this TH: DANLEY | TAPPED HORN SUBWOOFER
If the path-lengths are different, then they will comb filter into a series of peaks, nulls, with a rolled off bottom end and HF Chop Suey similar to what one gets when a BLH is designed to overlap with the driver's output well up into the hundreds of Hz.
GM
If the path-lengths are different, then they will comb filter into a series of peaks, nulls, with a rolled off bottom end and HF Chop Suey similar to what one gets when a BLH is designed to overlap with the driver's output well up into the hundreds of Hz.
GM
GM,
Ya, true enough. I understand that making the paths different will give a variety of sometimes chilling results. What I am ultimately after is being able to model this architecture within AkAbak. See, what I need to be able to do is use AkAbak with confidence. So far, I can't do that. At this point I don't honestly care about the quality, or lack thereof, of this or any design. Its all about being able to use AkAbak. I find soho54's comment about replacing the horn with waveguide and radiator very informative. I need to mess with that a bit. I need to reread the instruction set to see how to use a waveguide with 2 inputs. Curse of the ancient: "don't remember so good no more!"
Now, once that has been accomplished, the fun really begins!!!!!!!!!!!!!! We will then be in a position where we can argue philosophy until our tonsils bleed. Until then I just don't count, frankly.
Ya, true enough. I understand that making the paths different will give a variety of sometimes chilling results. What I am ultimately after is being able to model this architecture within AkAbak. See, what I need to be able to do is use AkAbak with confidence. So far, I can't do that. At this point I don't honestly care about the quality, or lack thereof, of this or any design. Its all about being able to use AkAbak. I find soho54's comment about replacing the horn with waveguide and radiator very informative. I need to mess with that a bit. I need to reread the instruction set to see how to use a waveguide with 2 inputs. Curse of the ancient: "don't remember so good no more!"
Now, once that has been accomplished, the fun really begins!!!!!!!!!!!!!! We will then be in a position where we can argue philosophy until our tonsils bleed. Until then I just don't count, frankly.