I know...it's a learning curve that's taking longer than it should, but for me is very valuable. I apologise that I've tried your patience. It's certainly not intentional...my kids are probably giving me brain damage!
So, I think I've found an explanation as to why my Tractrix horn doesn't really look like one. After playing with the program on John Inlow's page, I read some of the descriptive text that had been written by the author.
This may seem overly obvious, but I thought I'd share this observation if it helped anyone else. The low frequency cut-off determines the flared end of the horn (the mouth), and this appears to be the same size regardless of the driver.
What this means is that the driver's throat is the same size as the driver itself...so what you do is to sever the Tractrix shape where the diameter matches your driver - a bit like cutting the end off an icing pipe to ice the cake.
I might have been taking the wrong approach with HornResp, but it looks like it was calculating the length of the horn from the throat end, rather than from the mouth end!
I plugged a 150Hz Tractrix horn into the program, and, well, the flare is 6 feet in diameter! On Phil Melhuish's site, he mentions that using a half or quarter wavelength version of the horn might 'do', though I might need a bit of convincing on that!
Still, this has been enlightening for me 😉
So, I think I've found an explanation as to why my Tractrix horn doesn't really look like one. After playing with the program on John Inlow's page, I read some of the descriptive text that had been written by the author.
This may seem overly obvious, but I thought I'd share this observation if it helped anyone else. The low frequency cut-off determines the flared end of the horn (the mouth), and this appears to be the same size regardless of the driver.
What this means is that the driver's throat is the same size as the driver itself...so what you do is to sever the Tractrix shape where the diameter matches your driver - a bit like cutting the end off an icing pipe to ice the cake.
I might have been taking the wrong approach with HornResp, but it looks like it was calculating the length of the horn from the throat end, rather than from the mouth end!
I plugged a 150Hz Tractrix horn into the program, and, well, the flare is 6 feet in diameter! On Phil Melhuish's site, he mentions that using a half or quarter wavelength version of the horn might 'do', though I might need a bit of convincing on that!
Still, this has been enlightening for me 😉
Hi unaHm,
The mouth area of a fully-formed tractrix horn (mouth flare tangent angle = 90 degrees) is given by:
S2 = c ^ 2 / (4 * Pi * fc ^ 2)
Where:
S2 = mouth area
c = velocity of sound in air
fc = tractrix horn flare cutoff frequency
This can be done directly using the Horn Segment Wizard in Hornresp. Simply specify the desired throat area, horn cutoff frequency and mouth flare tangent angle, as shown in the attached screenprint.
Correct. Hornresp treats the tractrix horn the same as any other horn in this respect. L12 is the axial length of the horn (the distance between the throat and the mouth).
Kind regards,
David
The mouth area of a fully-formed tractrix horn (mouth flare tangent angle = 90 degrees) is given by:
S2 = c ^ 2 / (4 * Pi * fc ^ 2)
Where:
S2 = mouth area
c = velocity of sound in air
fc = tractrix horn flare cutoff frequency
This can be done directly using the Horn Segment Wizard in Hornresp. Simply specify the desired throat area, horn cutoff frequency and mouth flare tangent angle, as shown in the attached screenprint.
Correct. Hornresp treats the tractrix horn the same as any other horn in this respect. L12 is the axial length of the horn (the distance between the throat and the mouth).
Kind regards,
David
Attachments
All the units in hornresp are metric, so you might want to check again. My understanding has been that the cutoff frequency of a tractrix horn is the frequency where one wavelength is equal to the circumference of the mouth of a circular horn. I get about 28" diameter for a 150 hz horn or about 72cm.
I did. I apologise...I'd been looking at the same numbers for too long.
I get the same as you for the horn now. Thanks David again for your brief HornResp tutorial here!
Art, does this look more like it?
After a bout of illness and lots of frustration elsewhere, I've concluded this particular project with the 'midbass box':
I keep re-reading this thread, and I'm embarrassed by the amount of direction changes I've taken, to the noticeable frustration of those providing advice! I apologise for the turmoil that I've caused.
The design of the box above is based on the discussion around post #346. I picked a 12"x12" front baffle, and then calculated the depth to get the desired volume (32 litres).
The speakers work well, as we know the Alpha 8a would, and I'm slowly tweaking my receiver's DSP settings to get an even output between all three elements. I noticed they're not as sensitive without the horn, but that's OK - power consumption isn't an issue, and I'm in a small room.
If you're on a budget, DON'T attempt a Midbass horn...unless you already have some really good compression drivers! Cone drivers work well.
One day I'll be able to afford some Community M4s, or a JBL equivalent...somewhere where I need a horn, but for now, this'll do, and it's a very good compromise.
My next project will be to build a pair of Open Baffle speakers using tactile transducers, based on xrk971's design. Cheap and fun, and initial tests are certainly nice.
I keep re-reading this thread, and I'm embarrassed by the amount of direction changes I've taken, to the noticeable frustration of those providing advice! I apologise for the turmoil that I've caused.
The design of the box above is based on the discussion around post #346. I picked a 12"x12" front baffle, and then calculated the depth to get the desired volume (32 litres).
The speakers work well, as we know the Alpha 8a would, and I'm slowly tweaking my receiver's DSP settings to get an even output between all three elements. I noticed they're not as sensitive without the horn, but that's OK - power consumption isn't an issue, and I'm in a small room.
If you're on a budget, DON'T attempt a Midbass horn...unless you already have some really good compression drivers! Cone drivers work well.
One day I'll be able to afford some Community M4s, or a JBL equivalent...somewhere where I need a horn, but for now, this'll do, and it's a very good compromise.
My next project will be to build a pair of Open Baffle speakers using tactile transducers, based on xrk971's design. Cheap and fun, and initial tests are certainly nice.
You'll want to pass on the Community M4 drivers. They are not suitable for hi-fi. They are more at home at race tracks and stadiums.
The M200 is okay. I used to have the alu version before I found better drivers. I mentioned that I am building a Trio based system to a salesman at Stockholms biggest hifi store Audio Concept. He thought that the idea of an amateur improving on the work of a group of audio engineers was preposterous! He then continued by explaining to me that nobody is listening to horns anymore and that I was about 15 years late. To prove how wrong I were he demoed the Sonus faber stradivari homage. I liked it a lot and so did my girlfriend although she said they were a bit dry compared to the horns at home.
Last edited:
What would you recommend instead of the Community M4? Something like one of the large format JBL compression drivers?
hmm
2-way or a 3-way ?
crossover points ?
I too find an 8" crossed to a super clean horn at 2khz enough for me (if a sub is added). Next is to swap a jbl 2407 for a transformer coupled motorola piezo (with zobel and a notch).
midbass horns (assuming flat to 100hz) are close to 4' deep and need 6ft2 of mouth. On the floor isn't really 2pi. 2pi would be a horn flush mounted to the middle of the wall. A horn on a concrete floor is around 3pi. And you need to be 10' away for bass to develop (i had a keele w-bin for a while). You can avoid that if your horn is a full wavelength long. Most horns are 1/4 wavelength long.
And if you shorten the horn because the throat is large, you lose low end.
norm
2-way or a 3-way ?
crossover points ?
I too find an 8" crossed to a super clean horn at 2khz enough for me (if a sub is added). Next is to swap a jbl 2407 for a transformer coupled motorola piezo (with zobel and a notch).
midbass horns (assuming flat to 100hz) are close to 4' deep and need 6ft2 of mouth. On the floor isn't really 2pi. 2pi would be a horn flush mounted to the middle of the wall. A horn on a concrete floor is around 3pi. And you need to be 10' away for bass to develop (i had a keele w-bin for a while). You can avoid that if your horn is a full wavelength long. Most horns are 1/4 wavelength long.
And if you shorten the horn because the throat is large, you lose low end.
norm
Last edited:
The Community M200 is a perfectly fine driver. I don't understand your sharp response to my comment about the M4.
That depends on if you prefer phenolic (JbL 2480) or metal diaphragms, and how big horns you can fit. For smaller mid horns with 60cm diameter mouths, common 1" exit compression drivers should suffice. I must mention that there are special cases where 1" exits compression drivers still can reach low in large horns.
In my experience big horns with compression drivers depends on what kind of music you listen to. With music that has vocals and instruments that are clearly defined, horns are magical, but if you like big orchestra or heavy metal, or brit pop with a lot of reverb, that can be more chaotic, I would recommend something else. Open baffle, at least from mids and up can sound great. Open baffle bass woofers, not so much. Although stigerik seems pleased with the bass from his open baffle array with Beyma 21" woofers.
Last edited:
Like JLH says: "A low Q cone driver loaded into a horn."
I owned ten or a dozen Community M4 on their big horns (don't recall the model #) and a pair of EV 12M or EV 10M on a similar size (but much shorter) horn had much more low mid output and better sound quality. Eminence (and many other companies) make a variety of cone drivers that are similar (or better) than the EV drivers I was comparing, a pair of any of them would be a fraction of the M4 cost.
Even though I purchased the M4 and horns at a very good price (they had been removed from a convention center where they had been used as "ceiling speakers" the high Q horn was a bad choice for that application) I was not pleased with any aspect of their response. Later versions had an improved diaphragm, but low mid output was no better, and nearly any large format compression driver has smoother response in the upper midrange.
Art
Art, I don't think a 12" can compete with a compression driver above 500Hz. The M4 may be a horrible driver but there are better built compression drivers.
Last edited:
You know, after listening for a while with the Alpha speakers in their boxes, I can't handle it any more! I really miss the horns in the system!
I must solve this challenge, so I'm going to do my best to get these horns done...at least in a matching pair.
Maybe I should save and get a pair of those cheap compression drivers we'd talked about a while ago...distortion-wise they don't look great on paper, but I have no idea what that 'sounds' like in real life. Maybe I can tame the beast?
I must solve this challenge, so I'm going to do my best to get these horns done...at least in a matching pair.
Maybe I should save and get a pair of those cheap compression drivers we'd talked about a while ago...distortion-wise they don't look great on paper, but I have no idea what that 'sounds' like in real life. Maybe I can tame the beast?
Compete in what way?
Near the start of my career in sound it was common practice to cross 4" diaphragm compression drivers at 500 Hz. After hearing some horn loaded 12" drivers, it was quite apparent using horn loaded 12" or 10" drivers to around 1200 Hz greatly reduced distortion, and made for much more available SPL in the 500-1200 Hz range than can be produced using the limited displacement of a 4" diaphragm compression driver.
low qts horn loaded work great crossed at 1.2khz to a large 3" or 4" compression driver.
Only problem is how big the bass horn is and where to cross to subs, 150hz seems about right to keep bass horn manageable size.
Didn't jbl have a paper on horn loading woofers?
I think they measured much better than trying to push a 4" to an iunder 1lkhz crossover point.
Once you leave the double 15" + horn, you go up to a 3-way, 4632 and 3632 (dual horn loaded 6.5" frivers).
Can't find their pdf on it.
Only problem is how big the bass horn is and where to cross to subs, 150hz seems about right to keep bass horn manageable size.
Didn't jbl have a paper on horn loading woofers?
I think they measured much better than trying to push a 4" to an iunder 1lkhz crossover point.
Once you leave the double 15" + horn, you go up to a 3-way, 4632 and 3632 (dual horn loaded 6.5" frivers).
Can't find their pdf on it.
- Status
- Not open for further replies.