These curve started to look like what your were getting with AJHorn. My SPL plot is in quarter space due to the floor boundary condition. Your assumption of eigth space indicates you are also using the side wall to reinforce the SPL from the mouth. This is one difference in our models.
The driver I am using is the Lowther DX2 which has a Qts of 0.3 which os on the low end of your recommended Qts values.
One other difference is the shape of the horn, I am using a linear taper to get more bass boost. The exponential taper only brought the SPL response up to flat with the driver's efficiency.
So bottom line, due to the assumption of radiation space and the Qts difference I am getting a little less gain for the bass frequencies. But even with this gain, it may be possible to counter the baffle step loss at low frequencies which is important to full range driver users.
Then I took the output form this model and put it inot my more advanced model. This model includes the geometry of the front baffl, the flor reflection, the baffle step response, and a rear wall some distance behind the enclosure. The model of the front baffle eis shown below.
The driver I am using is the Lowther DX2 which has a Qts of 0.3 which os on the low end of your recommended Qts values.
One other difference is the shape of the horn, I am using a linear taper to get more bass boost. The exponential taper only brought the SPL response up to flat with the driver's efficiency.
So bottom line, due to the assumption of radiation space and the Qts difference I am getting a little less gain for the bass frequencies. But even with this gain, it may be possible to counter the baffle step loss at low frequencies which is important to full range driver users.
Then I took the output form this model and put it inot my more advanced model. This model includes the geometry of the front baffl, the flor reflection, the baffle step response, and a rear wall some distance behind the enclosure. The model of the front baffle eis shown below.
Attachments
This is my first quick pass at calculating the system SPL response. The ripples can be improved by changes to the geometry. The most interesting conclusion I draw from this plot is that a BSC filter is probably not needed, this is a goal I have been trying to meet with my back loaded horn designs.
There is one more thing that I have taken for granted. I have modeled the coupling chamber as a lumped parameter and have not included any standing waves that will exist in such a large volume (40 liters). For this approach to work well, the control of standing waves in the coupling chamber is required.
Hope that is of some interest,
There is one more thing that I have taken for granted. I have modeled the coupling chamber as a lumped parameter and have not included any standing waves that will exist in such a large volume (40 liters). For this approach to work well, the control of standing waves in the coupling chamber is required.
Hope that is of some interest,
Thanx Martin, your contribution is much appreciated on this already interesting approach to BLHs.
dave
Hi MJK and @ll others,
what do u think of our theory? Do u have any questions?
@MJK: your simulations look good. Actually like mine.
Till
what do u think of our theory? Do u have any questions?
@MJK: your simulations look good. Actually like mine.
Till
@ Tyimo
Hi Tyimo
actually it is not only to understand the Replicon, we wanted to show a different way to build BL-hornspeakers. Thats why I thought it would be interesting for others and open up a new thread.
But we will discus quite a lot when the side is open for everyone.
best regards Till
Hi Tyimo
actually it is not only to understand the Replicon, we wanted to show a different way to build BL-hornspeakers. Thats why I thought it would be interesting for others and open up a new thread.
But we will discus quite a lot when the side is open for everyone.
best regards Till
Hello Till,
I cannot find the length of the horn in your explanation. Is it a quarter wavelength of fl ? Does it change if I want to calculate the horn for 1/4 space instead of 1/8 space placement?
I want to thank you so much for sharing your knowledge and experience and writing long and detailed answers. Thank you.
Bye, Elam
I cannot find the length of the horn in your explanation. Is it a quarter wavelength of fl ? Does it change if I want to calculate the horn for 1/4 space instead of 1/8 space placement?
I want to thank you so much for sharing your knowledge and experience and writing long and detailed answers. Thank you.
Bye, Elam
Hi Elam,
the length has nothing to do wit 1/4 or 1/8 length. You will get the length if you use a formula someting like:
l=ln[(Am/Ah)/......]
as soon as I got the exact formula (havent got it in mind right now) I will post it here.
have a nice weekend
Till
the length has nothing to do wit 1/4 or 1/8 length. You will get the length if you use a formula someting like:
l=ln[(Am/Ah)/......]
as soon as I got the exact formula (havent got it in mind right now) I will post it here.
have a nice weekend
Till
planet10 said:
I am very very interested about understanding the Replikon horn
I have build the ML TL with the Fostex FE 167 E with a bsc
they sound very good but i still have the feeling the Fostex
can perform better..
Would they do better in a replikon ... or not ?
Hi weidok,
The Replikon is designed only for the Fostex FE 208Sigma (the old one). We calculate the horns according the parameters of the driver. Only here you have the best performance. The fe167E is not very suitable for horn designs like we do. A horn without comprimize is always big. The Replikon is actually the minimum size of a real horn. The Unikon and all other small horns with 3 or 4 inch drivers in comparism are always a compromize. But as far as we know there is a great request and interest for such miniature horns. But in thruth you will never reach the bass output such as big horns do.
Greetings
Raffi
The Replikon is designed only for the Fostex FE 208Sigma (the old one). We calculate the horns according the parameters of the driver. Only here you have the best performance. The fe167E is not very suitable for horn designs like we do. A horn without comprimize is always big. The Replikon is actually the minimum size of a real horn. The Unikon and all other small horns with 3 or 4 inch drivers in comparism are always a compromize. But as far as we know there is a great request and interest for such miniature horns. But in thruth you will never reach the bass output such as big horns do.
Greetings
Raffi
Hello Till,
looking at your equations I understand that
l = log( Am / Ah ) / k
or
Ah = Am * exp( -k * l )
An equation for either Ah or l was missing in your post or I just can find it
Bye, Elam
looking at your equations I understand that
l = log( Am / Ah ) / k
or
Ah = Am * exp( -k * l )
An equation for either Ah or l was missing in your post or I just can find it
Bye, Elam
Hi Till and Raffi,
I think that your theory has some real advantages. No questions at this time, I think I understand your method and will continue to run a few more simulations to see what can be achieved.
I have run many of the back loaded horn designs that have plans posted on the Internet and the results have not been very good. I have a design of my own that I am working on but your approach is definitely more compact and probably easier to design and build. Thank you for sharing what you have learned.
That is a very good result.
I think that your theory has some real advantages. No questions at this time, I think I understand your method and will continue to run a few more simulations to see what can be achieved.
I have run many of the back loaded horn designs that have plans posted on the Internet and the results have not been very good. I have a design of my own that I am working on but your approach is definitely more compact and probably easier to design and build. Thank you for sharing what you have learned.
That is a very good result.
Hi!
Raffi and Till!
I have one more practical question:how do you calculate/design the horn bending? How can I measure the correct horn line lenght, if I bend it in your style?
Martin!
What is the linear horn? I know the Hyperbolic, the Exponential and theTractrix horns.
Greets:
Tyimo
Raffi and Till!
I have one more practical question:how do you calculate/design the horn bending? How can I measure the correct horn line lenght, if I bend it in your style?
Martin!
What is the linear horn? I know the Hyperbolic, the Exponential and theTractrix horns.
Greets:
Tyimo
A linear horn is one where the area varies linearly along the length.
x = 0 then Area = S_throat
x = L/4 then Area = S_throat + 1/4 (S_mouth - S_throat)
x = L/2 then Area = S_throat + 1/2 (S_mouth - S_throat)
x = 3L/4 then Area = S_throat + 3/4 (S_mouth - S_throat)
x = L then Area = S_mouth
x = 0 then Area = S_throat
x = L/4 then Area = S_throat + 1/4 (S_mouth - S_throat)
x = L/2 then Area = S_throat + 1/2 (S_mouth - S_throat)
x = 3L/4 then Area = S_throat + 3/4 (S_mouth - S_throat)
x = L then Area = S_mouth
Greets!
There's obviously nothing terribly wrong with it since as I previously alluded to in post #5, you're claiming a well proven design philosophy as your own, though from the sims it looks like you tweaked it some. Not enough IMO to claim it as your own though. "What was once old is yet new again." Oh well, as they say, "plagiarism is the sincerest form of flattery". 😉
You are one quick 'study'! Your WSs save you a lot of time too. I had to experiment with a mini Jensen Imperial, do lots of head scratching, and build several small 'proofs of concept' before I figured it out 'well enough', and that was just for expo loading and slow expansion conical (~hyperbolic, my preference). Your WS was the first reference I'd seen for a linear horn expansion, though at a glance it yields too much gain over too narrow a BW to be of much general use IMO. Still, one can't have too many tools in today's hi-tech world, so thanks for sharing!
In the end though, as a DIYer more interested in performance than parts count/cost, I didn't see the point since I could get more performance overall in the same, or smaller package by adding an appropriate LF/midbass system, so never built any for HIFI apps. Not much chance a single driver BLH could survive one of my good ol' boy 'clients' cranking up 'Freebird', 'Get it While You Can', or similar.
As always though, "different strokes for different folks", YMMV, etc.,
GM
Roemhild said:
There's obviously nothing terribly wrong with it since as I previously alluded to in post #5, you're claiming a well proven design philosophy as your own, though from the sims it looks like you tweaked it some. Not enough IMO to claim it as your own though. "What was once old is yet new again." Oh well, as they say, "plagiarism is the sincerest form of flattery". 😉
MJK said:
You are one quick 'study'! Your WSs save you a lot of time too. I had to experiment with a mini Jensen Imperial, do lots of head scratching, and build several small 'proofs of concept' before I figured it out 'well enough', and that was just for expo loading and slow expansion conical (~hyperbolic, my preference). Your WS was the first reference I'd seen for a linear horn expansion, though at a glance it yields too much gain over too narrow a BW to be of much general use IMO. Still, one can't have too many tools in today's hi-tech world, so thanks for sharing!
In the end though, as a DIYer more interested in performance than parts count/cost, I didn't see the point since I could get more performance overall in the same, or smaller package by adding an appropriate LF/midbass system, so never built any for HIFI apps. Not much chance a single driver BLH could survive one of my good ol' boy 'clients' cranking up 'Freebird', 'Get it While You Can', or similar.
As always though, "different strokes for different folks", YMMV, etc.,
GM
Hi GM,
This type of work is what I have been doing for a living for over 25 years. I do exactly the same thing at work using the same tools (ANSYS and MathCad) just different products Math model, test, correlate with reality, learn, and document. The next math model is even better. This type of work has served me well, two months ago I got a better offer from another local company and said good bye to the one I had been at for almost 15 years. My old managers were not happy. I just wish I could find a way to work on speaker theory full time, I have a lot of ideas and only an hour or two a week to work on them. Only problem is I really suck at building things, takes me forever.
I don't disagree, I am going to try your approach when I have time since I already have the parts. For me, the BLH is just too interesting of an intellectual challenge. You never know what you will learn and maybe there will be something there that has not been tried before and really works out well. Who knows unless you try.
Your gonna make me break out those Skynyrd albums and crank um up to see how ithey sound played through my Lowthers. I have not heard those old southern rock albums in a long time ..... they are calling me now from the back of the closet.
This type of work is what I have been doing for a living for over 25 years. I do exactly the same thing at work using the same tools (ANSYS and MathCad) just different products Math model, test, correlate with reality, learn, and document. The next math model is even better. This type of work has served me well, two months ago I got a better offer from another local company and said good bye to the one I had been at for almost 15 years. My old managers were not happy. I just wish I could find a way to work on speaker theory full time, I have a lot of ideas and only an hour or two a week to work on them. Only problem is I really suck at building things, takes me forever.
I don't disagree, I am going to try your approach when I have time since I already have the parts. For me, the BLH is just too interesting of an intellectual challenge. You never know what you will learn and maybe there will be something there that has not been tried before and really works out well. Who knows unless you try.
Your gonna make me break out those Skynyrd albums and crank um up to see how ithey sound played through my Lowthers. I have not heard those old southern rock albums in a long time ..... they are calling me now from the back of the closet.
GM said:
Hi GM,
as far as we know our approach inclusive our formulas has not yet been published or quoted in any literature, in fact the simulation curves might look like not unused.
Greetings
Raffi
- Status
- Not open for further replies.