3D Folded Horn Design - diyAudio
Go Back   Home > Forums > Loudspeakers > Subwoofers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 4th September 2013, 02:58 AM   #1
diyAudio Member
 
Join Date: Aug 2013
Location: Abingdon, Virginia
Default 3D Folded Horn Design

I've been working on a folded horn design intended to maximize the use of internal volume in three dimensions as so maximize horn length. It is essentially a cube with seven 90 degree folds. In its simplest form it contains seven chambers. Only the mouth and one of the internal chambers are of increasing cross section along their length. All the rest are constant cross section, however, each is larger than the previous, and the output end of each is larger than the input end. Obviously aspects of this are non-optimal. However, a 3D folded horn is a difficult beast and if some aspects of the simple model can remain, construction is greatly simplified. I'm trying to balance the problems against each other if possible, or find simple additions to the design which can help ameliorate them.

My questions are:

To what extent does a series of straight pipes with outlet always larger than input equate a horn?

Most folded horns I see have rectangular cross sections. This design has several pipes of an assymetrical trapezoid cross section. Might this have an impact?

I'm assuming that the constant cross section segments will be prone to resonances from standing waves. Can these be calculated individually and the dimensions adjusted so they present peaks at different frequencies, or will the many turns and expanding mouth smear them back out? Also, several chambers are a wall thickness away from a different chamber. Might there be any conduction through the walls that might affect (for better or worse) those in the other chamber?

Finally, most folded horns have sharp bends (some with a small baffle at an angle to smooth the bend slightly). Would it be of any use, either with respect to resonance problems or just the transmission in general, for the bends to be curved?

Plans of the simplest design to follow. Note that no correction is made here for wall thickness, etc. This is simply for modeling purposes.

The box is a cube of 24" on a side. The shell is then six pieces of 24" x 24", one of which has a square cut out from one corner, 16" on a side. (fig. 1) The cut out piece is the "front".

One internal baffle (vertical) is 12" x 24". The other (horizontal starts as a 16" x 24" rectangle and has a notch cut from a 24" side so that that edge goes from the original corner to a point midway along that side but only 12" from the other long side, then back to the other original corner. (fig 2).

The only other piece is a side wall for the mouth, 12" in height on one end, 16" on the other, and 13" wide. (fig 3)

Build the main part of the box, bottom sides and back, using 24" squares. Install the vertical baffle inside the "left" wall, 12" from the back. Install two blocks on the inside of the back wall, 12" from the bottom, and one block on the inside of the "right" wall just at the "front" edge, 16" from the bottom. Set the horizontal baffle on the blocks. Place the mouth side wall under the diagonal edge of the horizontal baffle, touching the inside vertical edge of the vertical baffle and coming to the front edge of the bottom piece. Attach all loose parts. (fig 4)

Put the "front" piece on the box so that the mouth opening is on the bottom right, with the 16" cut edges mating with the front edge of the horizontal baffle and mouth side wall. Attach and seal all internal edges. With the last 24" square placed on top, the box is complete. But don't because it needs a driver.

The driver compartment is front bottom left. The driver can be installed facing up or on another baffle fitting inside this compartment at an angle, making this compartment also expanding. It could also be placed through the mouth side wall in the manner of a tapped horn. The sound path starts here and goes up, then front left to right across the upper half of the front. Here it
turns towards the back along the right wall (an expanding compartment), turns left across the back, then down along the back left edge to the bottom. It then turns right across the bottom of the back, to the bottom right and comes forward through the mouth. Internal path length through the centers of the compartments is approximately 7 feet. Expansion is (cross section of driver compartment = 120 in^2; cross section of mouth = 256 in^2) 2.13, approximately half of which (1.53) occurs at the mouth compartment.

Keep in mind this was conceived as a way of maximizing use of internal area in the smallest total volume. It was not designed from an acoustic standpoint. This project was intended to produce the design then see what could be done to maximize performance. Of course that will in part depend on the intended use of the speaker, which obviously includes driver selection. Also, this seems to better suited to subwoofer use than the full range as it was intended. I'd prefer to resolve the issues from that standpoint, but I'm already convinced this is where it might find better acceptance and use.

This design was actually built and tested using an SAS Bazooka 12". No filter or cross over was used. The garage door 150 feet across the street was rattling. Power was provided by an Audiosource AMP100 bridged to 160 watt mono and running at 100 watts. The impact was unexpectedly good. The tonal quality however was not impressive. This is what I'd like to tweak, with your help. Theoretical assertions are welcome as long as they're stated as such, and same for opinions. But if you make saw dust, your input is particularly appreciated.
Attached Images
File Type: gif 3D12.gif (6.7 KB, 424 views)
File Type: gif 3D3.gif (2.8 KB, 415 views)
File Type: jpg 3D4.jpg (73.3 KB, 419 views)
  Reply With Quote
Old 4th September 2013, 04:30 PM   #2
diyAudio Member
 
Join Date: Jan 2008
Quote:
Originally Posted by drmcclainphd View Post
1)To what extent does a series of straight pipes with outlet always larger than input equate a horn?
2) This design has several pipes of an assymetrical trapezoid cross section. Might this have an impact?
3) I'm assuming that the constant cross section segments will be prone to resonances from standing waves. Can these be calculated individually and the dimensions adjusted so they present peaks at different frequencies, or will the many turns and expanding mouth smear them back out?
4)Also, several chambers are a wall thickness away from a different chamber. Might there be any conduction through the walls that might affect (for better or worse) those in the other chamber?
5) Would it be of any use, either with respect to resonance problems or just the transmission in general, for the bends to be curved?
6)This design was actually built and tested using an SAS Bazooka 12".
1) The closer the series of pipes equate to the horn expansion ratio, the better, but for bass horns three or four straight (non expanding) sections is often enough.
That said, expansion can be designed in to most shapes, just takes a little head scratching and re-setting your saw blade to accomplish.
2) Shape does not matter much as long as the cross sectional area remains the same and the aspect ratio of the shape is not too high.
3) Resonances of sections probably can be calculated using Hornresp or Akaback.
4) The design should be adequately braced so wall members do not vibrate and resonate. Wall vibration can impart a ringing sound, and wastes upper output level.
5) Curved bends are helpful for high frequencies, which behave similar to beams of light (ray tracing) but actually waste valuable cabinet space in a sub 200 Hz (or so) cabinet, as low frequencies tend to flow in a laminar fashion.
6) The raw response of the speaker will be reflected in the output of the horn.
"Sub" type speakers are not appropriate for full range use, and folded horns designed for full range (or even up to 500 Hz) have different design consideration than subs.

Rattling (resonating) garage doors is fairly easy, getting a smooth response from a folded horn is more difficult. Sometimes small physical changes (which can't be simulated) can make fairly large acoustic changes. If you plan to make a good sounding cabinet, having some decent measurement tools (in addition to trained hearing) and learning how to use them will speed up the process quite a bit.

Art
  Reply With Quote
Old 4th September 2013, 05:11 PM   #3
AllenB is offline AllenB  Australia
diyAudio Member
 
Join Date: Oct 2008
I think the answer to each of your questions is essentially the same...Is the section/dimension etc. significantly smaller than a wavelength at the upper working frequencies of the horn.
  Reply With Quote
Old 4th September 2013, 05:44 PM   #4
diyAudio Member
 
Join Date: Aug 2013
Location: Abingdon, Virginia
Quote:
Originally Posted by AllenB View Post
I think the answer to each of your questions is essentially the same...Is the section/dimension etc. significantly smaller than a wavelength at the upper working frequencies of the horn.
Of the horn? You mean the entire internal length?
  Reply With Quote
Old 4th September 2013, 06:09 PM   #5
diyAudio Member
 
Join Date: Aug 2013
Location: Abingdon, Virginia
Quote:
Originally Posted by weltersys View Post
1) The closer the series of pipes equate to the horn expansion ratio, the better, but for bass horns three or four straight (non expanding) sections is often enough.
That said, expansion can be designed in to most shapes, just takes a little head scratching and re-setting your saw blade to accomplish.
2) Shape does not matter much as long as the cross sectional area remains the same and the aspect ratio of the shape is not too high.
3) Resonances of sections probably can be calculated using Hornresp or Akaback.
4) The design should be adequately braced so wall members do not vibrate and resonate. Wall vibration can impart a ringing sound, and wastes upper output level.
5) Curved bends are helpful for high frequencies, which behave similar to beams of light (ray tracing) but actually waste valuable cabinet space in a sub 200 Hz (or so) cabinet, as low frequencies tend to flow in a laminar fashion.
6) The raw response of the speaker will be reflected in the output of the horn.
"Sub" type speakers are not appropriate for full range use, and folded horns designed for full range (or even up to 500 Hz) have different design consideration than subs.

Rattling (resonating) garage doors is fairly easy, getting a smooth response from a folded horn is more difficult. Sometimes small physical changes (which can't be simulated) can make fairly large acoustic changes. If you plan to make a good sounding cabinet, having some decent measurement tools (in addition to trained hearing) and learning how to use them will speed up the process quite a bit.

Art
1 - In the one I built the notch in the horizontal baffle was smaller than described on the piece over the driver compartment, and larger than described over the rear, rather a bit of plywood sculpturing. I expect to do more.
2 - The compartment for the driver is the only really odd shaped section, nearly triangular.
3 --\I have a learning curve ahead of me. Or the need for more help.
4 - With the baffles firmly attached along all walls, the box is quite solid. Also, I typically attach odd shaped scraps in random places and orientations to break up wall resonance.
5 - I've used self-adhesive floor tiles for bends and can get bends of 2" radius. Helpful? Harmful? Useless?
6 - Absolutely a xover will be used, selected to match the box and driver.

I have an analog SPL meter with a jack so I can lock it in a box and run a wire out, to measure response in the sealed cab. It's not the best, but I always take several measurements at each freq, average them and test them for variance.

Might this design be used as a tapped horn? Placing the driver in the wall of the driver compartment, facing out? If so, might it also be possible to mate two drivers there, face to face and out of phase, for an isobaric driver? An isobaric requires only half the internal volume. Does the reverse hold (ie. does an isobaric driver on a given box make it effectively double the volume)?
  Reply With Quote
Old 4th September 2013, 07:23 PM   #6
diyAudio Member
 
Join Date: Jan 2008
Quote:
Originally Posted by drmcclainphd View Post
5 - I've used self-adhesive floor tiles for bends and can get bends of 2" radius. Helpful? Harmful? Useless?
6-Might this design be used as a tapped horn?
7-Placing the driver in the wall of the driver compartment, facing out?
8-If so, might it also be possible to mate two drivers there, face to face and out of phase, for an isobaric driver?
9-An isobaric requires only half the internal volume. Does the reverse hold (ie. does an isobaric driver on a given box make it effectively double the volume)?
5) Curved bends are helpful for high frequencies, which behave similar to beams of light (ray tracing) but actually waste valuable cabinet space in a sub 200 Hz (or so) cabinet, as low frequencies tend to flow in a laminar fashion.
The enclosed area behind self-adhesive floor tiles would be "squishy" like a balloon, not good.
6) Tapped horns can be pretty much any shape and layout.
7) A TH does not use a driver compartment, one side of the driver is near the mouth, the other near the throat.
8) Reverse polarity clam shell isobaric pairs have been used in all types of sub enclosures.
9) It takes displacement to make SPL, an isobaric pair has only the same displacement as one speaker.
An isobaric pair requires approximately half the internal volume to achieve the same low frequency cut off, but uses twice the power (and drivers) to achieve that. Doubling the enclosure volume would increase sensitivity, but would still be less sensitive than the drivers used individually.

In designing any horn enclosure maximizing the use of all internal space is important, and generally expanding sections are preferable.
  Reply With Quote
Old 5th September 2013, 03:14 PM   #7
diyAudio Member
 
Join Date: Aug 2013
Location: Abingdon, Virginia
Quote:
Originally Posted by weltersys View Post
5) Curved bends are helpful for high frequencies, which behave similar to beams of light (ray tracing) but actually waste valuable cabinet space in a sub 200 Hz (or so) cabinet, as low frequencies tend to flow in a laminar fashion.
The enclosed area behind self-adhesive floor tiles would be "squishy" like a balloon, not good.
6) Tapped horns can be pretty much any shape and layout.
7) A TH does not use a driver compartment, one side of the driver is near the mouth, the other near the throat.
8) Reverse polarity clam shell isobaric pairs have been used in all types of sub enclosures.
9) It takes displacement to make SPL, an isobaric pair has only the same displacement as one speaker.
An isobaric pair requires approximately half the internal volume to achieve the same low frequency cut off, but uses twice the power (and drivers) to achieve that. Doubling the enclosure volume would increase sensitivity, but would still be less sensitive than the drivers used individually.

In designing any horn enclosure maximizing the use of all internal space is important, and generally expanding sections are preferable.
Not responding to some does not mean I'm ignoring them, rather just my acceptance. Thanks, in each case.

7 - The driver compartment is just my terminology, as it can be configured various ways. It is the first pipe and thus the throat. It shares the side wall with the mouth.

9 - I'm not after SPL necessarily, but rather improving response. Power and drivers I have many. What I was asking after I found in the Wikipedia article on isobarics: "The two drivers operating in tandem exhibit exactly the same behavior as one loudspeaker in twice the cabinet". However, "if a speaker is optimized for performance in a 40 liter enclosure, one iso-group of the same speakers can achieve the same low frequency extension and overall response characteristics in a 20 liter enclosure." So were I to use an iso-pair I should use drivers optimized for an enclosure of 453 liters (16 cubic feet)? If so, it seems I'm better off sticking with a single driver. I have some that will do 8 cu ft, but none that will do 16, and fit in the space allowed.
  Reply With Quote
Old 5th September 2013, 07:19 PM   #8
diyAudio Member
 
Join Date: Jan 2008
Quote:
Originally Posted by drmcclainphd View Post
However, "if a speaker is optimized for performance in a 40 liter enclosure, one iso-group of the same speakers can achieve the same low frequency extension and overall response characteristics in a 20 liter enclosure." So were I to use an iso-pair I should use drivers optimized for an enclosure of 453 liters (16 cubic feet)? If so, it seems I'm better off sticking with a single driver. I have some that will do 8 cu ft, but none that will do 16, and fit in the space allowed
I have no idea how you got from 20 liters to 453 liters, or how you are figuring what the optimum enclosure size should be, but an optimum TH design will use a different enclosure size than a BR or sealed cabinet for the same low cut-off frequency.

At any rate, if you want to achieve any goal, you need to define what the goal is.
  Reply With Quote
Old 6th September 2013, 05:37 AM   #9
diyAudio Member
 
Join Date: Aug 2013
Location: Abingdon, Virginia
Quote:
Originally Posted by weltersys View Post
I have no idea how you got from 20 liters to 453 liters, or how you are figuring what the optimum enclosure size should be, but an optimum TH design will use a different enclosure size than a BR or sealed cabinet for the same low cut-off frequency.

At any rate, if you want to achieve any goal, you need to define what the goal is.
The quoted material was referenced as coming from the Wikipedia article, including the 20/40 liter example. The 453 is twice the volume of the design I presented, that being calculated from the volume of the box * 2, as per the example. And the goal was stated in the original post -- to start with the design intended to maximize the use of space, and then attempt to maximize performance with that design. I had assumed that a TH would be considered a vented design.
  Reply With Quote
Old 6th September 2013, 05:13 PM   #10
diyAudio Member
 
Join Date: Jan 2008
Quote:
Originally Posted by drmcclainphd View Post
And the goal was stated in the original post -- to start with the design intended to maximize the use of space, and then attempt to maximize performance with that design. I had assumed that a TH would be considered a vented design.
Dennis,

Your original post does not mention TH, nor does it show speaker position, or exit, or horn path, making it difficult to understand what type of design it is.

Vented designs are known as BR (bass reflex), a Helmholtz resonator.
Front loaded horns (FLH) use a compression chamber.
TH and FLH use a path that expands from throat to mouth.

"Maximize performance" is not a goal unless "performance" is specified in terms of efficiency, SPL, frequency response, distortion and anything else important to you.
A design for maximum output from 30-120 Hz and one designed for 50-150 or smooth response from 50-500 Hz are all different goals and would require different designs (and drivers) to achieve.

Using the free software program Hornresp you could model various designs using the same cabinet volume and speaker parameters and decide which fits your version of "maximum performance".

Art

Last edited by weltersys; 6th September 2013 at 05:17 PM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
easy-to-build 3D-folded horn el`Ol Multi-Way 0 24th January 2009 09:57 AM
a new horn speaker design in 3D angeloitacare Multi-Way 3 7th January 2009 10:59 PM
how to go from horn response to a folded bass horn design? paulspencer Subwoofers 8 4th November 2005 10:44 PM


New To Site? Need Help?

All times are GMT. The time now is 03:02 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2