A midrange horn project - opinions and proposals
Printed From: Speakerplans.com
Category: Plans
Forum Name: New Projects Forum
Forum Description: Forum for new speakerplans projects, in memory of Tony Wilkes, 1953 - 2014
URL: https://forum.speakerplans.com/forum_posts.asp?TID=19203
Printed Date: 27 March 2026 at 11:09am
Software Version: Web Wiz Forums 12.08 - https://www.webwizforums.com
Topic: A midrange horn project - opinions and proposals
Posted By: technofreak
Subject: A midrange horn project - opinions and proposals
Date Posted: 24 August 2008 at 2:15pm
|
First, hello everyone. I've been reading this forum for quite some time, but now I felt it's time to register. So, after reading Klipsch's paper on loudspeaker distortion and doing some simplified tests myself, I decided that horns are the way to go. I've bought a pair of Eminence Alpha 6 (at the moment playing in my bedroom on small OB-s and already very pleasing:) ). 300 Hz conical flare is on paper (or actually on hard drive). But before I incarnate this project any opinions and proposals are welcome, to further refine the design. Also, I have not enough time and money for extensive prototyping, so I'm trying design principles that (according to information found after extensive data digging on Internet) are proven to work and keep it as simple as possible. So, here are the details: Primary goal is as good sound as possible, ear-piercing SPL is secondary. 9x9 cm rectangular throat with phase plug (4 cm diameter initially, so throat area 68cmˇ2, compression ratio 1,86. Phase plug final shape and size is determinated by some experimentations) 1600 cmˇ2 mouth (40x40 cm), terminated with a foam or other similar material roundover to ~50x50 cm to smooth out the response and reduce mouth reflections back to throat. Horn lenght 31,2 cm (36,2 inc. roundover) Back chamber as small as possible, no larger than 2 litres. Intended bandwith is 300 to 3000 Hz, or as high as possible. 2...2,5 kHz will do it also, but not below 2 kHz. Fairly large throat gives me a chance to experiment with various phase plug shapes and sizes. Dispersion is approx. 50 degrees. Tweeter coaxially mounted in the mouth, at first trying to reuse my ol'trusty piezo array (a là Bill Fitzmaurice), compression driver with horn if that doesn't work well enough. They will be used in a multitude of roles - as my main hi-fi, for house parties, some band practice and occasional discos and lives, mainly closed parties in smaller venues up to 200 persons max. Proper bass support is to come next year (at the moment I have to stick with some "leftovers" - two 2x8", which are enough for home and band practices), for time being I can always rent (some Altec 816 style midbass bins that are around here would to the job). This is my hobby, just for fun and self-education and I'm not planning to make any money from it (yet). A concept image http://img380.imageshack.us/my.php?image=alpha6hornss3.jpg">  And Hornresp graph http://img98.imageshack.us/my.php?image=graphou6.jpg">  Schematic diagram http://img380.imageshack.us/my.php?image=schematicft0.jpg">  So, all opinions and proposals are welcome  |
Replies:
Posted By: Deadbeat
Date Posted: 24 August 2008 at 3:22pm
Looks fine and dandy. I like the fact that you have taken in mouth termination,  . I like to cut foam to the size of the horn and plug it actually. Another mouth termination technique is to double up the ends and route a radius in the wood. But I think for the most part you will be fine, saved by the small flare at the end.If you're even more worried, it is quite simple to rework the design so it is not axisymmetric, as this can tend to have more mouth reflection (and standing wave blah blah) problems etc (besides, vertical coverage is fine at 40 degrees or less). A typical falling response, it won't look so bad in real life. Do you have the hornresp inputs and have you checked that the system does not reach xmax? Have you also included the area immediately in front of the driver as front chamber? Also, have you considered using other horn flares such as tractrix/exponential/JMMIC? Another route you could go down is biradial/smith horn. ------------- Away on extended leave. |
Posted By: technofreak
Date Posted: 24 August 2008 at 4:39pm
|
I've done some research on horns and am aware of various methods to terminate mouth correctly. Again, roundover is the simplest and most accessible method. I got the foam idea from reading about an experiment where you make an improvised horn fom a rolled up magazine, talk into it and compare the sound with and without mouth termination made of a towel that is rolled up and wrapped around mouth. I've considered non-axisymmetric construction, but I don't want to make it wider - it's roughly 50x50x50 cm cube and seems to be optimal as far as doors, pole mounting (a compromise solution with horn on the pole and one or two 15" subs covering 50...300 Hz on the floor) and stacking goes (I live on the 8th floor and have to consider elevators). Also, my piezo array won't fit if I make the cab lower. I also considered tractrix etc but conical is the simplest (and arguably sonically one of the best) flare rates, also gives good control over dispersion so I'll go with this. I've double-checked the excursion figures, and with smaller than 2 litre back chamber it's OK even without crossover. Front chamber volume is estimated to be 100 cmˇ3 or less. Again, experimantation with phase plugs will be done. Construction shall begin in September. I'll try to get a camera and post some pics of the process. Also I'll try to make some plans up and post them here. Actually one more question about group delay - it always looks quite horrible with any horn on Hornresp graphs, but I quess it's not really relevant, right? |
Posted By: Deadbeat
Date Posted: 24 August 2008 at 5:08pm
|
So you have read Jack Bouska's excellent essay on his 'factors determining the sonic quality of what i can't remember', I see. The foam plug I proposed actually has come from the geddes HOMless plug design (which I think a few of us, including me stumbled across a while ago, but did not know of it's higher order mode killing abilities). The thing about the towel thing is that the stuff is actually in the horn, if you continued the flare of a horn, with let's say everyone's favourite choice acoustic foam (30ppi open cell), the diffraction (and other effects of a sharp edged mouth as you know) around the side when you transition between materials which you're trying to control with the termination makes the foam much much less effective than simply stapling it to the sides of the horn itself. This technique of lining the horn has been used successfully several times by people other than myself and works very well, is second to carving a plug. Unless I've misread you and you're just lining the horn with foam, as opposed to continuing it. 'Arguably sonically the best'. I could go into an appraisal of HF horn flares, but I'm not going to.  . Don't tempt me.... Post the group delay graphs - remember to not ignore any monsters. It really matters when you have a very large peak or dip somewhere. ------------- Away on extended leave. |
Posted By: technofreak
Date Posted: 24 August 2008 at 5:52pm
|
Here's the GD graph with 2 L back chamber (black) and 0,5 L chamber (light grey): http://img240.imageshack.us/my.php?image=groupdelaypp3.jpg">  At first yes, my idea was to extend the mouth with foam or something similar, but guess I'll make that last bit of flare from plywood and cover with foam. Or maybe try to get some foam like used for air filtering in amp air intakes and such and make a plug of several layers that also doubles as speaker grill (and also hides those "nasty piezos" from snobbish eyes:D)? Then, another idea - to extend the horn with eg. styrofoam and apply some acoustic foam over it and also ~10 cm into the main flare? |
Posted By: _djk_
Date Posted: 25 August 2008 at 8:49am
|
http://www.frazierspeakers.com/download/c40.pdf - http://www.frazierspeakers.com/download/c40.pdf
http://www.frazierspeakers.com/download/c79a.pdf - http://www.frazierspeakers.com/download/c79a.pdf ------------- djk |
Posted By: technofreak
Date Posted: 25 August 2008 at 5:51pm
|
Thanks for the info, _djk_, those look interesting boxes (although foxit reader goes bananas when trying to scroll daown the first PDF). Anyway, I threw some plans together (my CAD skills are not the best, actually these are the first CAD plans I've ever made): http://www.speedyshare.com/952908459.html - http://www.speedyshare.com/952908459.html . |
Posted By: _djk_
Date Posted: 26 August 2008 at 10:33am
|
I wanted you to see their use of foam, they have been doing it this way for about 20 years now.
The CAT-40 is about the size you are building, but driven with an 8" to 2Khz, and the midrange response looks quite smooth.
Here is a little better picture of one of their speakers.
 This is one of my early efforts, circa 1981:
 It's similar in size to the CAT-79, and has the break-back on the mouth that is at about 2/3rds the width.
------------- djk |
Posted By: biotec
Date Posted: 26 August 2008 at 10:55am
|
have a look at the OAP audio coaxial horns. They use dense foam on the end of the 10inch horn and have a very nice TAD 2inch coaxially mounted.
The 1x2" + 1x10" + 2x15" box they do suprised me with how nice it sounded. ------------- me so horny, me love you long throw. |
Posted By: Deadbeat
Date Posted: 26 August 2008 at 11:03am
|
djk, is the horn 'extended' by the foam, or does it line the horn? And is it open or closed cell? I can't tell, sorry.
------------- Away on extended leave. |
Posted By: Hi_Varu
Date Posted: 26 August 2008 at 11:05am
| the second immage isn't jeff robinson horn? |
Posted By: Deadbeat
Date Posted: 26 August 2008 at 11:31am
|
No, djk designs his horns. And...those mids are massive  ------------- Away on extended leave. |
Posted By: _djk_
Date Posted: 03 September 2008 at 6:40am
|
"is the horn 'extended' by the foam, or does it line the horn? And is it open or closed cell? I can't tell, sorry. "
The foam is 30-PPI open cell and there is a solid wall of the horn behind it.
"No, djk designs his horns.
And...those mids are massive " About 48" wide. They had a space frame (like the newer JBL HLA rigs), so they weren't heavy, just big.
The first pair of those I built started out with dual 15" drivers, this was OK as I was 4-way with phenolic compression drivers. Later, I changed them to 12s so they could go high enough to go 3-way with metal compression drivers. ------------- djk |
Posted By: technofreak
Date Posted: 03 September 2008 at 8:53am
| OK, seems like my ideas would work, right? So, after I get my salary, I´ll start building ASAP and post some pics here too. |
Posted By: tb_mike
Date Posted: 03 September 2008 at 8:58pm
| technofreak - it looks as if youve read the good papers and the best thing to do now is to build :-D! Did you see JBL CMCD(?) technote regarding direct radiator vs 6" cone mid. |
Posted By: doober
Date Posted: 03 September 2008 at 10:08pm
Can someone post the link to this article? I'm sure I saw it recently but I can't find it now. Kieran ------------- Blahblahblah |
technofreak wrote:Posted By: Deadbeat
Date Posted: 04 September 2008 at 8:43am
|
Jack Bouska on lansingheritage: http://www.audioheritage.org/vbulletin/showthread.php?t=12967 - http://www.audioheritage.org/vbulletin/showthread.php?t=12967 ------------- Away on extended leave. |
Posted By: doober
Date Posted: 04 September 2008 at 4:03pm
|
Thanks Deadbeat, that'll keep me entertained this evening. ------------- Blahblahblah |
Posted By: technofreak
Date Posted: 25 September 2008 at 6:37pm
|
Hi all, Due to the economy going south here I have suspended the horn project (this would have cost me ~200€ for materials and carpenter, since I don't have tools or room for such things) for time being. But I certainly WILL build them one day! For now I put the Alphas in small sealed boxes and called it a day. With those old double-8" they make great compact DJ monitors, micro rig for house parties and very loud hi-fi speakers - played the system at full blast yesterday and that was ridiculous in my 20 sq. meters room, I actually started to feel bass with my hair! The amp is 200W/ch at 4 ohms. http://picasaweb.google.com/tehnofriik - Here are some pics of the stuff. I used a cheap 2€ computer mic to get some idea of the freq. response - pretty flat from 50 Hz to 10 kHz, starts to roll off after that. I don't know what to blame more - the mic or the piezo tweeter, probably both are equally quilty. Slight EQ-ing of the top octave is required. The piezos are those 1€ things, slightly modified - added some wadding behind the diafragm, of course proper filter is applied (8,2 ohm 10W resistor across terminals, 1,5 uF cap in series and another 8,2 ohm 10 W resistor in series with all this). No harshness at 5 kHz. Alpha is rolled off at 5,8 kHz w/ 0,22 mH inductor and high-passed w/ 68 uF cap. Eight-inchers are low-passed w/ 3 mH inductor. It all works pretty well, though using 2. order x-overs would be better - the excursion of Alphas gets quite high, although doesn't exceed X-max. I deliberately left so much room around the three sides of piezo - so I can fit some smaller proper HF horns if I wanted to. ------------- http://www.youtube.com/watch?v=53Zq1I5_WAA - It's techno and it's freak |
Posted By: tommysb
Date Posted: 25 September 2008 at 11:54pm
|
Good to see. I love your Burberry Computer chair, and huge vodka bottle under the desk! Are all women in Estonia as beautiful as I've been lead to believe? Tom |
Posted By: technofreak
Date Posted: 26 September 2008 at 11:37am
This chair is comfortable as hell  . The vodka bottle wasn't emptied by me or my friends, we found it in a forest during a hike. No idea why we carried it all the way, it was empty, you know Must be crazy green thinking people... And yes, women here are really beautiful ------------- http://www.youtube.com/watch?v=53Zq1I5_WAA - It's techno and it's freak |