I think you're right, also the single horn expansion model from the table seems to array perfectly every 19 degrees as intended so I might skip trying to scale the drawing
@AllenB I'm making a 3D model in Fusion. I think I'll just have to 'average' out some of the dimensions as there are some details on the plan that don't make sense such as it showing one end of the fin beginning at the R=34.5mm arc and ending at the R=232 arc which would mean a fin length of 197.5mm not 192mm
I used CAD CNC to build my 290’s. I used 1mm increments to create the large curve. Ideally, I’d use a 3D/multi-axis router. I also CNC’d the throat housing, but used putty/bog to create a smooth square-to-round transition. I used Radian 950 beryllium drivers, crossed at 500Hz with LR2 xover. My custom bass horn uses Altec 416.
Attachments
There are conflicting dimensions that cannot be completely reconciled. I picked some that were obvious and worked together and fitted a smooth spline between them. I wouldn't get too hung up on it.
Here's a recent picture of the Athos Audio Yuichi A290 as a real-world reference. It's definitely large, 25.8" wide, and 11" high. Plenty heavy, too, maybe 25 lbs, minus the machined aluminum 1.4" circular to 2" square adapter, and the 1.4" compression driver.
Yes, that's an original EMI SQ-encoded DSotM. I used this record to demo the Audionics Shadow Vector quadraphonic decoder at the 1975 CES in Chicago, so a bit of history there.
Yes, that's an original EMI SQ-encoded DSotM. I used this record to demo the Audionics Shadow Vector quadraphonic decoder at the 1975 CES in Chicago, so a bit of history there.
Last edited:
Very nice! Thank you for sharing. Have you posted more details? Did you EQ the Radian? I would be particularly interested in waterfall measurements of this midrange driver/ horn combo…
I translated the parts marked in blue.
……part of the mark……
このように、ホーンの断面積は小ホーン5個分ではありますが、最も外側は、少しでも自由空間と馴染むようにわずかに曲面で拡大してあります。実際的な効果は少ないと思いますが、見た目も直線よりは曲面の方が柔らかく、美しく見えます。
……Translate……
In this way, the cross-sectional area of the horn is the same as that of five small horns, but the outermost part is slightly curved so that it blends in with the free space. I think the actual effect is small, but the curved opening looks softer and more beautiful than the straight opening.
Attachments
That's very interesting, I initially used 950PB with AL diaphragms on my clone A-290s, and had a lot of problems, huge dip at 1kHz and a pretty ragged response, both JBL2440 with 2441 diaphragms and TAD-4001 measure flatter and sound better. There was a lot of discussion about driver exit angles and horn entrance angles.
Note there is a level mismatch between the 2440/2441 - it is not significantly more sensitive than the others.
Edit: There is something going on in these plots at ~4kHz, which was identified as a resonance in the safety shield of my mini-lathe 20 feet away. I successfully eliminated it a few weeks ago. (See my post below)
Last edited:
Hi,guys
I would like to report on the wood horn I made five years ago, hoping that it will lead to someone's inspiration.
I'm in the process of building a 4-way speaker system using JBL130, BMS 4592ND, JBL LE 85, and JBL2405.
BMS 4592ND is combined with Yuichi’s wood horn A-290 made from Burmese Rosewood Laminated Lumber , and BMS 4592ND does not use a high frequency driver unit . EL85 is also combined with Yuichi’s wood horn A-480 made from Burmese Rosewood Laminated Lumber.
The major improvement theme at the moment is the crossover area between the 2405 and the LE85, aiming to match the sense of energy of the sound emitted by each unit. It can be paraphrased as wanting to match the clarity of sound.
The improvement approach I took for the A-480 was the following two points.
・Replace with materials that provide a sense of resolution.
・Improve the high-frequency side characteristics of the wood horn A-480 that is combined with the LE85.
I tried to replace the throat adapter from Burmese Rosewood to brass to replace it with a material that gives a sense of resolution. With this replacement, the nuances of the sound derived from the brass material were added, but the sense of resolution increased, and the difference in sound quality with the 2405 was no longer bothering me.
Next, as an approach to further enhance the characteristics on the high frequency side, we challenged to idealize the shape of the upper and lower surfaces of the A-480 where the fins are attached.
For wood horns with fins such as A-290 and A-480 of the original design , the surface to which the fins are attached is made flat to reduce the difficulty of work. However, the upper and lower plates sandwiching the fins form parallel planes. There is a concern that such parallel planes in the path of sound may act as a low-pass filter. Or maybe boost standing waves.
A wood horn without fins and a wood horn with fins have a big difference in the degree of freedom of work behind the horn. I thought that if this shape was improved, it would be possible to obtain higher fidelity playback.
I tried to eliminate the element of low-pass filtering by making the parallel plane formed by the upper and lower plates starting from the throat adapter joint surface curved.
This is my personal opinion, but even if the transition of the cross-sectional area slightly deviates from the design value, it is allowed as long as the difference is not large. I adjusted the dimensions by prioritizing the connection of smooth curved surfaces rather than the accuracy of the numerical values. As a result, the fin size was longer and thinner than the standard design value, requiring careful processing of the bonding surface between the upper and lower surfaces, which was extremely difficult. I wonder how easy it would be to work with a CNC. All my work is done by hand.
In addition, unnecessary low frequencies were cut off because the goal was to achieve high-fidelity reproduction in the high frequencies. Specifically, we changed Fc from 480 to 600Hz to make it smaller. I think that it can be used at frequencies above about 1000Hz.
The finished horn was named S-600.
The match with the J BL2405, which was a goal we had raised from the beginning of the design, was excellent, and the smoothness of the connection between the sound quality and the sense of energy, which did not feel a seam, was satisfactory at a high level.
However, the affinities obtained between A-290 and A-480 were not obtained. It can be said that the problem in the high range of the A-480 was just shifted to the low range side.
From now on, my quest will continue and I'll make some progress in the next installment.
thank you.
santoukun
I would like to report on the wood horn I made five years ago, hoping that it will lead to someone's inspiration.
I'm in the process of building a 4-way speaker system using JBL130, BMS 4592ND, JBL LE 85, and JBL2405.
BMS 4592ND is combined with Yuichi’s wood horn A-290 made from Burmese Rosewood Laminated Lumber , and BMS 4592ND does not use a high frequency driver unit . EL85 is also combined with Yuichi’s wood horn A-480 made from Burmese Rosewood Laminated Lumber.
The major improvement theme at the moment is the crossover area between the 2405 and the LE85, aiming to match the sense of energy of the sound emitted by each unit. It can be paraphrased as wanting to match the clarity of sound.
The improvement approach I took for the A-480 was the following two points.
・Replace with materials that provide a sense of resolution.
・Improve the high-frequency side characteristics of the wood horn A-480 that is combined with the LE85.
I tried to replace the throat adapter from Burmese Rosewood to brass to replace it with a material that gives a sense of resolution. With this replacement, the nuances of the sound derived from the brass material were added, but the sense of resolution increased, and the difference in sound quality with the 2405 was no longer bothering me.
Next, as an approach to further enhance the characteristics on the high frequency side, we challenged to idealize the shape of the upper and lower surfaces of the A-480 where the fins are attached.
For wood horns with fins such as A-290 and A-480 of the original design , the surface to which the fins are attached is made flat to reduce the difficulty of work. However, the upper and lower plates sandwiching the fins form parallel planes. There is a concern that such parallel planes in the path of sound may act as a low-pass filter. Or maybe boost standing waves.
A wood horn without fins and a wood horn with fins have a big difference in the degree of freedom of work behind the horn. I thought that if this shape was improved, it would be possible to obtain higher fidelity playback.
I tried to eliminate the element of low-pass filtering by making the parallel plane formed by the upper and lower plates starting from the throat adapter joint surface curved.
This is my personal opinion, but even if the transition of the cross-sectional area slightly deviates from the design value, it is allowed as long as the difference is not large. I adjusted the dimensions by prioritizing the connection of smooth curved surfaces rather than the accuracy of the numerical values. As a result, the fin size was longer and thinner than the standard design value, requiring careful processing of the bonding surface between the upper and lower surfaces, which was extremely difficult. I wonder how easy it would be to work with a CNC. All my work is done by hand.
In addition, unnecessary low frequencies were cut off because the goal was to achieve high-fidelity reproduction in the high frequencies. Specifically, we changed Fc from 480 to 600Hz to make it smaller. I think that it can be used at frequencies above about 1000Hz.
The finished horn was named S-600.
The match with the J BL2405, which was a goal we had raised from the beginning of the design, was excellent, and the smoothness of the connection between the sound quality and the sense of energy, which did not feel a seam, was satisfactory at a high level.
However, the affinities obtained between A-290 and A-480 were not obtained. It can be said that the problem in the high range of the A-480 was just shifted to the low range side.
From now on, my quest will continue and I'll make some progress in the next installment.
thank you.
santoukun
Attachments
@santohkun
At what frequency will you cross from the A-290 to A-480? I was crossing at 10kHz from the A-290/TD-4001 to FaitalPro STH100/HF10AK, I recently changed to a much lower crossover point of 5kHz for evaluation, so far I think I like the results better at the lower crossover point.
At what frequency will you cross from the A-290 to A-480? I was crossing at 10kHz from the A-290/TD-4001 to FaitalPro STH100/HF10AK, I recently changed to a much lower crossover point of 5kHz for evaluation, so far I think I like the results better at the lower crossover point.
Thank you! 
I started with LR48, switched to LR24 and ended with Bessel 24, currently crossing the mains at 620Hz, and 6.3kHz. Bass bins are diy Onken boxes with Iconic 165-8G (AKA GPA 515-8G). Subs 17Hz - 45Hz with Butterworth 24dB.
It's all one big experiment with a lot of learning along the way, and a lot more to learn at a guess.
I started with LR48, switched to LR24 and ended with Bessel 24, currently crossing the mains at 620Hz, and 6.3kHz. Bass bins are diy Onken boxes with Iconic 165-8G (AKA GPA 515-8G). Subs 17Hz - 45Hz with Butterworth 24dB.It's all one big experiment with a lot of learning along the way, and a lot more to learn at a guess.
Last edited:
I have a Yuichi A290, and need to know what problems (issues) may arise when I use a CD that has an exit angle that isn't in the narrow range that the A290 was designed for ( 8 to 10 degrees).
The CD I am thinking about using is a B & C DCM 50. Due to the particular design of the phase plug on the DCM 50 there is no exit angle.
(0 degree exit angle)
The CD I am thinking about using is a B & C DCM 50. Due to the particular design of the phase plug on the DCM 50 there is no exit angle.
(0 degree exit angle)
I own a pair of Yuichi A-290 clones and did some response measurements 7.5cm from the mouth of the horn. Others are more expert than me, but with a Radian 950PB I had a huge notch right at 1kHz, shorter throat, and wider exit angle than the recommended drivers. I would recommend sticking to one of the drivers known to work with the Yuichi, there are several JBL and other options discussed by other members that work well.
No EQ, no crossover.
Note: I replaced the chart with one expressed in dBSpl rather than dBFS. The JBL is a 2440 fitted with 2441 diaphragms. SPL level for JBL is higher due to higher power being applied. Efficiency is relatively similar between these drivers.
I was able to EQ out most of the hole at 1kHz with the Radian driver, but it never sounded quite right which led to testing with the 2441 and finally the purchase of a pair of late 1980s vintage TAD TD-4001 drivers.
No EQ, no crossover.
Note: I replaced the chart with one expressed in dBSpl rather than dBFS. The JBL is a 2440 fitted with 2441 diaphragms. SPL level for JBL is higher due to higher power being applied. Efficiency is relatively similar between these drivers.
I was able to EQ out most of the hole at 1kHz with the Radian driver, but it never sounded quite right which led to testing with the 2441 and finally the purchase of a pair of late 1980s vintage TAD TD-4001 drivers.
Last edited:
Drivers with zero degree exits can work well enough in any horn or waveguide, it is much more problematic to have a wide opening conical section right at the throat of a horn designed to open up slowly, kevin's measurement shows what can happen when there is a severe mismatch in the wrong direction.