A few months ago, there were a few threads that rekindled my interest in radial horns and the Yuichi A-290 in particular. I wanted to make a BEM sim of the A-290 to see it’s directivity and if there were any simple modifications that might result in an improvement. I thought it would be a simple task of transferring the values from the document into a CAD drawing. This turned out not to be the case as the values from the drawings did not create a smooth curve for the horizontal sides. I chose to use a conic curve with a visual best fit between the points in Fusion. The radius dimensions to align the flat section with the fin ends and vertical expansion didn’t quite line up and I decided to fudge it to make it work. The dimensions for the fins produced reasonable curves without any intervention. The original file had the sides being quite flat before curving out as this seemed to match with images of completed horns, I had seen.

I was able to make a mesh and get a basic simulation going. There were a few problems with the model but overall, the directivity seemed to be reasonably constant in the horizontal as touted. The Radiation Impedance was very spiky and odd looking and looked to me like it could be from simulation errors. I knew @DonVK had simulated some similar horns with @docali so I got in touch with Don to have a look and see if he could offer some advice on what I was seeing. It turned out that there wasn’t too much wrong and that the Rad Imp was reasonably consistent with other horns of the same type.

This sparked off several different simulations to improve the model and look to improve the horn without making it into something completely different. Don made all the model iterations. Ideas on fin placement and number from Kolbrek and Dunker’s book were tried along with asymmetric, golden ratios and some other variations. The results iteratively got better and in tandem docali’s Radial Fin spreadsheets were reworked by him and simulated by Don looking for improvements. The horn’s resulting from docali’s spreadsheets surpassed the best attempts at improving the A-290. As a result of lessons learned there, I took another look at the A-290 documents and reworked the CAD from scratch. I drew three-point arcs based on defined coordinates from the drawings for the side profiles and used the data taken from the excel spreadsheet calculations (included in the A-290 doc) for the vertical curves. This time the fin ends lined up exactly at the end of the flat section and the overall result was much cleaner. The best result for a horn based off Yuichi’s plans used 6 fins with a spacing of 7.5,15,15,15,7.5 degrees. The outer two channels being half fins between the fin and wall.

It is still not clear to me if there is a definitive way to produce an exact A-290 as Yuichi intended. What you can find in the attached step and Fusion archives is the blueprint to build either of the versions I created with simulated performance of what to expect.

The V1 is a full 3D solid model, the V2 is a surface model that needs further work to turn it into a full 3D solid for production. Results of selected models to follow in further posts.

Greater simulated performance can be had from similar horns and the information may be released in another thread describing that process or via docali’s sphericalhorns.net website, currently those experiments are still ongoing.

I was able to make a mesh and get a basic simulation going. There were a few problems with the model but overall, the directivity seemed to be reasonably constant in the horizontal as touted. The Radiation Impedance was very spiky and odd looking and looked to me like it could be from simulation errors. I knew @DonVK had simulated some similar horns with @docali so I got in touch with Don to have a look and see if he could offer some advice on what I was seeing. It turned out that there wasn’t too much wrong and that the Rad Imp was reasonably consistent with other horns of the same type.

This sparked off several different simulations to improve the model and look to improve the horn without making it into something completely different. Don made all the model iterations. Ideas on fin placement and number from Kolbrek and Dunker’s book were tried along with asymmetric, golden ratios and some other variations. The results iteratively got better and in tandem docali’s Radial Fin spreadsheets were reworked by him and simulated by Don looking for improvements. The horn’s resulting from docali’s spreadsheets surpassed the best attempts at improving the A-290. As a result of lessons learned there, I took another look at the A-290 documents and reworked the CAD from scratch. I drew three-point arcs based on defined coordinates from the drawings for the side profiles and used the data taken from the excel spreadsheet calculations (included in the A-290 doc) for the vertical curves. This time the fin ends lined up exactly at the end of the flat section and the overall result was much cleaner. The best result for a horn based off Yuichi’s plans used 6 fins with a spacing of 7.5,15,15,15,7.5 degrees. The outer two channels being half fins between the fin and wall.

It is still not clear to me if there is a definitive way to produce an exact A-290 as Yuichi intended. What you can find in the attached step and Fusion archives is the blueprint to build either of the versions I created with simulated performance of what to expect.

The V1 is a full 3D solid model, the V2 is a surface model that needs further work to turn it into a full 3D solid for production. Results of selected models to follow in further posts.

Greater simulated performance can be had from similar horns and the information may be released in another thread describing that process or via docali’s sphericalhorns.net website, currently those experiments are still ongoing.